United States
      Environmental Protection
      Agency
Office of Water (4303)
Washington, DC 20460
EPA-821-B-01-006
February 2002
v>EPA Economic Analysis of Proposed
       Effluent Limitations Guidelines
       and Standards for the Meat and
       Poultry Products Industry

-------

-------
Economic Analysis of Proposed Effluent Limitations
           Guidelines and Standards for the
         Meat and Poultry Products Industry
                    Christine Todd Whitman
                        Administrator

                        Tracy Mehan
              Assistant Administrator, Office of Water

                        Sheila E. Frace
             Director, Engineering and Analysis Division

                        Samantha Lewis
                       Project Manager

                    William Wheeler, Ph.D.
                          Economist
                 Engineering and Analysis Division
                 Office of Science and Technology
               U.S. Environmental Protection Agency
                    Washington, D.C. 20460
                        February 2002

-------

-------
                 ACKNOWLEDGMENTS AND DISCLAIMER
This document was prepared with the support of Eastern Research Group, Incorporated under
Contract 68-C-01-073.

Neither the United States government nor any of its employees, contractors, subcontractors, or other
employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for
any third party's use of, or the results of such use of, any information, apparatus, product, or process
discussed in this report, or represents that its use by such a third party would not infringe on privately
owned rights.

-------

-------
                                     CONTENTS
                                                                                      age
FIGURES	viii
TABLES	:.'	 ix

EXECUTIVE SUMMARY

       ES.l  Background	ES-1
       ES.2  Industry Overview	ES-1
       ES.3  Data Sources	'.	ES-2
       ES.4  Economic Impact Methodology  .	ES-3
       ES.5  Results	 .  ES-6

             ES.5.1 Regulatory Options and Costs  	ES-6
             ES.5.2 Impacts	ES-12
             ES.5.3 Small Business Impacts''.	'.	 - ES-12
             ES.5.4 Environmental Benefits	ES-16

       ES.6  References	-	ES-18


CHAPTER 1        INTRODUCTION

       1.1     Scope and Purpose	 1-1
       1.2     Data Sources .	 1-2
       1.3     Report Organization 	,	 1-7
       1.4     References	 1-8


CHAPTER 2       INDUSTRY PROFILE

       2.1     Industry Overview Based on Census Data .  . .	 2-1

              2.1.1  Industry Sectors . .	 2-2

                    2.1.1.1 Animal (Except Poultry) Slaughtering: NAICS code 311611  	 2-2
                    2.1.1.2 Meat Processed from Carcasses: NAICS code 311612 ...	... 2-6
                     2.1.1.3 Poultry Processing: NAICS code 311615	   2-11
                     2.1.1.4 Rendering and Meat Byproduct Processing: NAICS code 311613 .   2-13

              2.1.2   Sector Overview	'.	   2-21

                     2.1.2.1 Relative Industry Shares	   2-21
                     2.1.2.2 Geographic Distribution of Industry 	   2-23

-------
                             CONTENTS (cont.)


2.2    Screener Survey and Subcategorization  .	  2-26

       2.2.1   Meat Type and Process Classes	  2-26

               2.2.1.1 Method of Classification	  2-26
               2.2.1.2 Facility Count by Class, Discharge, and Size	  2-27

       2.2.2   Proposal 40 CFR 432 Subcategories	  2-28

               2.2.2.1 Method of Subcategorization	•. . .  2-28
               2.2.2.2 Facility Count by Subcategories	  2-32

2.3    Trends in Production, Prices, and International Trade 	  2-34

       2.3.1   Aggregate Industry Trends	  2-34

               2.3.1.1 Domestic Production and International Trade Trends  	   2-34
               2.3.1.2 Price Trends	,	  2-40

       2.3.2   Industry Response to Changing Consumer Preferences	  2-40

2.4    Industry Concentration	•	  2-45

       2.4.1   Trends in Industry Concentration		  2-46
       2.4.2   Facility Size and Economies of Scale	  2-61
       2.4.3   Industry Concentration and Market Power	  2-63

2.5    Profile of Industry Leaders	  2-7G

       2.5.1   Beef Slaughtering Operations	  2-71
       2.5.2   Hog Slaughtering Operations     			  2-79
       2.5.3   Poultry Slaughtering Operations	  2-86

               2.5.3.1 Broiler Companies		  2-92
               2.5.3.2 Turkey Companies	  2-97

       2.5.4   Overall Ranking of Meat Processing Companies	•;	2-100

2.6    References	2-105

-------
                                   CONTENTS (cont.)
                                                                                          'age
CHAPTERS        ECONOMIC IMPACT METHODOLOGY

       3.1    Methodology for the Proposed Rule .	  3-2

              3.1.1   Cost Affliualization Model	• • •	  3-2
              3.1.2   Facility-Level Impact Analysis	  3-5

                     3.1.2.1  Overview of Basic Model Framework	  3-6
                     3.1.2.2  Development of Model Facility Income Measures	  3-7
                     3.1.2.3  Distribution of Income Represented by Model Facilities	 3-11
                     3.1.2.4  Use of Model Facility and Distribution to Project
                             Closure Impacts	 3-14
                     3.1.2.5  Negative Baseline Facility Income	 3-19
                     3.1.2.6  Matching Economic Model Facilities to Engineering Model
                             Facilities		 3-23

               3.1.3  Financial Ratio Analysis  .	••	 3-26

                      3.1.3.1  Return on Assets  	•  • •	 3-27
                      3.1.3.2 Corporate Financial Distress Analysis	 3-28

               3.1.4   MarketModel	 3-33

                     . 3.1.4.1 Market Model Approach	 3-36
                      3.1.4.2 Data Sources for Market Model Analysis	 3-38

               3.1.5   National Direct and Indirect Impacts  . .	 3-43

        3.2    Methodology for the Final Rule	 3-44

               3.2.1   Cost Annualization Model  . . .	 3-45
               3.2.2   Facility Closure Model		 3-46

                      3.2.2.1  Assumptions and Choices	 3-48
                      3.2.2.2  Present Value of Future Earnings  	,	 3-50
                      3.2.2.3  Projecting Site Closures As a Result of the Rule	 3-54
        3.3
References
                                                                                           3-56
                                               HI

-------
CHAPTER 4
                                  CONTENTS (cont.)
POLLUTION CONTROL OPTIONS
                                                                                      Page
       4.1     Effluent Limitations Guidelines and Standards	 4-1
       4.2     Technology Options		 4-2.
       4.3     References	 4-6


CHAPTERS         ECONOMIC IMPACTS

       5.1     Total and Average Compliance Costs	 5-2

              5.1.1   Total and Average Compliance Costs by Subcategory	 5-3

                     5.1.1.1  Upper-Bound Costs	 5-3
                     5.1.1.2  Upgrade Costs	 . 5-7

              5.1.2   Total and Average Compliance Costs by Class	   5-11

                     5.1.2.1  Upper-Bound Costs	  5-11
                     5.1.2.2  Upgrade Costs	  5-19

              5.1.3   Comparison of Upper-Bound and Retrofit Compliance Costs by Class ....  5-26


       5.2     Facility Closure Analysis .	 .  5-26

              5.2.1   Projected Closure Impacts by Subcategory	  5-32

                     5.2.1.1  Upper-Bound Cost Closure Impacts	  5-32
                     5.2.1.2  Upgrade Cost Closure Impacts	. . . .	•  5-36

              5.2.2   Projected Closure Impacts by Meat Type and Process Class	  5-40

                     5.2.2.1  Upper-Bound Cost Closure Impacts	  5-40
                     5.2.2.2  Upgrade Cost Closure Impacts  	  5-46

       5.3     Facility Nonclosure Impacts	  5-52

              5.3.1   Nonclosure Impacts by Subcategory	  5-54

                     5.3.1.1  Upper-Bound Cost Nonclosure Impacts	  5-54
                     5.3.1.2  Upgrade Cost Nonclosure Impacts	  5-58
                                             IV

-------
                                  CONTENTS (cont.)
              5.3.2  Nonclosure Impacts by Meat Type and Process Class	  5-59

                    5.3.2.1  Upper-Bound Cost Nonclosure Impacts	  5-59
                    5.3.2.2  Upgrade Cost Nonclosure Impacts	  5-69

       5.4     Financial Ratio Analysis	  5-76

              5.4.1  Financial Ratio Analysis by Subcategory .  . .	  5-76

                    5.4.1.1  Upper-Bound Cost Financial Ratio Analysis	5-76
                    5.4.1.2  Upgrade Cost Financial Ratio Analysis	5-79

              5.4.2  Financial Ratio Analysis by Meat Type and Process Class	  5-79

                    5.4.2.1  Upper-Bound Cost Financial Ratio Analysis	  5-79
                    5.4.2.2  Upgrade Cost Financial Ratio Analysis	 .  5-86

       5.5     Corporate Financial Distress	  5-91
       5.6     Market and Trade Impacts	  5-93
       5.7     Impacts on Output and Employment	  5-98
       5.8     New Sources	'. • •	5-101
       5.9     Summary and Observations	5-111
       5.10   References . .	•	• 5-114


CHAPTER 6  SMALL BUSINESS INITIAL REGULATORY FLEXIBILITY ANALYSIS

       6.1     Introduction	  6-1
       6.2     Initial Assessment . .	  6-1
       6.3     Regulatory Flexibility Analysis Components 	  6-2

              6.3.1   Need for Objectives of the Rule	  6-2
              6.3.2   Estimated Number of Small Business Entities to Which the Regulation
                     Will Apply	  6-3
              6.3.3   Description of the Proposed Reporting, Recordkeeping, and Other Compliance
                     Requirements	  6-5
              6.3.4   Identification of Relevant Federal Rules That May Duplicate, Overlap, or Conflict
                     with the Proposed Rule	•	r •	.'....  6-9
              6.3.5   Significant Regulatory Alternatives	- .  6-9

       6.4     Small Business Analysis	• • •  6-10

              6.4.1   Total and Average Compliance Costs	  6-14

-------
                                CONTENTS (cont.)
                                                                                 Page
                    6.4.1.1 Total and Average Compliance Costs by Subcategory	  6-15
                    6.4.1.2 Total and Average Compliance Costs by Meat Type
                          and Process Class	  6-21

             6.4.2   Closure Impacts	  6-26

                    6.4.2.1 Projected Closure Impacts by Subcategory	  6-31
                    6.4.2.2 Projected Closure Impacts by Meat Type and Process Class	  6-37

             6.4.3   Facility Nonclosure Impacts	  6-41

                    6.4.3.1 Nonclosure Impacts by Subcategory	 .  6-45
                    6.4.3.2 Nonclosure Impacts by Meat Type and Process Class	  6-52

       6.5    Regulatory Flexibility Analysis	;	  6-57
       6.6    References	  6-67


CHAPTER? ENVIRONMENTAL BENEFITS

       7.1    Benefit Valuation Methodology	'.	  7-1

             7.1.1   A Continuous Approach to Valuation	  7-2
             7.1.2   Use Category Approach to Valuation	  7-10

       7.2    Benefit Valuation Results		...  7-12
       7.3    References	  7-22


CHAPTERS COST-BENEFIT COMPARISON AND
             UNFUNDED MANDATES REFORM ACT ANALYSIS

       8.1    Cost-Benefit Comparison .	 . .	  8-1
       8.2    Unfunded Mandates Reform Act Analysis	  8-2
       8.3    References-	,	  8-3
APPENDIX A
COST ANNUALIZATION MODEL	A-l
APPENDIX B
FACILITY LEVEL ANALYSIS	:  B-l
                                          VI

-------
APPENDIX C
MARKET MODEL METHODOLOGY		 C-l
APPENDIX D
SUMMARY OF DEMAND AND SUPPLY ELASTICITY
LITERATURE	
                                                                D-l
APPENDIX E
SENSITIVITY ANALYSES	  	, . E-l
APPENDIX F
COST EFFECTIVENESS ANALYSIS	 F-l
APPENDIX G     SURVEY FORMS
                                 Ytt

-------
                                       FIGURES
Figure
2-1    Meat Products Industry: Percentage of Employment, Total Shipments, and Value Added by
       NAICS Sector	  2-22
2-2    Value of Meat Products Shipments and Meat Products as a Percentage of Shipments
       by State	i	:	  2-24
2-3    Meat Products Employment by State and Employment as a Percentage of State
       Employment	  2-25

3-1    Cost Annualizatkra Model	3-3
3-2    Baseline Distribution Function for Model Establishment Cash Flow  	  3-18
3-3    Impact of the Effluent Guideline on Market for Meat Product i 	  3-35

7-1    Cumulative Willingness to Pay for Changes in WQI, f(W)	 7-8
                                            vui

-------
                                        TABLES
Table
Pas
ES-1   Proposal 40 CFR Subcategories, Meat Type and Process Class,
       Discharge Type, and Size	ES-7
ES-2   Meat Products Industry Treatment Technology Options	 ES-8
ES-3   Total Estimated Compliance Costs (Upper-Bound & Retrofit by 40 CFR Subcategories  . .  . ES-9
ES-4   Summary of Impacts Under Proposed Options	• • • •	ES-13
ES-5   Projected Compliance Cost Impacts on Meat Product Markets
       Proposed Option: BAT 2 for Subcategory J, BAT 3 All Other Subcategories	ES-15
ES-6   Summary of Impacts Under the Proposed Options, Small Business Owned Facilities	ES-17

1-1    EPA Effluent Limitations Guidelines for Meat Products Industry	  1-3

2-1    1997 Animal Slaughter Industry: NAICS Code 311611 Statistics for Selected States	  2-4
2-2    1997 Animal Slaughter Industry: NAICS Code 311611 Statistics by Employment Size	  2-5
2-3    1997 Animal Slaughter Industry: NAICS Code 311611 Output by Selected Product Codes  . .  2-7
2-4    1997 Meat Processing Industry: NAICS Code 311612 Statistics for Selected States T ......  2-8
2-5    1997 Meat Processing Industry: NAICS Code 311612 Statistics by Employment Size	  2-10
2-6    1997 Meat Processing Industry: NAICS Code 311612 Output by Selected Product Codes .  .  2-12
2-7    1997 Poultry Processing Industry: NAICS Code  311615 Statistics for Selected States	  2-14
2-8    1997 Poultry Processing Industry: NAICS Code 311615 Statistics by Employment Size ....  2-15
2-9    1997 Poultry Processing Industry: NAICS Code 311615 Output by Selected Product Codes   2-16
2-10   1997 Rendering Industry: NAICS Code 311613 Statistics for Selected States	  2-18
2-11   1997 Rendering Industry: NAICS Code 311613 Statistics by Employment Size	  2-19
2-12   1997 Rendering Industry: NAICS Code 311613 Output by Selected Product	  2-20
2-13   Facility Count by Meat Type and Process Class,  Discharge Type, and Size  	,  2-29
2-14   Facility Count by Proposal 40 CFR 432 Subcategories, Discharge Type, and Size . .	.  .  2-33
2-15   Beef Production, Exports and Imports, 1980-2000	 :	  2-35
2-16   Pork Production, Exports andlmports, 1980-2000	 : . . .		  2-36
2-17   Broiler Production andExports, 1980-2000	  2-38
2-18   Turkey Production andExports, 1980-2000		  2-39
2-19   Consumer Price Index for All Items, Food and Meat, 1980-2000		  2-41'
2-20   Consumer Price Index for Meat Products, 1980-2000	  2-42
2-21   Annual Heifer and Steer Slaughter by Plant Size, 1972-1998		  2-48
2-22   Heifer and Steer Slaughter Plants by Plant Size, 1972-1998			  2-49
2-23   Concentration Ratios and Herfindahl-Hirshman Index for
       Steer and Heifer Slaughter, 1980-1998	 .	  2-51
2-24   Firms Performing Steer and Heifer Slaughter, Number of Plants Owned
       byFirmSize, 1980-1998	.2-52
2-25   Annual Hog Slaughter by Plant Size 1972-1998		 . .  .  2-53
2-26   Hog Slaughter Plants by Plants Size 1972-1998	.'.	  2-55
2-27   Concentration Ratios and Herfindahl-Hirshman Index for Hog Slaughter, 1980-1998 . .	  2-56
2-28   Firms Performing Hog Slaughter, Number of Plants Owned by Firm Size, 1980-1998  : . . .  .  2-58
2-29   Annual Poultry Production 1972-1995 . .	 . . . :	  2-59
                                              IX

-------
                                     TABLES (cont)
2-30   Concentration Ratios for Poultry Industry, 1963-1992	  2-60
2-31   Firms with Beef Slaughter Plants Ranked by 1999 Slaughter	  2-72
2-32   Firms with Pork Slaughter Plants Ranked by 1999 Slaughter	'.	.•.  2-80
2-33   Firms with Broiler Slaughter Plants Ranked by 1999 Slaughter	  2-87
2-34   Firms with Turkey Slaughter Plants Ranked by 1999 Slaughter	  2-88
2-35   Firms with Poultry Slaughter Plants Ranked by 1999 Slaughter	  2-90
2-36   Meat Processing Firms with 1999 Revenues Exceeding $250 Million	2-101

3-1    Model Facility Income Mean and Standard Deviation by Employment Class	  3-15
3-2    Price Elasticities of Supply and Demand Identified in Feedlots Literature Searches	  3-41
3-3    EPA Estimates of Armington Trade Elasticities with Respect to Domestic Price	  3-41

4-1    Meat Products Industry Treatment Technology Options	  4-3
4-2    Technology Options for Meat Product Industry Subcategories	  4-5

5-1    Total and Average Upper-Bound Costs, 40 CFR 432 Subcategories	  5-4
5-2    Total and Average Retrofit Costs, 40 CFR 432 Subcategories	  5-8
5-3    Total and Average Upper-Bound Costs, Meat Type and Process Classes	,	  5-12
5-4    Total and Average Retrofit Costs, Meat Type and Process Classes	  5-20
5-5    Comparison of Upper-Bound and Retrofit Capital Costs	  .	  5-27
5-6    Economic Closure Impacts: Upper-Bound Costs, 40 CFR 432 Subcategories  	  5-33
5-7    Economic Closure Impacts: Retrofit Costs, 40 CFR 432  Subcategories	  5-37
5-8    Economic Closure Impacts: Upper-Bound Costs, Meat Type and Process Classes	  5-41
5-9    Economic Closure Impacts: Retrofit Costs, Meat Type and Process Classes	  5-47
5-10   Nonclosure Impacts: Upper-Bound Costs, 40 CFR 432 Subcategories .	  5-55
5-11   Nonclosure Impacts: Retrofit Costs, 40 CFR 432 Subcategories	  5-60
5-12   Nonclosure Impacts: Upper-Bound Costs, Meat Type and Process Classes	  5-63
5-13   Nonclosure Impacts: Retrofit Costs, Meat Type and Process Classes	  5-70
5-14   Impacts to Return on Assets Ratio: Upper-Bound Costs, 40 CFR 432 Subcategories .......  5-77
5-15   Impacts to Return on Assets Ratio: Retrofit Costs, 40 CFR 432 Subcategories	  5-80
5-16   Impacts to Return on Assets Ratio: Upper-Bound Costs, Meat Type and Process  Classes . .  5-82
5-17   Impacts to Return on Assets Ratio: Retrofit Costs, Meat Type and Process Classes	  5-87
5-18   Altman Z' Results	  5-92
5-19   Estimated Compliance Cost per Pound of Output by Meat Type and Options	  5-95
5-20   Projected Compliance Cost Impacts on Meat Product Markets Proposed Option
       Scenario 1: BAT 3 Costs for Direct Dischargers Only with Cross Market Impacts,
       Armington Trade	 .  5-96
5-21   Projected Compliance Cost Impacts on Meat Product Markets Proposed Option
       Scenario 2: BAT 3 Costs for Direct and Indirect Dischargers with Cross Market Impacts,
       Armington Trade	  5-97
5-22   Output and Employment Impacts	  5-99
5-23   Ratio of Capital Costs to Total Assets Presented by Subcategory and Option	5-104
5-24   Ratio of Capital Costs to Total Assets Presented by Meat Class and Option	5-107

-------
                                    TABLES (cont)
5-25   Summary of Impacts Under the Proposed Options ................... ........... 5-112

6-1    Meat Product Industry Estimated Small Business Owned Facilities
       by 40 CFR 432 Subcategories  ........... . ..................................  6-6
6-2    Meat Product Industry Estimated Small Business Owned Facilities
       by Meat Type and Process Classes  ....................................... •  • •  6-7
6-3    Meat Product Industry Estimated Direct and Indirect Discharge Small Business
       Owned Facilities by 40 CFR 432 Subcategories  ........ ......... ...............  6-11
6-4    Meat Product Industry Estimated Direct and Indirect Discharge Small Business
       Owned Facilities by Meat Type and Process Classes  .... ................. ........  6-12
6-5A   Total and Average Costs: Small Model Facilities, 40 CFR 432 Subcategories  . ..........  6-16
6-5B   Total and Average Costs: Nonsmall Model Facilities Owned by Small Businesses,
       40 CFR 432 Subcategories  ............................... ................  6-18
6-6A   Total and Average Costs: Small Model Facilities, Meat Type and Process Classes  .......  6-22
6-6B   Total and Average Costs: Nonsmall Model Facilities Owned by Small Businesses,
       Meat Type and Process Classes ......... ................. ....... ..........  6-27
6-7A   Economic Closure Impacts: Small Model Facilities, 40 CFR 432 Subcategories ... ..... ,  .  6-32
6-7B   Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses,
       40 CFR 432 Subcategories  ........ ........... . ...........................  6-35
6-8A   Economic Closure Impacts: Small Model Facilities, Meat Type and Process Classes ......  6-38
6-8B   Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses,
       Meat Type and Process Classes ............. ..............................  6-42
6-9A   Nonclosure Impacts: Small Model Facilities, 40 CFR 432 Subcategories ...............  6-46
6-9B   Nonclosure Impacts: Nonsmall Model Facilities Owned by Small Businesses,
       40 CFR 432 Subcategories  ...................... '. ........................  6-49
6-10A Nonclosure Impacts: Small Model Facilities, Meat Type and Process Classes ....... ....  6-53
6-10B Nonclosure Impacts: Nonsmall Model Facilities Owned by Small Businesses,
       Meat Type and Process Classes  ........... ................................  6-58
6-11   Meat Product Industry Estimated Direct and Indirect Discharge Affected Small Business
       Owned Facilities by 40 CFR 432 Subcategories  ............. 1 ..................  6-63
6-12   Meat Product Industry Estimated Direct and Indirect Discharge Affected Small Business
       Owned Facilities by Meat Type and Process Class ..............................  6-64
6-13   Summary of Impacts Under the Proposed Options, Small and Nonsmall ...............  6-66

7-1    Applying WQI to Vaughn's Use Category Criteria ... ........... . ................ 7-4
7-2    Empirical Calculation of Criteria from the Baseline Scenario ............ ..... ; ...... 7-5
7-3    Comparison of Baseline Scenario Categorization under Most Restrictive
       Use and Mean + SD Criteria.  ........... .................................... 7-6
7-4    WTP Values for Changes in Use Category  ....... ....... . ....................  7-11
7-5    National Reach Use Category Changes from Alternative Scenarios ..................  7-13
7-6    Summary of Monetized Benefits  .............. . .......... .............. - - - -  7-14
7-7    Households and River Mileage Affected by State, Proposed Scenario 7 ...............  7-15
7-8    Households and Changes in WQI by State, Proposed Scenario 7 ....................  7-16

-------
                                     TABLES (cont)
7-9    Total Benefits by State, by Use Category Change Method	 7-17
7-10   Total Benefits by State, by Continuous Method	 . 7-19


-------
                                EXECUTIVE SUMMARY
ES.l   BACKGROUND

       The U.S. Environmental Protection Agency is proposing to revise subcategorization and effluent
limitations guidelines and standards for the meat products industry point source category. The current
meat products rule, 40 CFR Part 432, sets effluent guidelines and limitations for the beef, pork, and
rendering sectors of the meat products industry. These standards were set and revised over a number of
years, most recently in 1995. This proposed rule revises the existing subcategories in the industry as well
as guidelines for those subcategories, and proposes new standards for facilities that perform poultry
slaughter and processing operations. Prior to this proposed rule, EPA has set no national effluent
limitations guidelines or standards for poultry slaughterers or processors.

        With the exception of small processors (Subcategory E), EPA is proposing revisions to Best
Practicable Control Technology Currently Available (BPT), Best Available Technology Economically
Achievable (BAT), Best Conventional Pollutant Control Technology (BCT), and New Source Performance
Standards (NSPS) in Subcategories A through D (red meat facilities that perform slaughter operations),
Subcategories F through I (red meat facilities that process meat not slaughtered at the facility), and
Subcategory  J (rendering facilities). EPA is proposing to create two new subcategories (K and L) for
facilities that slaughter and process poultry, and to set BPT, BAT, BCT, and NSPS for these poultry
subcategories. EPA is not proposing any revisions to current guidelines and standards for indirect
dischargers in the red meat subcategories, nor is it proposing to set new standards for indirect dischargers
in the poultry subcategories.
 ES.2   INDUSTRY OVERVIEW
        The meat products industry includes establishments that primarily slaughter livestock and/or
 process meat into products for further processing or for final sale to consumers. The industry can be
 roughly divided into red meat facilities, primarily producing beef or pork products, and poultry facilities,
                                               ES-1

-------
which primarily produce chicken (excluding eggs) and turkey products.  (Red meat facilities may also
process lamb or veal. Poultry facilities may also process other birds, such as ducks and geese, and also
small game, such as rabbits.) Facilities may perform slaughtering operations, processing operations from
carcasses slaughtered at other facilities, or both. In addition, rendering operations may be performed either
at stand alone facilities, or in combination with slaughter and/or further processing operations. Companies
that own meat product facilities may also own facilities that perform "upstream" or "downstream"
operations involved in getting meat products from the farm to the consumer (e.g., livestock raising,
wholesale distribution), but these facilities are not considered part of the meat products industry.

        The meat products industry is primarily composed of four North American Industrial
Classification System (NAICS) codes: 311611 (animal - except poultry - slaughtering), 311612 (meat
processed from carcasses), 311613 (rendering), and 3U615 (poultry processing). Based on 7997
Economic Census data (U.S. Census,  1999a - 1999d), the industry employs 464,000 workers in 3,400
establishments, and produced $30.9 billion of products (value added basis) in 1997. The industry
component sectors, however, are quite distinct. For example, red meat slaughtering facilities (NAICS
311611) employ 142,000 workers in about 1,400 establishments, while red meat processors (NAICS
311612) employ 89,000 workers in 1,300 establishments.  However, total value added by meat processors
exceeds that of slaughterers (29 percent and 28 percent of total industry value added respectively). Poultry
plants (NAICS 311615) account for only 14 percent of industry establishments (470), yet employ almost
50 percent of the industry work force (225,000 workers) and produce 39 percent of industry value added
output.  Rendering facilities (NAICS 311613) employ 2 percent of industry labor and produce 4 percent of
output.
ES3  DATA SOURCES

       The economic analysis relies on a wide variety of sources. Both data availability and relevance
determined the relative reliance EPA placed on different sources for various components of the economic
profile, methodology, and analysis.
                                             ES-2

-------
       EPA surveyed the meat products industry under authority of the CWA Section 308 (U.S. EPA,
2002a). EPA administered 1,650 screener surveys and 350 detailed surveys.  EPA used data from the
screener survey to classify and subcategorize facilities by meat type, processes performed, and facility size
to determine the relevant industry population potentially affected by the proposed rule and to provide a
framework for the estimation of compliance costs and economic impacts. EPA will use facility and
company specific financial data from the detailed survey to develop models for estimating impacts of the
final rule; this data was not available in time for use in analyzing impacts for the proposed rule.

       EPA  relied heavily on the U.S. Census Bureau's 1997 Economic Census to profile the meat
products industry. Furthermore, data from the same source were used to develop economic model facilities
for estimating impacts of the proposed rule.  EPA also obtained special tabulations of Census data to
statistically model the distribution of facilities represented by each model facility. EPA used U.S.
Department of Agriculture (USDA) publications as data sources for the baseline economic models and the
analysis of changes and trends in the industry over time.  Publications by USDA's Economic Research
Service were a rich source of information and analysis on important issues such as the demand for meat
products, industry concentration, competitiveness, and technological change.

        Academic journals were an important source of information on the nature of competition in the
meat products industry, technological change, and industry trends. EPA also used academic research to
provide econometric estimates of key industry parameters — such as the price elasticities of demand and
 supply — for its economic impact models. EPA used industry sources such as trade journals and trade
 associations to develop its industry profile, to formulate a better understanding of industry changes, trends,
 and concerns, and to highlight significant firnis and their role in the industry.
 ES.4   ECONOMIC METHODOLOGY

        EPA developed capital and operating and maintenance (O&M) costs for incremental pollution
 control. The capital cost, a one-time cost, is the initial investment needed to purchase and install equipment
 involved in pollution control. The O&M cost is the annual cost of operating and maintaining that
 equipment; a site incurs its O&M cost each year. For this proposal, EPA estimated average compliance
                                               ES-3

-------
costs for a series of model facilities based on subcategory, size, and discharge type (for details, see the
Development Document, U.S. EPA, 2002b).

       EPA then annualized the estimated capital and O&M compliance costs. Annualized costs are
calculated as the equal annual payments of an annuity that has the same present value as the stream of cash
outflow over the project life and includes the opportunity cost of money or interest. An annualized cost is
analogous to a mortgage payment that spreads the one-time investment of a home over a series of constant
monthly payments. EPA annualizes capital and O&M costs because: (1) capital costs are incurred only
once in the equipment's lifetime and the initial investment should be expended over the life of the
equipment, and (2) money has a time-based value,  so expenditures incurred at the end of the equipment's
lifetime or O&M expenses in the future are not the same as expenses paid today.

       EPA used its estimated annualized compliance costs in four different levels of analysis:

        •       Facility-level impacts model (see Section 3.1.2 for details),
        •       Financial ratio analysis (see Section 3.1.3 for details),
        •       Market model (see Section 3.1.4 for details), and
        •       National impacts (see Section 3.1,5 for details).

Each is discussed briefly, below.

        EPA used 7997 Economic Census data at the employment class level from the four meat product
industry NAICS codes to develop model facilities representing red meat slaughter plants, red meat
processing plants, rendering plants, and poultry combined slaughter and processing plants. EPA used
Census revenue and cost data to estimate facility revenues, earnings before interest and taxes, net income,
and cash flow.  EPA also obtained from Census special tabulations of the variance of key revenue and cost
measures that it used to estimate the variance of each model facility's income.  Combining this with the
assumption that facility income is normally distributed, EPA estimated a cumulative probability
distribution function for each model facility. This  allows EPA not only to estimate impacts to each model
facility, but to the entire class of facilities the model represents as  well.  Thus, EPA presents two types of
model facility impacts. First, EPA provides impact measures such as the ratio of annualized compliance
                                              ES-4

-------
costs to revenues and net-income to the model facility itself.  Second, EPA uses its estimated probability
distributions to project impacts to the group of facilities represented by the model. These include impacts
such as the percentage and number of facilities that incur costs exceeding 100 percent of cash flow, or 1
percent of revenues.

       EPA used financial ratio analysis to examine whether a company can afford the aggregate costs of
upgrading all of its sites.  Many banks use financial ratio analysis to. assess the credit worthiness of a
potential borrower.  If regulatory costs cause a company's financial ratios to move into an unfavorable
range, the company will find it more difficult to borrow money. EPA considers a company in such a
condition to be in financial distress.  Financial ratio  analysis is performed at the company level rather than
the facility level.  This is because: (1) many firms maintain complete financial statements (balance sheet
and income statement) at the business entity or corporate level, but not the site level, (2) significant
financial decisions, such as expansion of a site's capacity, are typically made or approved at the corporate
level, and (3) the business entity (or corporate parent) is the legal entity responsible for repayment of a
loan, and therefore the lending institution evaluates the credit worthiness of the business entity, not the site.
EPA selected the Airman Z' score, a weighted-average of several financial ratios, to characterize the
baseline and post-regulation financial conditions of potentially affected firms. The Airman Z' score
simultaneously considers measures of liquidity, leverage, profitability, and asset management. It addresses
the problem of how to interpret the data when some financial ratios look  "good" while other ratios look
"bad." Also, it provides well defined thresholds for classifying firms as in good, indeterminate, and  poor
financial health. For proposal, EPA could only perform the Airman Z' score analysis for a select group of
facilities due to a lack of data availability; all firms will be examined for the final rule.

        EPA developed a market model to examine the impacts of the meat products industry effluent
guidelines on the price and output of various meat products. The distinguishing feature of EPA's market
model is that it explicitly incorporates cross-market impacts among meat types into the analysis. This is
for two reasons. First, the demand for meat products such as beef, pork, broilers, and turkey is closely
related; a change  in the price of pork will also tend to cause a change in the demand for beef because it is a
substitute for pork.  Second, EPA's proposed effluent guidelines will simultaneously affect the price of
beef, pork, chicken, and turkey, thus the market analysis for each product depends not only on the
compliance costs for that product but also on the impact of compliance costs on the prices of the other three
                                               ES-5

-------
meat products. The market model also examines international trade effects of the proposed rule; the export
of meat products is becoming an increasingly important source of growth for U.S. meat producers.

       Finally, EPA uses the U.S. Department of Commerce's Bureau of Economic Analysis (BEA)
"input-output" multipliers (RIMS II) to examine indirect and induced impacts of the proposed rule on the
national economy. Impacts on the meat product industry are known as direct effects, impacts on industries
that supply inputs to the meat products industry economy are known as indirect effects, and effects on
consumer demand are known as induced effects.
ES.5  RESULTS

       ES.5.1  Regulatory Options and Costs

       Table ES-1 presents EPA's proposed subcategories for the meat products industry along with the
facility process combinations (meat type and process classes) and EPA's count of potentially affected
facilities (based on survey data) contained in each subcategory.

       Table ES-2 summarizes the pollution control options considered for each subcategory. EPA is
proposing option 3 for BAT and NSPS in all subcategories except Subcategory J, for which option 2 is
proposed. EPA proposes to exclude small red meat processors (facilities producing less than 6,000 pounds
of finished product per day; Subcategory E) from revisions to the current guidelines.  EPA proposes to set
less stringent requirements (option 1) for small processors in subcategories K and L.  EPA does not
propose revisions to PSES in red meat-subcategories, nor does it propose to set PSES for subcategories K
and L.

       Table ES-3 provides estimated compliance costs by subcategory and option.  Note that EPA
estimated two sets of costs: "upper-bound" and "retrofit." Upper-bound costs represent the estimated cost
of purchasing new capital equipment for each option. However, in options 3 and 4, it is possible to retrofit
or upgrade already purchased wastewater treatment technologies to meet the more stringent standard rather
than purchase new equipment.  Thus EPA provides retrofit costs as a lower-bound compliance cost
                                             ES-6

-------
                                            Table ES-1
                               Proposal 40 CFR 432 Subcategories,
                                  Meat Type and Process Class,
                                     Discharge Type, and Size
Meat Type
Processes -

Size
Number of Facilities
Direct
Dischargers
Subcategory A through D
Red Meat
(1) First Processing;
(2) First Processing and Further Processing;
(3) First Processing and Rendering;
(4) First, Further Processing, and Rendering
Small
Non Small
59
66
Subcategory E through I
Red Meat
( 1 ) Further Processing;
(2) Further Processing and Rendering;
(3) Mixed Meat Further Processing;
(4) Mixed Meat Further Processing and
Rendering1
Small
Non Small
48
19
Subcategory J
Red Meat
or Poultry
(1) Rendering
Small
Non Small
6
21
Subcategory K
Poultry
(1) First Processing;
(2) First Processing and Further Processing;
(3) First Processing and Rendering;
(4) First, Further Processing, and Rendering
Small
Non Small
0
88
Subcategory L , '
Poultry
(1) Further Processing;
(2) Further Processing and Rendering;
(3) Mixed Meat Further Processing;
(4) Mixed Meat Further Processing and
Rendering1
Small
Non Small
4
15
Indirect
Dischargers

1,001
60

2,940
234

17
75

39
138

568
208
1 EPA allocated 61 percent of facilities from the mixed further processing and mixed further processing and
rendering classes to Subcategory E through I, and the remaining 39 percent to Subcategory L. For small facilities,
the allocation is 59 percent in Subcategory E through I and 41 percent in Subcategory L. EPA designated facilities
as "small" based on production (See Chapters 4 and 6 for details).
                                                ES-7

-------
                                        Table ES-2
                   Meat Products Industry Treatment Technology Options
Option
Treatment Unit
<* < * ป- •*•• .' I  * *• T^>* * 4 5.1 ซ. * •. ) * ซ"^ •*
"* : Indirect Dischargers
>ซ. ซ i ปป 1 1 .. , *ป ~ _-•
Preliminary Treatment, Dissolved Air Flotation, Equalization
Preliminary Treatment, Dissolved Air Flotation, Equalization, Nitrification -
Suspended Growth, Drying Beds
Preliminary Treatment, Dissolved Air Flotation, Equalization, Biological
Nitrogen Removal, Drying Beds
Preliminary Treatment, Dissolved Air Flotation, Equalization, Biological
Nutrient Removal - 3/5 Stage, Drying Beds
Changes between technology options indicated by italics.
                                           ES-8

-------


t— I

'I
a
,**•

" a
*
-
1




X

"j
1
> s
, <
4 2
cฃ
i-2
e

r
s
•v
I'S
33 N

a 'S
S g
^'4
"45

O
'3
U
c
" t
'11
C
<4j


fc
J*c
fl
11
c
.*" c
t: 1
c



o
^
>f
"e
•4
Q
C
i
~- ~ ฃ


P





r-
o
s


oo
oo
o
c
fe"




o
r-

^
CQ




































<

Z

K


<


s
n





o<
V
&


cr





ฃ
rr
fee

TT
oo

o

<

ฃ

%


55


cr
0
fee





c-
o
&

r-
•6^




t-~
c-

cs
&•
fT
OO



OO
in
CO
n
OS
IO
fee

fee


-H
'\f

0
a
fe<





0
OS
oo
f^V


c
c
fe1




oo
ot
\1
fe

CO
c/
oo



ฃ
3
00
OS
fee

ฃ
*


oo
fee

co
o
1





R


1
'6^:




r-
r--
\j
fee

o
n
FH













|
:[ii
?"

bl
~;-22
c/
















g

g

^


g


0





fee


o





o
fee


H
CQ

2

g

5^

^


g


c<
fee





•<
CO
fee


V
co
fee





' r~
fee

o
ซ



o
CO

&e

oo
CO
fee


2
fee

ir
:fee





S
•&e


ex
CO
•fee





p-
\o
cs
fee

CO
tt.



00
CO
Cs
1 — I
fee
oo
co
&e

cr-


co
S
3

co
oo
o
fee





cs
a
VD
•&e


c
fe<





in
VO
o
CM
CO
fee

'g
tฃ




































g

*

*


55


p-





fe<


00
oo
oo
G





cr
Vฃ
fee

5
oo

c

1

2

^


g


fe^





OS
CN
ir
G
fee


?





OS
CO
oo
CO
fee

O)
o-
W
c



oo
)0
O
ซ•
)O
ie

OS
CO
o>
/=


^r
u-
bซ

m~
tx
ir
fee





CM
oo
fee


OS
CO
os
CO





ง
VI
CO
fee

CO
cr
w
00



o
OS
m
fee
OS
X3
3

i
&e


ฃ
[•N
\f.

oc
feซ





S
o
fe<

c





>n
CM
OS
CM
>n
fee

c
oo











>

'I
. ^
' K
, C.
JC


















g

X

55


55


fee





o


o





o
fee


ฃ
05

CN

g

^

55


55


c
co
fee





ir
fee


fe<





o


S
cc



n
r-
m

CM
r-
oc
cr
fee





c<
Ol
V
fee


c
a
fef





co
oo
CO
OS

3
PC



CO
W

-------
1
 o
 8?
                                10
                                in
                   
-------
  1
   o
   2
0)
                                                                                                                                        oo

-------
estimate and expects that the true cost to industry will lie somewhere between the two figures.1 EPA
currently believes that the retrofit costs are the more realistic of the two sets of costs.
        ES.5.2 Impacts

        Table ES-4 presents facility level impacts under the proposed options.  Total posttax annualized
compliance costs are estimated to range from $50.4 million to $73.8 million. Posttax annualized costs per
facility range from $14,500 in Subcategory J to $550,000 in Subcategory A through D.  These costs
compose from 0.29 percent (Subcategory E through I) to 4.23 percent of net income (Subcategory L).
EPA estimates that annualized compliance costs per facility will average less than 0.5 percent of facility
revenues.

        Of the 20 major meat product companies for which EPA was able to perform the Altaian Z'
analysis, none are projected to incur financial distress under the proposed options.  Two firms, however,
are projected to experience some worsening of their financial condition, moving from ''financially healthy"
to "indeterminate" status.

        Table ES-5 provides projected market level and international trade impacts under the proposed
options.  The largest impacts are incurred in the market for chicken products. Estimated compliance costs
decrease supply of chicken products by 0.4 percent, causing a 0.12 percent increase in price, a 0.5 decrease
in domestic supply, and a 0.14 percent decrease in exports.  Impacts in other markets are smaller.
        ES.5.3 Small Business Impacts

        Based on Small Business Administration size standards, EPA estimates that 91 percent (5,174 out
of 5,670) of facilities in the meat products industry are small business owned (that is, they employ 500
workers or fewer). However, the vast majority of these facilities (4,991) are indirect dischargers, and thus
   1 As explained in Chapter 5, EPA was unable, for the purpose of this proposal, to allocate 65 "certainty"
facilities by Subcategory, hence costs for these facilities are estimated by multiplying total industry costs by 1.08.
                                              ES-12

-------
2 ' 11
ซ5 O Cซ
Jfe SJ
Qnt <&
- fO **
o -0 4ij '
ja ^*8 ง S
I1"" II
';|Iij|
. ; - v3_
" ',""* -fซ .ฃป ป
ป t Q j5?
- r. -*|f 8
" — ~ & <&

> Tปt •>* v , *
jx/ "^ ^
' "' o I" jfฃ i
" 5 S %ปil
- "' /'i3 .
S) S^*S* S ^s 6*
A ij1 S fi I*? ปs
SJ g,5! a ^ 'ง
5 fc S S" fa-
- "i- ^ar^
1-^ fJltll
P"H *1 ffl *3 ฃป^t p5
f *. i t * ^g^f S Vfr, r^S
*" ^^ "*fi3 -^ Istf
5 ,-^ur Pf *S-
* * ^ " "^S-r*
* "" V$ ST
* > ., *ซ•*.•*•* ^^ ^3
"* ^^|
' . ' " f "
3S S3^
gซtง
U






































Q
t
•s
Subcategory A
o




4
04

04
C5

04
O'


*
CO
O




O
ON
i— 1

0
^5
O
in
m

CO
VC)
co
•ฃซ•




VO



CO
PQ
Upper-Bound
CO
O




^.
, — i

ON
O
o

',_,
o


CO
04
0



0
o
CO
"— '

O
o
o
F-.
CO

ง

1








































Subcategory L
•st
o




in
04

oo

1—1
O


ฃ
p-
o



0
CO
01
"*

O
O
O
o
O4


OO





in
1— 1



CO
1
1
0
1
CO
o




in
-M

co
0

t— I
o


•n
m
0



0
T-H
o
co

o
in
in
oo
ffi

CO
S





in



CO
ง
|| Retrofit
CO

-------
         3 S's.s.-S-
         o -g S is S
^a
ซ1 *™ซ
             S.-2.J?
             M ^2 T^
             11
              ,i:
                  o\
Total Upper-Bound
                        trt
                       • 00

                        ฃ
                        \o
li
lli >O
o> ve
& tt>
S3*"

1-3
H5
Total Retrofit Includi

65 Certainty Fities
                                                             00
                                                             W
            a
                                 in
3.S
cd ''O
D-, C


11

ซl
8e
"S 8

ง1
* "
  n)
ss
o M
Ti a>
o -2
r^ S

ps
pe

-------
•a ซ •ซ
ง ™* S3
cj ปฃ3 M
Ma
tซ S "^
3 "2 "ft
*j g s
'<ฃ i1
o
Q"
111
-Idf"
-: 1
Gj *"Ci '"•j-
a 9 .2.
ป5
•S
' 0
-til
ง 2 ,_,
oปS *
'-*
**ti
'j|*j
izr-gf
- ง3S
'*-' ?
-Ji
-^
.-a
-2

o


a

(i










f-H


CO
OC
CN



OC
CN


VO
p-
\c
V
o
c-
c
S
5
0>



0








ON
o
9

R




o
o


o
c
9
VC
o
 — 1
o
oo
T-H
o
0
o


oo
p-


00
oo



oc



00
p-
9
C
C
OO
c

4


o>
p^
1
c

a)
C


-------
will not be affected by the proposed rule. EPA estimates that 183 small business owned facilities are direct
dischargers, 112 of which are likely to be excluded due to low levels of production, leaving 71 small
business owned facilities affected by the proposed rule.

        Table ES-6 presents EPA's projected small business impacts.  Four small processing facilities
(based on production) in Subcategory L incur posttax annualized costs of $700 per facility: about 2.4
percent of facility net income, and 0.2 percent of facility revenues. The 67 nonsmall (by production level)
affected facilities owned by small businesses incur, on average, posttax annualized costs of $119,000 per
facility. Note, however, that in subcategories A through D, E through I, and J, the average cost per facility
is $26,000 or less (less than 0.7 percent of net income and 0.2 percent of revenues).  Conversely, average
costs per facility for the 40 facilities in subcategories K and L range from $126,000 to $215,400, about 4.9
percent to 6.8 percent of net income.
        ES.5.4 Environmental Benefits

        The proposed meat products industry effluent limitations guideline will reduce emissions into the
waters of the United States. The reduction in emissions will reduce the levels of fecal eoliform and
biological oxygen demand and improve other indicators of water quality. As water quality improves,
waters may become suitable for increasingly demanding human uses. A primary benefit of the regulation is
the restoration of waters to conditions conducive to fishing and swimming.

        Each use category can be defined in terms of a set of water quality indicators. If the indicators
meet or exceed all of the criteria for a given use, then the water body can be used for that use. Vaughari
(1986) developed a water quality criteria ladder which describes the type of recreational use that a water
body can support (none, boating, fishing, or swimming). Once the use of the water body is defined by the
Vaughan ladder, the public willingness to pay for changes in use category can be estimated.

        One criticism of the water quality ladder approach is that a rule is only credited with a benefit
when it results in a change from one category to another.  Thus* even if a regulation causes significant
improvements hi water quality, but does not result in a change in use, no benefits are attributed to it. When
                                              ES-16

-------
a I
11
25  ป>
ฃ4 ^
^&
 =  ง
 rt
so
        f
        J
             p
             o
             o
             0
             01
             p
             0
             oo
            00

            CO
            CO
            V3-
             m
                 1
                 tt)
-S
1
                      01
                      o
                      O
                      O
                       v
                       (N
                       1
                           ^
                           ซ
IT)
0
                                 vo
                                 c5
                                v
                                ^-H
                                -ae-
                                ffl
                                 r—
                                 •a
                                      -
          05
                                          I
                                          00
ON
iri
               OO

               o
                                                cs
                                                oo
               00
               C5
               S
ffi
               I
                                                           CO
                                                           o
                         o
                         •ae-
          1
          00
                                                                cs
                                                                CN
                                                                ON
                                                                00
                                                                OO
               0
               ^5
               CM
                                                                01
                                                                CO

                                                               1
                                                                     0
                                                                   g
                                                                     o
                                                                     •Sfi-
I
I
                                                                          oo
                                                        oo
                                                        06
                                                                              o
                                                                              on

                                                                              cs
                                                                             |
                                                                             .S
                                                                              CO
                                                                              (U
                                                                              3
                                                                             X)
                                                                              S)
                                                                             CO
                                                                              I
                                                                             Hง'
                                             S 3
                                             O 'O
                                             r*>. M
                                             o a
"co C


It
03 SH

•<|
                                                                                         r~-

                                                                                         eo

-------
a marginal change in water quality measures results in a change in use category, large benefits are ascribed
to it. Therefore, EPA has also developed a continuous approach in order to value improvements in water
quality that do not result hi a change in use category (see Section 7.1.1 for details).

       EPA presents the results of the benefits evaluation for both the discrete and continuous methods of
determining the value of improvements in water quality. Under the proposed rule, EPA estimates that
about 21 miles of river reaches nationwide experience improvements.in water quality from nonswimmable
to swimmable levels.  EPA estimates that the public's willingness to pay for these improvements ranges
from $1.1 million (discrete method of valuation) to $15.6 million (continuous method of valuation).  These
benefits estimates reflect only the 36 plants actually analyzed for water quality improvements. The
corresponding annualized costs for these facilities are $33.7 million. If the ratio of costs to benefits for
these facilities is the same as the ratio of costs to benefits for all facilities, the .total (continuous) benefits of
the rule would be $37.0 million.
ES.6  REFERENCES
U.S. Census Bureau. 1999a. Animal (Except Poultry) Slaughtering. EC97M-3116A. 1997 Economic
       Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
       November.
U.S. Census Bureau. 1999b. Meat Processed From Carcasses. EC97M-3116B. 1997 Economic Census:
       Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.
U.S. Census Bureau. 1999c. Poultry Processing. EC97M-3116D. 1997 Economic Census: Manufacturing
       Industry Series. Washington, D.C.: U.S. Department of Commerce. November.
U.S. Census Bureau. 1999d. Rendering and Meat Byproduct Processing. EC97M-3116C. 1997 Economic
       Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
       December.
U.S. EPA. 2002a. 2001 Meat Products Industry Survey.  Washington, DC: OMB Control No. 2040-
       0225. Expiration Date February 29, 2004.
U.S. EPA. 2002b. Development Document for the Proposed Revisions to the Effluent Limitations
       Guidelines for the Meat Products Industry. EPA-821-B-01-007. Washington, D.C.: U.S.
       Environmental Protection Agency, Office of Water.
                                            ES-18

-------
Vaughan, William J. 1986. The RFF Water Quality Ladder, Appendix B in Robert Cameron Mitchell and
       Richard T. Carson, The Use of Contingent Valuation Data for Benefit/Cost Analysis in Water
       Pollution Control, Final Report.  Washington:Resources for the Future.
                                            ES-19

-------

-------
                                         CHAPTER 1

                                      INTRODUCTION
1.1     SCOPE AND PURPOSE

        The U.S. Environmental Protection Agency (EPA) proposes and promulgates water effluent

discharge limits (effluent limitations guidelines and standards) for industrial sectors. This Economic

Analysis (EA) summarizes the costs and economic impacts of technologies that form the bases for setting

limits and standards for the meat products industry.1


        The Federal Water Pollution Control Act (commonly known as the Clean Water Act [CWA, 33

U.S.C. ง1251 et seq.1) establishes a comprehensive program to "restore and maintain the chemical,

physical, and biological integrity of the Nation's waters" (section 101(a)).  EPA is authorized under

sections 301, 304, 306, and 307 of the CWA to establish effluent limitations guidelines and standards of

performance for industrial dischargers. The standards EPA establishes include:

        •       Best Practicable Control Technology Currently Available (BPT). Required under section
               304(b)(l), these rules apply to existing industrial direct dischargers. BPT limitations are
               generally based on the average of the best existing performances, by plants of various sizes,
               ages, and unit processes within a point source category or subcategory.

        •       Best Available Technology Economically Achievable (BAT). Required under section
               304(b)(2), these rules control the discharge of toxic and nonconventional pollutants and
               apply to existing industrial direct dischargers.

        •       Best Conventional Pollutant Control Technology (BCT). Required under section
               304(b)(4), these rules control the discharge of conventional pollutants from existing
               industrial direct dischargers.2 BCT limitations must be established in light of a two-part
             , cost-reasonableness test. BCT replaces BAT for control of conventional pollutants.

        •       Pretreatment Standards for Existing Sources (PSES). Required under section 307.
               Analogous to BAT controls, these rules apply to existing indirect dischargers (whose
               discharges flow to publicly owned treatment works [POTWs]).
         The industry, however, is free to use whatever technology it chooses in order to meet the limit.

       2 Conventional pollutant
coliform, pH, and oil and grease.
2 Conventional pollutants include biochemical oxygen demand (BOD), total suspended solids (TSS), fecal
                                               1-1

-------
       •       New Source Performance Standards (NSPS).  Required under section 306(b), these rules
               control the discharge of toxic and noncdnventional pollutants and apply to new source
               industrial dkect dischargers.
       •       Pretfeatment Standards for New Sources (PSNS). Required under section 307.
               Analogous to NSPS controls, these rules apply to hew source indirect dischargers (whose
               discharges flow to POTWs).

       The current meat products rule, 40 CFR Part 432, set effluent guidelines and limitations for the
beef and pork sectors of the meat products industry.  These standards were set and revised over a number
of years, most recently in 1995. Table 1-1 presents a listing of the standards set for each of the  10 current
subcategories in the meat products industry along with the relevant Federal Register citation.  This
proposed rule revises the existing subcategories in the industry, and proposes new standards for facilities
that perform poultry slaughter and processing operations. Prior to this proposed rule, EPA has set no
national effluent limitations guidelines or standards for poultry slaughterers or processors.
1.2     DATA SOURCES

        The economic analysis relies on a wide variety of sources. Both data availability and relevance
determined the relative reliance EPA placed on different sources for various components of the economic
profile, methodology, and analysis. Data sources used in the economic analysis include:

        •       EPA survey of the Meat Products industry.
        •       Census data.
        •       USDAdata.
        •       Academic literature.
        •       Industry journals.
        •       General economic and financial references (these are cited throughout the report).

The use of each of these major data sources is discussed in turn below.  Citations for these data sources as
utilized will be found in the relevant chapters of this EA.
                                                1-2

-------
                         Table 1-1
EPA Effluent Limitations Guidelines for Meat Products Industry
Subcategory v ; ;
Simple Slaughterhouses
(Subpart A)
Complex Slaughterhouses
(Subpart B)
Low-Processing Packinghouse
(Subpart C)
Standard
BPT
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES
NSPS
PSNS
BCT
Federal Register Notice
39 FR 789,7, February 28, 1974;
amended at 60 FR 33964, June 29, 1995
Reserved
40 FR 6446, February 11, 1975;
amended at 60 FR 33964, June 29, 1995
39 FR 7897, February 28, 19.74;
39 FR 26423, July 19, 1974
60 FR 33964, June 29, 1995
5 1FR 25001, July 9, 1986
39 FR 7897, February 29, 1974;
39 FR 26423, July 19, 1974;
amended at 45 FR 82254, December 15, 1980;
60 FR 33964, June 29, 1995
Reserved
40 FR 6446, February 11, 1975;
amended at 60 FR 33965, June 29, 1995
39 FR 7897, February 28, 1974;
39 FR 26423, July 19, 1974
60 FR 33965, June 29, 1995
5 1FR 25001, July 9, 1986
39 FR 7897, February 28, 1974;
amended at 60 FR 33965, June 29, 1995
Reserved
40 FR 6446, February 1 1, 1975;
amended at 60 FR 33965, June 29, 1995
39 FR 7897, February 28, 1974;
39 FR-26423, July 19, 1974
60 FR 33965, June 29, 1995
5 1FR 25001, July 9, 1986
                           1-3

-------
                     Table 1-1 (cont.)
EPA Effluent Limitations Guidelines for Meat Products Industry
Subcategory
High-Processing Packinghouse
(Subpart D)
Small-Processor (Subpart E)
Meat Cutter (Subpart F)
Sausage and Luncheon Meats
Processor (Subpart G)
Standard
BPT.
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES
NSPS
PSNS
BCT
BPT .
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
Federal Register Notice
39 FR 7897, February 28, 1974;
amended at 60 FR 33965, June 29, 1995
Reserved
40 FR 6446, February 11, 1975;
amended at 60 FR 33965, June 29, 1995
39 FR 7897, February 28, 1974;
39 FR 26423, July 19, 1974
60 FR 33965, June 29, 1995
5 1FR 25001, July 9, 1986
40 FR 905, January 3, 1975;
amended at 60 FR 33965, June 29, 1995
Reserved
Reserved
40 FR 905, January 3, 1975
40 FR 905, January 3, 1975;
amended at 60 FR 33965, June 29, 1995
5 1FR 25001, July 9, 1986
40 FR 906, January. 3, 1975;
amended at 60 FR 33965, June 29, 1995
44 FR 50748, August 29, 1979
Reserved
40 FR 906, January 3, 1975
40 FR 906, January 3, 1975;
amended at 60 FR 33965, June 29, 1995
5 1FR 25001, July 9, 1986
40 FR 907, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
40 FR 50748, August 29, 1979
                            1-4

-------
                     Table 1-1 (cont.)
EPA Effluent Limitations Guidelines for Meat Products Industry
Siibcategory •'• ;- :Y ' .:- ;:;;-^V^ V' ';'-•

Ham Processor (Subpart H)
Canned Meats Processor
(Subpart I)
Renderer (Subpart J)
Standard
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES
NSPS
PSNS
BCT
BPT
BAT
PSES .
NSPS
PSNS
BCT
^EederarRegister Notice • ;:
Reserved
40 FR 907, January 3, 1975
40 FR 907, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
5 1FR 25001, July 9, 1986
40 FR 908, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
44 FR 50748, August 29, 1979
Reserved
40 FR 908, January 3, 1975
40 FR 908, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
51 FR 25001, July 9, 1986
40 FR 909, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
44 FR 50748, August 29, 1979
Reserved
40 FR 909, January 3, 1975
40 FR 909, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
51 FR 25001, July 9, 1986
40 FR 910, January 3, 1975;
40 FR 1 1874, March 14, 1975;
amended at 60 FR 33966, June 29, 1995
44 FR 50748, August 29, 1979 .
Reserved
42 FR 54419, October 6, 1977
40 FR 910, January 3, 1975;
amended at 60 FR 33966, June 29, 1995
51 FR 25001, July 9, 1986
                            1-5

-------
       EPA collected site- and company-specific data under authority of the CWA Section 308 (U.S.
EPA, 2002). EPA administered 1,650 screener surveys and 350 detailed surveys. EPA used data from the
screener survey to classify and subcategorize facilities by meat type; processes performed, and facility size
to determine the relevant industry population potentially affected by the proposed rule and to provide a
framework for the estimation of compliance costs and economic impacts. EPA also used production data
from the screener survey to match engineering model facilities with economic model facilities.  EPA will
use facility and company specific financial data from the detailed survey to develop models for estimating
impacts of the final rule.

       EPA relied heavily on the U.S. Census Bureau's  1997 Economic Census to profile the meat
products industry. In addition, data from the same source were used to develop economic model facilities
for estimating impacts of the proposed rule. EPA also obtained special tabulations of Census data to
statistically model the distribution of facilities represented by each model facility.

       EPA used U.S. Department of Agriculture publications for two major purposes. First,
publications such as Livestock, Dairy and Poultry Situation and Outlook, and the Packers and Stockyards
Statistical Report provided data for the baseline economic models and the analysis  of changes and trends in
the industry over time.  Second, publications by USDA's Economic Research Service were a rich source of
information and analysis on important issues such as the demand for meat products, industry concentration,
competitiveness, and technological change. Finally, data to model international trade in meat products was
obtained from the databases of USDA's Foreign Agricultural Trade of the U.S. (FATUS) and the Food and
Agriculture Organization of the United Nations (UN FAO).

       Academic journals were an important source of information on the nature of competition in the
meat products industry, technological change, and industry trends. EPA also used academic research to
provide econometric estimates of key industry parameters - such as the price elasticities of demand and
supply - for its economic impact models.

       EPA used industry sources such as trade journals and trade associations  to develop its industry
profile, to formulate a better understanding of industry changes, trends, and concerns, and to highlight
                                               1-6

-------
significant firms and their role in the industry.  EPA also accessed company specific websites to develop its

profiles of major industry "players."


       As necessary, EPA cites various economic and financial references used in its analysis throughout

the EA. These references may be in the form of financial and economic texts, or other relevant sources of

information germane to the impact analysis.
1.3     REPORT ORGANIZATION
        This Economic Analysis (EA) is organized as follows:
               Chapter 2—Industry Profile

               Provides background information on the industry and companies affected by this
               regulation.
               Chapter 3—Economic Impact Analysis Methodology Overview

               Summarizes the economic methodology by which EPA examines incremental pollution
               control costs and their associated impacts on the industry. More detailed information on
               the economic methodology is located in Appendixes A through D.
               Chapter 4—Pollution Control Options

               Presents short descriptions of the regulatory options considered by EPA. More detail is
               given in the Development Document (U.S. EPA, 2002).
               Chapter 5—Economic Impacts

               Using the methodology presented in Chapter 3, EPA presents the annualized costs
               reflecting both the capital and annual operating and maintenance costs that are associated
               with more stringent pollution control. EPA then presents the economic impacts associated
               with the regulatory costs, including impacts on facilities, companies, industry output,
               prices, international trade, and employment. In other words, this chapter presents the
               findings on which EPA based its determination of economic achievability under the CWA.
                                               1-7

-------
              Chapter 6—Initial Regulatory Flexibility Analysis

              Pursuant to the Regulatory Flexibility Act as amended by the Small Business Regulatory
              Enforcement Fairness Act, EPA examines whether the regulatory options have a
              significant adverse impact on a substantial number of small entities.
              Chapter 7—Benefits Methodology

              Summarizes the methodology by which EPA identifies, qualifies, quantifies, and—where
              possible—monetizes the benefits associated with reduced pollution.
              Chapter 8—Cost and Benefits of the Proposed Rule

              Using the benefits described in Chapter 7, EPA presents an assessment of the nationwide
              costs and benefits of the regulation pursuant to Executive Order 12866 and the Unfunded
              Mandates Reform Act (UMRA).
1.4    REFERENCES

U.S. EPA. 2002. 2001 Meat Products Industry Survey.  Washington, DC: OMB Control No. 2040-0225.
       Expiration Date February 29,2004.
                                              1-8

-------
                                        CHAPTER 2
                                  INDUSTRY PROFILE
       Chapter 2 presents a profile of the meat products industry. Section 2.1 provides a snapshot of the
meat products industry based on 7997 Economic Census data; Section 2.2 is a snapshot of the industry
based on Section 308 survey data. This data formed the basis for EPA's subcategorization of the industry
and the framework for projecting economic impacts. Section 2.3 discusses trends in industry output and
prices. Section 2.4 describes the trends in beef, pork, and poultry production toward market concentration
and summarizes analyses of whether the trend toward concentration has generated significant market power
for the large firms apparently dominating the industry. Section 2.5 provides a brief guide to the important
players in the industry.

2.1    INDUSTRY OVERVIEW BASED ON CENSUS DATA

       The meat products industry includes establishments that primarily slaughter livestock and/or
process meat into products for further processing or for final sale to consumers. The industry can be
roughly divided into red meat facilities, primarily producing beef or pork products, and poultry facilities,
which primarily produce chicken (excluding eggs) and turkey products. (Red meat facilities may also
process lamb or veal. Poultry facilities may also process other birds, such as  ducks and geese, and also
small game, such as rabbits.) Facilities may perform slaughtering operations, processing operations from
carcasses slaughtered at other facilities, or both. In addition, rendering operations may be performed either
at stand alone facilities, or in combination with slaughter and/or further processing operations. Companies
that own meat product facilities may also own facilities that perform "upstream" or "downstream"
operations involved in getting meat products from the farm to the consumer (e.g., livestock raising,
wholesale distribution), but these facilities are not considered part of the meat products industry.
        The 7997 Economic Census (U.S. Census Bureau, 1999a through 1999d) provides a snapshot of
the meat products industry based on factors such as facility size, employment, value of shipments, and
geographical distribution. The red meat industry segment is profiled in two separate North American
Industry Classification System (NAICS) codes: Animal Slaughtering (NAICS 311611) and Meat
                                               2-1

-------
Processed From Carcasses (NAICS 311612).1 Thus, the NAICS codes divide the red meat industry into

meat packers (or slaughterers) and meat processors, but do not distinguish beef production from pork
production at the facility level. Therefore, neither Sections 2.1.1.1 nor 2.1.1.2 of this profile distinguish

beef production from pork production. Although these two industry segments are relatively well defined,

they are not mutually exclusive.2 The poultry processing industry (NAICS 311615) is well defined and

distinct from the red meat industries; it is profiled in Section 2.1.1.3. Note that at the corporate level, a

single company may own facilities in all three industry segments, while at the facility level, a single facility
may manufacture some products classified in  other segments. The rendering industry is briefly discussed in

Section 2.1.1.4. Section 2.1.2 relates the NAICS sectors to each other and to the meat products industry as
a whole, and also describes the industry's geographic distribution.


       2.1.1   Industry Sectors


       2.1.1.1 Animal (Except Poultry) Slaughtering: NAICS Code 311611


       NAICS 311611 consists of establishments  primarily engaged in the slaughter of cattle, hogs, sheep,
lambs, calves, and horses for human consumption.3 These establishments may also cook, can, cure, and

freeze the meat after slaughtering. Some industry establishments manufacture prepared feeds and feed
ingredients for animals (except dogs and cats). These establishments may perform slaughtering operations
to manufacture the animal feed as well.
        1 NAICS 311611 was previously covered under Standard Industrial Classification (SIC) code 2011 (Meat
Packing Plants) and part of SIC 2048 (Prepared Feeds, not elsewhere classified). NAICS 311612 was covered
under SIC 2103 (Sausages and Other Prepared Meats) and part of SIC 5147 (Wholesale Distribution of Meat and
Meat Products).

        2 The coverage ratio for animal slaughtering is 99 percent, i.e., 99 percent of animal slaughter product
shipments are accounted for by establishments classified in the industry. Furthermore, 96 percent of animal
slaughtering product shipments are the primary product of establishments classified in the industry. (This number
is called the specialization ratio). For the meat processing industry, the coverage ratio is 96 percent and the
specialization ratio is 92 percent.

        3 For this industry, the 7997 Economic Census did not fully implement the conversion from the SIC to the
NAICS system. Therefore the Census data for NAICS 311611 does not include SIC 0751, which consists of
establishments engaged in custom slaughtering. Nevertheless, the SIC and NAICS data for this industry are
comparable (within 3 percent).

                                                2-2

-------
       The animal slaughtering industry comprises 1,300 companies with approximately 1,400
establishments. The industry employs 142,000 people with payroll expenditures in excess of $3.2 billion.
The total value of shipments for the industry is $54 billion, of which $8.5 billion is value added by
manufacture.                                   .

       Twelve states have industry shipments exceeding $1 billion. Table 2-1 presents statistics for these.
As can be seen, Texas, California, Illinois, Iowa,,and Wisconsin contain the largest number of animal
slaughtering establishments, with at least 60 establishments each (the five states account for 28 percent of
                                                          v
all animal slaughtering establishments). Nebraska ranks seventh in the number of slaughtering
establishments, but with 18,500 workers, it employs the most workers in the slaughtering industry. Iowa,
Kansas, and Texas also employ more than 14,000  workers each in the industry. Combined, these four
states account for 44 percent of all employment in the animal slaughtering industry. Nebraska alone
accounts for almost 17 percent of all value added and 16 percent of total shipments  in the industry. Iowa,
Minnesota, Nebraska, and Texas contribute almost 45 percent of value added in the industry, while
Nebraska, Kansas, and Texas account for 40 percent of industry shipments. Thus industry activity is most
heavily concentrated in Nebraska, Kansas, Iowa, and Texas.                        .

        Table 2-2 portrays the relative importance to the industry of different establishment size categories.
More than a thousand establishments—72 percent of the total—have fewer than 20  employees each,
employ less than 5 percent of the industry workforce, and contribute an even smaller percentage of value
added and value of shipments to the industry. Conversely, while the 39 establishments employing between
 1,000 and 2,500 workers make up only 3 percent of the total number of establishments, they provide 43
percent of industry employment and 55 percent of value added by manufacture. Forty-six percent of the
value of shipments in this industry also comes from these facilities.

        With the exception of the largest establishments (those with employment exceeding 2,500 workers),
 as employee size class increases, the relative contribution of the class to industry output increases—even
 though the number of establishments in the class decreases. Note that while the nine establishments with
 more than 2,500 employees employ 19 percent of the industry workers, and 21 percent of industry
 shipments, the value added by these establishments is relatively low: only 4 percent of industry value added
 by manufacture is attributed to these facilities. Thus, the largest establishments apparently perform a very
                                                2-3

-------
                                          Table 2-1
                     1997 Animal Slaughter Industry: NAICS Code 311611
                                 Statistics for Selected States
State
United States
California
Colorado
Illinois
Iowa
Kansas
Michigan
Minnesota
Nebraska
Pennsylvania
Texas
Washington
Wisconsin
Number of Establishments
All
1,393
77
37
85
60
39
42
32
55
56
102
25
60
20 or More
Employees
386
27
13
35
25
10
13
12
25
23
30
9
19
Number of
Employees
142,374
4,300
5,999
8,663
16,163
14,116
2,725
5,462
18,461
4,923
14,055.
2,464
4,728
Value Added
by
Manufacture
($1,000,000)
8,525
306
416
492
811
658
3.69
783
1,414
282
794
163
411
Value of
•l Shipments
($1,000,000)
54,501
• 1,571
2,858
2,795
5,291
• 7,044
1,266
2,720
8,690
1,751
6,047
1,211
2,043
Source: U.S. Census Bureau, 1999a.
                                            2-4

-------
                                         Table 2-2
                     1997 Animal Slaughter Industry: NAICS Code 311611
                                Statistics by Employment Size
Employment Size
Class
Total3
1 to 19
20 to 99
100 to 249
250 to 999
1,000 to 2,499
2,500 or More
Number of
Establishments
1,393
1,007
220
64
54
39
9
Number of
Employees
142,374
5,990
10,324
9,833
26,926
61,833
27,468
Value Added by
Manufacture *
($1,000,000)
8,525
220
602
729
1,936
4,706
331.
Value of
Shipments
($1,000,000)
54,501
1,081
2,758
4,133
10,047
24,892
1 159
source: U.i>. Census Bureau, 1999a.
a Numbers may not sum due to rounding.
                                           2-5

-------
high volume of low value-added operations—presumably just the initial slaughter and cutting operations
(e.g., whole and half carcasses)—with higher-value operations occurring at other facilities.

        Table 2-3 presents the value of shipments for selected animal slaughter industry primary products.
Beef products make up approximately 55 percent of total shipments; over half of beef production is
accounted for by boxed beef (30 percent of total shipments). Pork products make up 34 percent of
shipments; of $17 billion in total pork product shipments, approximately 30 percent are accounted for by
products requiring further processing such as curing and sausage making. The remainder of shipments
consists primarily of veal and lamb products, with a small fraction accounted for by hides, skins, and pelts.
        2.1.1.2 Meat Processed From Carcasses: NAICS Code 311612

        Establishments in NAICS 311612 are engaged in processing or preserving meat and meat
 byproducts (but not poultry or small game) from purchased meats. Many of the processing and canning
 operations are essentially identical to those undertaken in the animal slaughter industry .(NAICS 311611). It
 is not the final processed, canned, cooked, or cured meat product that differs between the two NAICS
 codes, but the fact that one industry produces that meat product from animals that it slaughters in its
 facility, while the second industry performs no slaughtering operations, purchasing its meat inputs from
 other facilities.

        The meat processing industry comprises 1,164 companies. These companies own and operate
 approximately 1,300 meat processing establishments/The industry employs 88,000 people, with a payroll
 exceeding $2.3 billion. The value of all shipments from this industry is more than $25.0 billion, of which
 $9.1 billion is value added by manufacture. Thus, although there are almost as many establishments in the
 meat processing industry as the animal slaughter industry, employment in meat processing is approximately
 60 percent of the employment in animal slaughter, and the value of shipments is 45 percent. However,
 value added in meat processing exceeds that of slaughtering by $600 million (i.e., it is 7 percent greater).

        Table 2-4 shows the geographic distribution of major meat processing establishments and the
 relative geographic concentration of the industry. Four states, California, Illinois, New York, and Texas,
                                                2-6

-------
                                             Table 2-3
                       1997 Animal Slaughter Industry: NAICS Code 311611
                                 Output by Selected Product Codes
'• ^NAICS,
Product
Code
311611
3116111
31161111
31161113
31161115
311611A
311611A121
311611G
311611J
311611P
Product Description
Animal slaughtering products, except poultry *
Fresh and frozen beef, not canned or made into sausage, made
from animals slaughtered in this plant
Fresh and frozen whole carcass and half carcass beef, not canned or
made into sausage, made from animals slaughtered in this plant
Fresh and frozen subprimal and fabricated cuts packaged in plastics
(boxed beef), not canned or made into sausage, made from animals
slaughtered hi this plant
Fresh and frozen boneless beef, including hamburger, not canned or
made into sausage, made from animals slaughtered in this plant
Fresh and frozen pork, not canned or made into sausage, made
from animals slaughtered in this plant
Fresh and frozen primal and fabricated cuts (including trimmings),
not canned or made into sausage, made from animals slaughtered in
this plant
Pork, processed or cured (not canned or made into sausage),
made from annuals slaughtered in this plant
Sausage and similar products (not canned), made from animals
slaughtered in this plant
Hides, skins, and pelts
Value of
Product
Shipments
($l,000,000)a
50,781
28,209
6,734
15,465
3,272
11,812
10,249
3,305
1,998
2,068
Source: U.S. Census Bureau, 1999a.
a Value of shipments by product class is not the same as value of shipments by industry. Value of shipments by
industry includes all products from establishments classified as animal slaughtering plants, whether those products
are primary to the industry or not; value of shipments by product class includes all shipments of that product
regardless of the industry classification of the establishment.
                                                2-7

-------
                                           Table 2-4
                      1997 Meat Processing Industry: NAICS Code 311612
                                  Statistics for Selected States
State
United States
California
niinois
Iowa
Kansas
Nebraska
New York
North Carolina
Ohio
Pennsylvania
Texas
Wisconsin
Number of '.-
Establishments .
All
1,297
123
94
40
20
21
96
40
46
74
99
53
20 or More
Employees
622
60
51
24
13
15
34
22
23
46
49
27
- Number of
Employees
87,966
4,779
6,515 •
4,764
2,574
3,369
2,419
3,290
4,638
5,169
7,296
10,000
Value Added
'•• ^:*ฃij;. •:••";
Manufacture
($l,000j000)
9,136
467
720
875
234
212
938
125
454
428
1,094
1,220
'-;^^yaiue;ofV'
Shipments
($l;0d0;000)
25,005
1,147
. 1,911
2,438
692
111
1,210
481
• 1,375
1,491
2,570
2,951
Source: U.S. Census Bureau, 1999b.
                                             2-8

-------
contain more than 90 meat processing establishments each, and account for almost 32 percent of industry
establishments. As with the animal slaughter industry, however^ employment in this industry is
concentrated in another set of states: Illinois, Pennsylvania, Texas, and Wisconsin. Together, these four
states employ one-third of the United States's meat processing employees. Thus, these states tend to have
larger establishments. In Wisconsin, for example, more than half the establishments employ more than 20
workers; Wisconsin also accounts for the largest share of both total shipments and value added in the
industry. Four states, Illinois, Iowa, Texas, and Wisconsin, account for almost 40 percent of industry
shipments. Of these four, two states, Iowa and Texas, are also among the largest four animal slaughtering
states, while the other two largest slaughtering states, Nebraska and Kansas, rank ninth and tenth
respectively in meat processing shipments.4 Thus, the meat processing industry partially, but not entirely,
overlaps the slaughter industry in terms of geographical distribution. It is not as regionally concentrated as
the slaughter industry.

        Table 2-5 presents meat processing establishments according to employment class. From the table
it can be seen that more than half of the establishments have fewer than 20 employees, but this group
contributes only a small fraction of value added and value of shipments of this industry. The bulk of
employment (54 percent), value added (55 percent) and total shipments (57 percent) is accounted for by
facilities employing between 100 and 500 workers.

        A comparison between meat processing and animal slaughtering facilities by employment class is
illuminating. The distribution of employment, value added, and value of shipments in the meat processing
industry is relatively equally divided among facilities in the 100 to 249 and 250 to 999 employment classes,
with smaller, but still considerable percentages accounted for by establishments in the 20 to 99 and the 500
to 999 employment classes. The largest (more than 1,000) and smallest (fewer than 20) employment classes
account for relatively small percentages of employment and production. In the animal slaughtering
        4 Note that while New York is ranked seventh in value of shipments, it is third in value added by
manufacture. This relatively high share of value added could be something of an anomaly: if New York processors
purchase meat inputs (which account for the largest share, by far, of production costs) at approximately the same
price as meat processors in other states, but need to pay New York wages to meet the high cost of living in the
greater metropolitan New York area (and selling their product at a higher price as well), these establishments
would show a greater percentage of value added per dollar of shipment relative to areas where the cost of living is
lower. Some of the difference may also be attributable to differences in product mix (e.g., veal compared to
bologna).
                                               2-9

-------
                                           Table 2-5
                      1997 Meat Processing Industry: NAICS Code 311612
                                 Statistics by Employment Size
Employment Size
Class
Total"
Itol9
20 to 99
100 to 249
250 to 499
500 to 999
1 000 or Moreb
Number of
Establishments
1,297
675
386
143
68
22
3
Number of
Employees
87,966
. 4,661
17,566
23,298
23,983
18,458
4,946
Value Added by
Manufacture
($1,000,000)
9,136
366
1,506
2,755
2,264
2,245
714
Valueof
Shipments
($1,000,000)
25,005
930
4,332
7,697
6,618
5,427
1,538
Source: U.S. Census Bureau, 1999b.
" Numbers may not sum due to rounding.
b Two establishments employ between 1,000 and 2,449 workers, one establishment employs between 2,500 and
4,999 employees; the Census Bureau did not provide detail due to confidentiality issues.
                                              2-10

-------
industry, meanwhile, the distribution of employment and output is heavily skewed toward the largest
establishments. Establishments employing more than 1,000 workers account for 63 percent of employment
and 66 percent of shipments. There are 48 establishments employing more than 1,000 workers in animal
slaughtering, but only 3 in meat processing. Thus the animal slaughter industry is dominated by a handful
of very large facilities, while output from the meat processing industry is spread relatively evenly over a
large number of moderately sized facilities.

        Table 2-6 lists the value of shipments for selected product codes in the animal processing industry.
The share of industry shipments by type of meat is reversed in the meat processing industry compared to
the animal slaughtering industry. Pork makes up the biggest share of total shipments at 52 percent, while
beef products account for roughly a third of shipments.5

        2.1.1.3 Poultry Processing: NAICS Code 311615

        Establishments in the poultry processing industry primarily slaughter poultry or small game, and
may also process the meat and prepare meat byproducts.6 Under the SIC system, the code for poultry
processing (SIC 2015) includes facilities that dry, freeze, and break eggs. Therefore, data for the SIC and.
NAICS codes for this industry are not comparable. SIC sales or receipts cannot be estimated within 3
percent from NAICS data, and only 95 percent of SIC 2015  sales and receipts are classified under NAICS
311615.                .

        Poultry processing operations are performed by 260 companies, which own 470 poultry processing
establishments. Together, these companies employ a total of 224,000 employees, with a payroll exceeding
$4.0 billion. The poultry processing industry's total value of shipments is $31.7 billion, of which $12.1
billion is value added by manufacture.
        5 Note that although the pork product percentage of shipments exceeds 50 percent, the absolute value of
pork shipments is lower in the meat processing industry than in the animal slaughter industry.
        6 The coverage ratio for poultry processing is 96 percent. The specialization ratio is 97 percent. Thus, the
poultry processing industry is well defined by this NAICS code.
                                               2-11

-------
                                             Table 2-6
                       1997 Meat Processing Industry: NAICS Code 311612
                                 Output by Selected Product Codes
NAICS
Product
Code
311612
3116121
31161212
31161214
3116124
31161241
31161242
31161243
31161244
311612A
311612A1
311612A2
311612A3
311612A4
Product Description ;
Meat processed from carcasses
Pork, processed or cured, including frozen, (not canned or made into
sausage), not made in meat packing plants
Smoked pork hams and picnics (not otherwise cooked), except canned, not
made into sausage
Smoked pork sliced bacon (not otherwise cooked), except canned, not made in
meat packing plants
Sausage and similar products, (not canned), not made in meat packing
plants
Fresh sausage (pork sausage, breakfast links, etc.), except canned, not made in
meat packing plants
Dry or semidry sausage and similar products (salami, cervelat, beef jerky,
pepperoni, summer sausage, pork roll, etc.), except canned, not made in meat
packing plants
Frankfurters, including wieners, except canned, not made in meat packing
plants
Other sausage, smoked or cooked, and jellied goods and similar preparations,
not canned, not made in meat packing plants
Other processed, frozen, or cooked meats, not made in meat packing
plants
Boxed meat (beef, pork, lamb, etc.) not made in slaughtering plants
Frozen ground meat patties (processed, frozen, or cooked), not made in meat
processing plants
Frozen portion control meats (processed, frozen, or cooked), not made in meat
packing plants
Other processed, frozen, or cooked meats, not made in meat packing plants
Value of
Product
Shipments
($1,000,000)"
22,245
5,068
2,208
1,628
6,527
1,088
1,189
1,546
2,701
7,737
1,463
1,759
1,061
3,241
Source: U.S. Census Bureau, 1999b.
8 Value of shipments by product class is not the same as value of shipments by industry. Value of shipments by
industry includes all products from establishments classified as meat processing plants, whether those products are
primary to the industry or not; value of shipments by product class includes all shipments of that product regardless
of the industry classification of the establishment.
                                                2-12

-------
       Table 2-7 presents data on poultry processing establishments for nine states in which the value of
poultry product shipments exceeded $1 billion per state. Unlike the red meat industries described above, the
poultry processing industry has a large percentage of establishments—82 percent—that employ more than
20 workers. Among these are almost all the establishments in Arkansas and Georgia.7 Five states,
Alabama, Arkansas, California, Georgia, and North Carolina, account for 36 percent of the nation's
poultry processing establishments. Output and employment are dominated by four of these states: Alabama,
Arkansas, Georgia,  and North Carolina account for approximately 44 to 45 percent of industry workforce,
value added,  and total shipments of processed poultry in the United States.

        The poultry processing industry has relatively few very small facilities. Like the red meat animal
slaughtering  industry, it is dominated by a few very large facilities. This is shown in Table 2-8. Almost 50
percent of industry employment and over 40 percent of industry shipments are accounted for by the 75
facilities that employ more than 1,000 workers each. Facilities with more than 500 workers account for 80
percent of employment and 74 percent of total shipments. Yet facilities employing more than 500 workers
each make up only  36 percent of poultry processing establishments.

        Output of the poultry processing industry can be divided into three key components: broilers and
fryers, turkeys, and processed poultry. Shipments by the industry for selected product codes are presented
in Table 2-9. Broilers and fryers are by far the most important product, making up over half of the
industry's shipments. Processed poultry accounts for approximately 30 percent of shipments, and turkey
products account for about 12 percent of shipments.

        2,1.1.4 Rendering and Meat Byproduct Processing: NAICS Code 311613

         NAICS 311613 consists of establishments engaged in rendering inedible stearin, grease, and tallow
 from animal fat, bones, and meat scraps. It also includes establishments manufacturing animal oils,
         7 Red meat processing establishments most likely include many relatively small butcher shops and
 specialty meat processors. Poultry production as a specialized industry, on the other hand, is a relatively recent
 development that started directly with industrialized production, resulting in relatively large facilities.
                                                2-13

-------
                                          Table 2-7
                    1997 Poultry Processing Industry: NAICS Code 311615
                                 Statistics for Selected States
State
United States
Alabama
Arkansas
California
Georgia
Mississippi
Missouri
North Carolina
Texas
Virginia
Number of
Establishments
All
474
30
43
29
42
25
24
29
18
15
20 or More
Employees
387
28
42
19
40
22
19
26
15
13
Number of
Employees
224,511
19,944
33,409
7,671
30,435
15,952
12,215
18,166
. 10,792
10,162
Value Added
• :;:by:: " •;.
Manufacture
($1,000,000)
12,062
1,088
1,869
577
1,201
665
994
1,111
586
386
Value of
Shipments
($1,000,000)
31,656
' 2,340
4,908
1,327
3,833
1,672
1,988
2,852
1,620
1,518
Source: U.S. Census Bureau, 1999c.
                                             2-14

-------
                                         Table 2-8
                    1997 Poultry Processing Industry: NAICS Code 311615
                                Statistics by Employment Size
Employment Size
Class
Total3
ItolP
20 to 99
100 to 499
500 to 999 .
1,000 to 2,499
2,500 or More
Number of
Establishments
474
87
69
146
97
70
5
: Number of
Employees
224,511
407
3,421
40,418
70,625
95,187
14,453
Value Added by
Manufacture V;
($1,000,000)
12,062
34
345
, 2,558
4,111
4,634
379.
Value of
Shipments
($i#OOiOOO)
31,656
79
851
7,186
10,536
11,621
1,383
Source: U.S. Census Bureau, 1999c.
a Numbers may not sum due to rounding.
                                           2-15

-------
                                            Table 2-9
                     1997 Poultry Processing Industry: NAICS Code 311615
                                Output by Selected Product Codes
NAICS
Product
Code
311615
3116151
31161511
31161513
31161514
3116157
31161572
31161573
311615D
311615D
121
311615D
131
311615D
151
- --• . ,v. ;. -. ; .:, .;.,;-. ••*,.•:•: .. • •• .. •.. • •.. • - •..._. ; .. : ,
-•'•• .-'- V- -^'"'- •'.'."> .-•-' "• ": '.'-' •".:'.- : ' '
Product Pescriptioh
Poultry processing
Young chickens (usually under 20 weeks of age), whole or parts
Wet ice pack broilers and fryers (usually under 20 weeks of age), bulk
Tray pack (consumer packaged) broilers and fryers (usually under 20
weeks of age), chilled
Other broilers and fryers (usually under 20 weeks of age), including
frozen
Turkeys (including frozen), whole or parts
Young turkeys (mature) (usually 4 to 7 months of age), whole, including
frozen
Old turkeys, whole, and turkey parts
Processed poultry and small game (except soups) containing 20
percent or more poultry or meat
Cooked or smoked turkey, including frozen (except frankfurters, hams,
and luncheon meats), containing 20 percent or more poultry
Cooked or smoked chicken, including frozen (except frankfurters, hams,
and luncheon meats), containing 20 percent or more poultry
Cooked or smoked poultry hams and luncheon meats, including frozen,
containing 20 percent or more poultry
Value of
Product
Shipments
($1,000,000)"
30,998
16,527
6,702
4,030
3,449
3,802
1,705
1,915
9,200
1,403
4,125
1,838
Source: U.S. Census Bureau, 1999c.
a Value of shipments by product class is not the same as value of shipments by industry. Value of shipments by
industry includes all products from establishments classified as meat processing plants, whether those products are
primary to the industry or not; value of shipments by product class includes all shipments of that product regardless
of the industry classification of the establishment.
                                                2-16

-------
 including fish oil, and fish and animal meal.8 Many establishments not classified as rendering plants
perform rendering operations; only 62 percent of primary product shipments are accounted for by
establishments classified in this industry (the coverage ratio).9

        The rendering industry consists of 137 companies, which own and operate 240 establishments. The
industry employs 8,800 employees, with $269 million in payroll expenditures. The total value of shipments
in 1997 was $2.6 billion, with value added by manufacture of $1.3 billion.

        Table 2-10 displays employment and output data for the six states with more than $100 million in
rendering product shipments. Texas and California are the two states accounting for the largest share of the
rendering industry. The six states listed in Table 2-10 contain establishments with 34 percent of total
industry shipments.                                       .

        Table 2-11 summarizes rendering industry establishments according to employment size class. In
general, rendering plants are relatively small; only 11 plants employ more than 100 workers each, and only
one employs more than 250 workers. The  132 establishments that employ between 20 and 99 workers
account for the largest share of industry shipments (66 percent) and employment (72 percent).

        Table 2-12 lists  the value of shipments for selected rendering industry product codes. The industry
has two primary product classes: (1) rendering and meat byproducts (primarily lard), accounting for 31
percent of shipments, and (2) animal and marine feed and fertilizer products, accounting for 63 percent of
shipments. Miscellaneous rendering products, none of which are significant, account for the remainder.
        8 Prior to 1997, this industry was classified as SIC 2077: Animal and Marine Fats and Oils. The 7997
Economic Census did not fully implement the conversion from the SIC system to NAICS for this industry. NAICS
311613 does not include establishments engaged in manufacturing lard from purchased materials. Hence, the SIC
and NAICS codes for this industry are not comparable.
        9 However, 97 percent of product shipments by establishments classified as rendering facilities are
rendering products (the specialization ratio).
                                               2-17

-------
                                        Table 2-10
                        1997 Rendering Industry: NAICS Code 311613
                                 Statistics for Selected States
State
United States
California
Georgia
Minnesota
Nebraska
Pennsylvania
Texas
Number of Establishments
All
240
21
10
12
15
9
20
20 or More
Employees
143
14
8
5
9
4
12
Number of
Employees
8,804
770
432
358
474
301
789
Value Added
.:••:•;*ฃ••; -:
Manufacture
($1,000,000)
1,257
77
44
45
75
54
103
Value of
Shipments
($1,000,000)
2,572
178
109
101
159
128
208
Source: U.S. Census Bureau, 1999d.
                                            2-18

-------
                                          Table 2-11
                         1997 Rendering Industry: NAICS Code 311613
                                 Statistics by Employment Size
Employment Size
Glass
Total2
1 to 19
20 to 49
50 to 99
100 or Moreb
Number of
.Establishments
240
97
81
51
11
•- ' " "-"-_".
Number of
Employees
8,804
839
2,803
3,550
1,612
Value Added by
Manufacture
($1^000^000)
1,257
136
435
417
269
Value of
Shipments
($1,000,000)
2,572
380
879
811
502
Source: U.S. Census Bureau, 1999d.
a Numbers may not sum due to rounding.
*? Ten establishments employ between 100 and 249 workers, one establishment employs between 250 and 499
employees; the Census Bureau did not provide detail due to confidentiality issues.             •'•  •
                                              2-19

-------
                                           Table 2-12
                         1997 Rendering Industry: NAICS Code 311613
                                Output by Selected Product Codes

NAICS
Product
Code
311613
3116131
31161311
3116134
31161341


• •--'. -;"-•'' • --- • ••' ' * •• - .", •• • '' - ' • '-' ••

Product Description
Rendering or meat byproducts
Rendering and meat byproduct processing
Lard, except canned, not made in meat packing plants
Animal and marine feed and fertilizer byproduct
Animal and marine feed and fertilizer byproducts
Other feed and fertilizer byproducts..
Value of
Product
Shipments
($1,000,000)?
3,839
1,209
1,142
2,406
1,096
1,060
Source: U.S. Census Bureau, 1999d.
" V<-lae of shipments by product class is not the same as value of shipments by industry. Value of shipments by
industry includes all products from establishments classified as meat processing plants, whether those products are
primary to the industry or not; value of shipments byproduct class includes all shipments of that product; regardless
of the industry classification of the establishment.
                                                2-20

-------
       2.1.2   Sector Overview

       Sections 2.1.1.1 through 2.1.1.4 present a detailed overview of the principal sectors of the meat
product industry. This section places those component sectors in context to each other: the importance of
each of the four component NAICS code sectors relative to the overall size of the industry. Then it places
the industry in geographical context: which states are the most important producers in the industry and,
therefore, which states may be most affected by the proposed effluent guideline.

       2.1.2.1 Relative Industry Shares

       Figure 2-1 shows what percent of the industry each NAICS sector occupies. Industry output as
measured by value of shipments for the meat products industry in 1997 totaled $113.9 billion (the sum of
the value of shipments for NAICS 311611, 311612, 311613, and 311615). Almost 50 percent of that
output (47.9 percent) v/as produced in plants that perform (nonpoultry) animal slaughter operations
(NAICS 311611). The poultry sector (NAICS 311615)—slaughterers, processors, and entities that
slaughter and process—produced 27.8 percent of shipments. Plants that process but do not slaughter
animals (NAICS 311612) produced 22 percent of shipments, and plants that primarily perform rendering •
operations (NAICS 311613) account for 2.3 percent of shipments.

       In fact, value of shipments does not express the relative significance of industry segments as well
as value added by sector. Value added subtracts the cost of material inputs from the value of shipments, so
it includes an estimate of the additional value to materials already produced that can be attributed to an
industry or sector. A prime example of the significance of measuring value added can be observed in the
relationship between slaughter plants and further processing plants (NAICS 311611 and 311612). Further
processing plants use the output of slaughter plants as raw materials in their production process: Including
the value of meat purchased from slaughter plants in the value of processing plant output means double-
counting goods produced by the slaughter sector. Comparing the relative shares of industry value added,
the poultry sector is the largest, accounting for 38.9 percent of the meat product industry's $30.9 billion
value added. Further processing accounts for a larger share of industry value added than slaughter plants:
29.5 percent to 27.5 percent. Rendering accounts for 4.1 percent of industry value added.
                                               2-21

-------
                                        Figure 2-1
     Meat Products Industry: Percentage of Employment, Total Shipments, and Value Added
                                     by NAICS Sector
         27.8%
       2,3%
             22.0%
                                      47.9%
 Slaughters
 Processing
I Rendering
 Poultry
                                                               Percentage of Total Shipments
      38.9%
             4.1%
                                   27.5%
                                                    Slaughter:
                                                    Processing
                                                    Rendering
                                                    Poultry
                               29.5%
                                                               Percentage of Value Added
      48.4% \k
                    1.9%
                                    30.7%
 Slaughters
 Processing
 Rendering
 Poultry
                                19.0%
                                                               Percentage of Employment
Source: U.S. Census Bureau, 1999a through 1999d
                                           2-22

-------
       The poultry sector accounts for almost 50 percent of the 464,000 total jobs provided by the meat
products industry. Plants that perform slaughter operations account for 30.7 percent of employments, while
19 percent and 1.9 percent of jobs can be attributed to further processing and rendering, respectively. Note
that this suggests that the value added by an employee in the processing sector is much higher than the
value added by an employee in either the slaughter or poultry sectors.
       2.1.2.2 Geographic Distribution of Industry

       EPA presents two comparisons to demonstrate the relative importance of geographical regions to
the meat product industry, as well as the importance of the industry in state economies.

       The top panel of Figure 2-2 presents the value of meat product shipments by state. Texas is the
leader, producing more than $10 billion worth of meat product shipments in 1997. Nebraska, Iowa, and
Kansas follow Texas, with shipments valued between $7.5 billion and $10 billion. In the third tier are
states such as Arkansas, Wisconsin, and Illinois, with 1997 shipments between $5 billion and $7.5 billion
in 1997. These seven states account for 46 percent of meat product shipments.

       The lower panel of Figure 2-2, however, indicates that the significance of the meat products
industry within these states varies widely. For Nebraska, Kansas, Iowa, and Arkansas, the meat products
industry accounts for a minimum of 12 percent of manufacturing production within the state; in Kansas,
for example, it accounts for almost 35 percent of state manufacturing output. While Texas is the largest
producer of meat products by value, meat products only make up 3.5 percent of state manufacturing
production (Texas is seventeenth in percentage of production devoted to meat products). Conversely, while
Delaware is only ranked twenty-ninth in the value of meat production—its total value of production is 8
percent of Texas' production—meat products are relatively more important to its economy than to the
Texas economy, accounting for more than 6 percent of manufacturing output.

       A similar pattern can be observed in industry employment (Figure 2-3). Arkansas, Georgia, and
Texas each employ more than 30,000 meat product industry workers. North Carolina, Iowa, Nebraska, and
Alabama each employ more than 20,000 workers. Nebraska is first, though, in percent of employment: the
                                              2-23

-------
                     Figure 2-2
   Value of Meat Products Shipments by State and
Meat Products as a Percentage of Shipments by State
                                                           less than $1 b.
                                                           $lb. to $2.5 b.
                                                           $2.5b. to$5b.
                                                           $5b. to $7.5 b.
                                                           $7.5b. to$10b.
                                                           more than $10 b.
                                                             4% to 8%
                                                             8% to 12%
                                                             more than \2"f
                           2-24

-------
                   Figure 2-3
     Meat Products Employment by State and
Employment as a Percentage of State Employment
                                                           [216,000
                                                           1! 12,000
                                                           Hi 8,000
                                                           124,000
                                                           130,000
                                                           • 36,000
                                                            3.60%
                                                            7.20%
                                                            10.80%
                                                            14.40%
                                                            18.00%
                                                            21.60%
                     2-25

-------
meat products industry accounts for 21 percent of Nebraska's manufacturing employment. Although
Delaware only provides 6,400 meat product industry jobs, making it twenty-fourth among all states, the
meat product industry accounts for 16 percent of manufacturing employment in Delaware. Meat product
industry employment accounts for 3.4 percent of manufacturing employment in Texas, which is the third
largest industry employer.
2.2    SCREENER SURVEY AND SUBCATEGORIZATION

       For the proposal analysis, EPA used the 2.001 Meat Products Industry Screener Survey
(hereinafter referred to as the "Screener Survey") to obtain information on a sample of meat product
facilities potentially affected by the rule. EPA used its authority under Section 308 of the Clean Water Act
to collect information not available otherwise, such as site-specific employment, production, and
wastewater data.

       EPA used this detailed data to construct a framework for the proposed effluent guideline.  Site
level data on production and wastewater flow was used to determine classifications based on meat type,
type of process, size, and discharge type. These meat type and process classes were then grouped into the
40 CFR 432 subcategories (hereafter, subcategories).  Effluent limitations and guidelines are set on the
basis of these subcategories.
        2.2.1   Meat Type and Process Classes

        2.2.1.1 Method of Classification

        To generate the meat type and process classes, EPA first evaluated the screener survey population
based on the type of meat produced at the facility:
               red meat (primarily beef and pork),
               poultry (primarily chicken and turkey),

                                            - 2-26

-------
        •       mixed (both red meat and poultry),
        •       rendering, or meat byproducts (either red meat or poultry),

'and second, the type of processes performed at the facility:

        •       first processing (slaughter),
        •       further processing,                                                             •
        •       rendering (the process resulting in meat byproducts).

 This results in a classification of facilities consisting of combinations of the processes for each meat type.
 For example, a poultry facility may perform any of the following six combinations of processes, each one
 of which will place it in a different class: (1) first processing; (2) further processing; (3) first and further
 processing; (4) first processing and rendering; (5) further processing and rendering; or (6) first processing,
 further processing, and rendering. Facilities that only perform the process of rendering are classified as
 Tenderers; if rendering is performed in combination with the other two processes the facility is classified
 with the appropriate meat type (red meat or poultry).

        EPA also classified facilities by discharge type and facility size. Discharge type distinguishes
 those facilities that discharge process wastewater directly into U.S. surface waters (direct dischargers) from
 those that discharge wastewater to treatment works (indirect dischargers).  Under the Clean Water Act,
 EPA may apply different standards to direct and indirect dischargers (see Section 4.2).  Size is determined
 by facility production and wastewater flow and was used to cost the appropriate treatment capacity for the
 facility.  For the purposes of costing, EPA divided facilities in each class into small, medium, large, and
 very large.  Detailed information  on classification can be found in the Development Document (EPA,
 2002).                    •                                                        .

        2.2.1.2 Facility Count by Class, Discharge, and Size

        As mentioned above, data from the Screener Survey sample was used to generate facility classes,
 based on meat type and process.  Moreover, each class is further divided into direct and indirect
                                                2-27

-------
dischargers, as well as four size groupings within the discharging group. Table 2-13 shows the number of
facilities in each of the meat type and process class by discharge type and size.  An analysis by meat type
shows that of the total 5,606 facilities, 70 percent produce red meat, followed by facilities producing mixed
meat at 15 percent of the total. Poultry .producing facilities make up 13 percent of total facilities.  By
process, further processors comprise of 71 percent of all facilities, and first and further processing facilities
are 14 percent of the total. By discharge type, 94 percent of facilities are indirect dischargers. Finally, a
size distribution is as follows: 84 percent of facilities are small, 11 percent are medium, 3 percent large,
and 2 percent very large.                                          .
        2.2.2   Proposal 40 CFR 432 Subcategories

        2.2.2.1 Method of Subcategorization

        The subcategories developed for the proposed rule modify and extend the existing industry
 subcategories. Currently, EPA has subcategorized the industry as follows:
               Subcategory A -
               Subcategory B -
               Subcategory C -
               Subcategory D -
               Subcategory E -
               Subcategory F -
               Subcategory G -
               Subcategory H •
               Subcategory I -
               Subcategory J -
- Simple Slaughterhouse
- Complex Slaughterhouse
- Low-Processing Packinghouse
- High- Processing Packinghouse
- Small Processor
- Meat Cutter
- Sausage and Luncheon Meats Processor
- Ham Processor
• Canned Meats Processor
-Renderer
 Using the meat type and process classes described in Section 2.2.1.1 above, EPA grouped the screener
 survey population into five subcategories.  For the proposed rule, the first four subcategories are combined
 to form Subcategory A through D, the next four are combined to form Subcategory E through I, while
                                               2-28

-------
                            Table 2-13
Facility Count by Meat Type and Process Class, Discharge Type, and Size
Vleat Type
ied Meat
Red Meat
Red Meat
Red Meat
Red Meat
Red Meat
. - .'".•-'•' --•.-"
;-.-," ----•" • " - . • --•
Process.' ;..•. • .."..: ••• '--\ ^.: - ,/':
First Processing



Further Processing

First and Further Processing
First Processing and Rendering
Further Processing and Rendering
First Processing, Further
Processing, and Rendering

''. ' : i •-.
^•••'..'Xi.
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Number of Facilities
Direct
Dischargers
17
6
0
0
43
10
1
1
0
0
0
0
17
17
7
12
0
4
0
0
25
17
7
Indirect
Dischargers
265
o
. 0
0
2,489
160
4
4
674
• . 28
0
0
12
7
3
5
32
7
0
0
50
12
5
                                2-29

-------
                         Table 2-13 (cont.)
Facility Count by Meat Type and Process Class, Discharge Type, and Size
Meat Type

Poultry
Poultry
Poultry
Poultry
Poultry
Poultry
Process

First Processing



Further Processing
First and Further Processing
First Processing and Rendering
Further Processing and Rendering
First Processing, Further
Processing, and Rendering

Size
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Number of Facilities
Direct
Dischargers
0
0
17
25
7
. 0
10
1
2
0
6
2
8
0
7
8
2
0
0
0
0
0
2
Indirect
Dischargers
0
19
32
48
12
272
133
4
18
20
11
4
14
0
2
. 2
1
4
9
6
0
0
3
                                2-30

-------
                         Table 2-13 (cont.)
FacUity Count by Meat Type and Process Class, Discharge Type, and Size
Meat Type

Mixed Meat
Mixed Meat
Either
Process- : '•-;;••...• :'""'-'":' ."••; --.." : ,

Further Processing
Further Processing and Rendering
Rendering

Size :--;,;.-
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Small
Medium
Large
Very Large
Number of Facilities
Direct
Dischargers .;
3
1
9
5
0
0
0
0
0
0
6
7
6
8
Indirect
Dischargers
7
2
707
97
0
0
4
0
0
0
17
26
21
28
                                2-31

-------
Subcategory J is unchanged. The proposed rule creates new subcategories for the poultry industry which is

not regulated under the current effluent guidelines.                                                    ,


        Thus, the structure of the subcategbrization for the proposed rule is as follows:
               red meat facilities that perform first processing alone or in combination with further
               processing and/or rendering are assigned to Subcategory A through D.

               red meat facilities that perform further processing alone or in combination with rendering,
               but no first processing, are assigned to Subcategory E through I.

               facilities that perform rendering but no other processes are assigned to Subcategory J.

               poultry facilities that perform first processing alone or in combination with further
               processing and/or rendering are assigned to Subcategory K.

               poultry facilities that perform further processing alone or in combination with rendering,
               but no first processing, are assigned to Subcategory L.

               mixed facilities — both red meat and poultry — performing further processing are split
               into the two subcategories consisting of further processors (Subcategory E through I and
               Subcategory L).' The facilities are assigned based on average production of each meat type.
               The mixed facilities are divided as such:

               —     for medium, large, and very large facilities, 61 percent are assigned to the red meat
                       further processors (Subcategory E through I), while for small facilities, the ratio is
                       59 percent.

                —     the remaining mixed facilities, i.e., 39 percent of medium, large, and very large
                       facilities and 41 percent of small facilities are assigned to the poultry further
                       processors (Subcategory L).
        2.2.2.2 Facility Count by Subcategories


        Table 2-14 presents the proposed subcategorization and the corresponding meat type and process

 classes that constitute each subcategory. Also shown on the table is the number of facilities in each

 subcategory by discharge type and size (smalls and non smalls only).  Of the total facilities, 58 percent are
 in Subcategory E through I, followed by 21 percent in Subcategory A through D. Subcategory L consists
                                                2-32


-------
                   Table 2-14
Facility Count by Proposal 40 CFR 432 Subcategories,
             Discharge Type, and Size
Meat Type
Processes
Size
Number of Facilities
Direct
Dischargers
Subcategory A through D
Red Meat
First Processing;
First Processing and Further Processing;
First Processing and Rendering; and
First, Further Processing, and Rendering
Small
Non Small
59
66
Subcategory E through I
Red Meat
Further Processing;
Further Processing and Rendering;
Mixed Meat Further Processing; and
Mixed Meat Further Processing and
Rendering
Small
Non Small
48
19
Subcategory J
Red Meat
or Poultry
Rendering
Small
Non Small
6
. 21
Subcategory K
Poultry
First Processing;
First Processing and Further Processing;
First Processing and Rendering; and
First, Further Processing, and Rendering
Small
Non Small
0
88
Subcategory L
Poultry
Further Processing;
Further Processing and Rendering;
Mixed Meat Further Processing; and
Mixed Meat Further Processing and
Rendering
Small
Non Small
4
15
Indirect
Dischargers

1,001
60

2,940
234

17
75

39
138

568
208
                      2-33

-------
of 14 percent of facilities, 5 percent of facilities belong to Subcategory L, and the remaining 2 percent
belong to Subcategory J.
2.3    TRENDS IN PRODUCTION, PRICES, AND INTERNATIONAL TRADE

       2.3.1   Aggregate Industry Trends

       2.3.1.1 Domestic Production and International Trade Trends

       The 7997 Economic Census provides a detailed snapshot of the meat products industry in 1997.
The screener survey provides detailed data that allows EPA to analyze and subcategorize the industry.
However, neither provide information on industry trends. Furthermore, due to the switch from the SIC
code system to the NAICS system, it is difficult to reliably interpret Census time series data for the     ..   •
industry. EPA used data from a variety of sources, primarily the U.S. Department of Agriculture (USDA),
to characterize industry trends.         .                                       *

        Table 2-15 presents data on aggregate beef production and trade from 1980 to 2000. Overall
 domestic production grew at an average annual rate of little more than 1 percent from 1980 to 2000,
 although the industry grew at a faster rate of 1.7 percent per year from 1990 to 2000. The significant role
 of international trade in sustaining industry growth is readily apparent from Table 2-15. Beef exports grew
 by 12 percent per year from 1980 to 2000, and at a somewhat slower but still robust 9.5 percent Since
 1990. Exports now make up almost 10 percent of domestic production. The year to year volatility of beef
 exports is also apparent in the data. Note that despite this substantial growth in beef exports, the U.S. is a
 net importer of beef.

         table 2-16 presents data on aggregate pork production and trade from 1980 to 2000. Domestic
 output of pork grew more slowly than either beef or chicken in this time frame. Although pork exports grew
 more quickly than beef exports, pork exports  as a percent of domestic production are less significant than
 beef exports. However, because pork exports  have maintained a double-digit annual growth rate since
                                               2-34

-------
                                          Table 2-15
                       Beef Production, Exports and Imports, 1980-2000

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994 .
1995
1996
1997
1998
1999
2000
Beef (million of pounds, carcass weight)
Domestic
Production
21,643
22,389
22,536
23,243
23,598
23,728
24,371
23,566
23,589
23,087
22,743
22,917
23,086
23,049
24,386
25,222
25,525
25,490
25,653
26,386
26,777
&vg., 1980-2000
Avg., 1990-2000
Avg., 1995-2000
Percent
Change

3.4%
0.7%
3.1%
1.5%
0.6%
2.7%
-3.3%
0.1%
-2.1%
-1.5%
0.8%
0.7%
-0.2%
5.8%
3.4%
1.2%
-0.1%
0.6%
2.9%
1.5%
1.1%
1.7%
1.6%
Imports
2,064
1,743
1,939
1,974
1,823
2,071
2,129
2,269
2,379
2; 178
2,356
2,406
2,440
2,401
2,369
2,103
2,073
2,343
2,642
. 2,874
3,032



Percent
Change*

-15.6%
11.2%
1.8%
-7.6%
13.6%
2.8%
6.6%
4.8%
-8.4%
8.2%
2.1%
1.4%
-1.6%
-1,3%
-11.2%
-1.4%
13.0%
12.8%
8.8%
5.5%
2.8%
3.3%
5.3%
. • - . = .
Exports
173
216
250
268
323
325
516
600
680
1,023
1,006
1,188
1,324
1,275
1,611
1,821
1,877
2,136
2,171
2,417
2,516



Percent
Change

24.9%
15.7%
7.2%
20.5%
0.6%
58.8%
16.3%
13.3%
50.4%
-1.7%
18.1%
11.4%
-3.7%
26.4%
13.0%
3.1%
13.8%
1.6%
11.3%
4.1%
12.4%
9.5%
7.7%

As Percent of
Domestic
Production
0.8%
1.0%
1.1%
1.2%
1.4%
1.4%
2.1%
2.5%
- 2.9%
4.4%
4.4%
5.2%
5.7%
5.5%
6.6%
7.2%
7.4%
8.4%
8.5%
9.2%
9.4%

Source: 1980-1997 data: Putnam & Allshouse,
Situation & Outlook, 12/27/00 and 8/29/01.
1999, extended through 2000 from Livestock, Dairy & Poultry:
                                              2-35

-------
                                          Table 2-16
                       Pork Production, Exports and Imports, 1980-2000

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Pork Cmillion of pounds, carcass weight)
Domestic
Production
16,617
15,873
14,229
15,199
14,812
14,807
14,063
14,373
15,684
15,813
15,354
15,999
17,233
17,088
17,696
17,849
17,117
17,274
18,981
19,278
18,928
\vg., 1980-2000
Vvg., 1990-2000
\vg 1995-2000
Percent
Change

-4.5%
-10.4%
6.8%
-2.5%
-0.0%
-5.0%
2.2%
9.1%
0.8%
-2.9%
4.2%
7.7%
-0.8%
3.6%
0.9%
-4.1%
0.9%
9.9%
1.6%
-1,8%
1.0%
2.2%
1.3%
Imports
550
542
612
707
954
1,128
1,122
1,195
1,137
896
898
775
645
740
743
664
618
633
704
827
967



Percent.
Change

-1.5%
12.9%
15.5%
34.9%
18.2%
-0.5%
6.5%
-4.9%
-21.2%
0.2%
-13.7%
-16.8%
14.7%
0.4%
-10.6%
-6.9%
2.4%
'11.2%
17.5%
16.9%
4.5%
2.6%
6.6%
Exports
252
307
214
219
164
128
86
109
195
262
238
283
420
446
549
787
970
1,044
1:229
1,278
1,305



Percent
Change

21.8%
-30.3%-
2.3%
-25.1%
-22.0%
-32.8%
26.7%
78.9%
34.4%
-9.2%
18.9%
48.4%
6.2%
23.1%
43.4%
23.3%
7.6%
17.7%
4.0%
2.1%
14.8%
16.3%
14.3%

\s Percent of
Domestic
Production
1.5%
1.9%
1.5%
1.4%
1.1%
0.9%
0.6%
0.8%
• 1.2%
1.7%
1.6%
1.8%
2.4%
2..6%
3.1%
4.4%
5.7%
6.0%
6.5%
6.6%
6.9%



Source: 1980-1997 data: Putnam & Allshouse, 1999, extended through 2000 from Livestock, Dairy & Poultry:
Situation & Outlook, 12/27/00 and 8/29/01.
                                              2-36

-------
1995, this may change. Note that despite the substantial growth in pork exports, the U.S. is a net importer
of pork.

       Table 2-17 presents data on aggregate broiler production and exports from 1980 to 2000; broiler
imports have generally been negligible and were not included in the table. Domestic broiler production
maintained an average 5 percent annual growth rate over the 20-year period. Carcass weight of broiler
production expanded from approximately 50 percent of beef production in 1980 to 114 percent of beef.
production in 2000. Broiler exports grew at 15 percent per year over the 20-year period, and now account
for more than 18 percent of domestic production.10

        Table 2-18 presents data on aggregate domestic turkey production from 1999 to 2000; turkey
imports have generally been negligible and were not included in the table. While overall turkey production
grew at 4 percent per year over the 1980 to 2000 time frame—much faster than beef or pork, and only
slightly slower than broilers—growth has slowed considerably since 1995. The pattern for turkey exports is
similar: a high growth rate over the entire 20-year period is offset by a significant slowdown in the last 5
years.                                                       .                                    .

        The above tables amply illustrate the importance of international trade for U.S. meat producers.
Most of this growth in trade of meat and poultry has been attributed to a liberalized trading environment.
Trade agreements  like NAFTA, for example, have spurred the growth of U.S.  beef exports to Canada and
poultry exports to Mexico (USDA, 1997b; USITC, 1998). Yet there exist trade barriers, such as health and
 sanitary concerns, that have been preventing the U.S. from open access to certain markets. For instance, the
European Union bans the importation of U.S. beef produced with growth-promoting agents (USDA,
 1997b). Industry organizations like the National Pork Producers Council and the National Cattlemen's
 Beef Association believe that the growth of U.S. exports has been limited by unfair and unscientific non-
 tariff trade barriers on meat imports imposed by some countries (NPPC, 1999b; NCBA, 2000a).
         10 Focus on aggregate trade statistics can sometimes obscure significant differences in traded goods. For
 example, Russia is a significant importer of U.S. poultry products. Russian consumers prefer dark poultry meat
 while U.S. consumers prefer white meat. Thus, trade with Russia provides an important element of complementary
 production for U.S. poultry processors (Standard & Poor's, 1999).
                                                2-37

-------
                                Table 2-17
                Broiler Production and Exports, 1980-2000

Year
1980
1981 .
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993 '
1994
1995
1996
1997
1998
1999
2000
Broilers (million of pounds, ready-to-cook carcass weight)
Domestic
Production
11,252
11,868
11,996
12,326
12,921
13,520
14,180
15,413
16,007
17,227
18,430
19,591
20,904
22,015
' 23,666
24,827
26,124
27,041
27,863
29,741
30,485
\vg., 1980-2000
\ve., 1990-2000
<\vg 1995-2000
Percent
Change

5.5%
1.1%
218%
4.8%
4.6%
4.9%
8.7%
3.9%
7.6%
7.0%
6.3%
6.7%
5.3%
7.5%
4.9%
5.2%
3.5%
3.0%
6.7%
2.5%
5.1%
5.0%
4.3%
Exports
567
719
501
432
407
417
566
752
765
814
1,143
1,261
1,489
1,966
2,876
3,894
4,420
4,664
4,673
4,920
5,548



Percent
Change

26.8%
-30.3%
-13.8%
-5.8%
2.5%
35.7%
32.9%
1.7%
6.4%
40.4%
10.3%
18.1%
32.0%
46.3%
35.4%
13.5%
5.5%
0.2%
5.3%
12.8%
15.4%
15.6%
11.4%
As Percent
of Domestic
Production
5.0%
6.1%
4.2%
3.5%
3.1%
3.1%
4.0%
4.9%
4.8%
4.7%
6.2%
6.4%
7.1%
8.9%
12.2%
15.7%
16.9%
17.2%
16.8%
16.5%
18.2%



Source: 1980-1997 data: Putnam & Allshouse, 1999; extended
Livestock, Dairy & Poultry: Situation & Outlook, 12/27/00 and
Broiler imports are negligible.
through 200U trom
8/29/01.
                                     2-38

-------
                                 Table 2-18
                Turkey Production and Exports, 1980-2000

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
Turkey (million of pounds, ready-to-cook carcass weight)
Domestic
Production
2,370
2,536
2,472
2,590
2,601
2,817
3,155
3,701
3,879
4,136
4,514
4,603
4,777
4,798
4,937
5,069
.5,401
5,412
5,28
5 297
5,402
Lvg., 1980-2000
Lvg., 1990-2000
Ikvg., 1995-2000
Percent
Change

7.0%
-2.5%
4.8%
0.4%
8.3%
12.0%
17.3%
4.8%
6.6%
9.1%
2.0%
3.8%
. 0.4%
2.9%
2.7%
6.5%
0.2%
-2.47
0.37
2.0<7
4.07
1.87
1.67
Exports
75
63
51
47
27
27
27
33
51
41
5^
122
202
244
280
. 348
438
598
446
379
458



Percent
Change :

-16.0%
-19.0%
-7.8%
-42.6%
0.0%
0.0%
22.27o
54.5%
-19.6%
31.7%
125.97o
65.67o
20.8%
14.87o
24.3%
25.97o
36.5%
-25.4%
-15.0%
20.87
18.57
20.57
12.87
As Percent
of Domestic
Production -
3.2%
2.5%
2.1%
1.8%
1.0%
1.0%
0.9%
0.9%
1.3%
1.0%
1.2%
2.7%
4.2%
5.1%
5.7%
6.9%
8.1%
11.0%
8.4%
7.27o
8.5%



OULULwt*. JL^UW—JL ^f -S I x*c*w*ป J. ป*"ปป**•" ***- ******-.-—	j	   i 	           v
Livestock, Dairy & Poultry: Situation & Outlook, 12/27/00 and 8/29/01.
Turkey imports are negligible.
                                      2-39

-------
       2.3.1.2 Price Trends

       Table 2-19 compares the overall trend in meat prices with all food prices and the Consumer Price
Index (CPI). While food prices have increased more slowly than the overall CPI, meat prices have
increased even more slowly than food prices. Thus,, the price of meat products has decreased relative to the
prices for other food products, and other products in general, over the last 20 years. Beef prices have
consistently increased more slowly than pork or poultry, which again reflects a relative decline in the price
of beef (see Table 2-20). The price of pork as a component of the aggregate meat measure has increased at
approximately the same rate as overall food prices for the 20-year period—with some fluctuation in
subperiods. As with beef, the price of poultry  has risen more slowly than the overall price of food,
indicating a relative price decrease for poultry. The decrease in the relative price of poultry combined with
the high growth rate of output indicates that productivity gains in the poultry industry were probably
substantial.
       23.2   Industry Response to Changing Consumer Preferences

       Meat consumption patterns in the U.S. have undergone important changes in the last three decades.
While total meat consumption increased from 1970 to 1998, the relative consumption of red meats to
poultry has not remained the same. Total annual per capita meat consumption (boneless, trimmed
equivalent) increased by 10 percent between 1970 and 1998. In the same period, annual per capita red meat
consumption declined by 12 percent and per capita poultry consumption increased by 92 percent. More
specifically, annual per capita beef consumption decreased 18 percent (from 79.6 pounds in 1970 to 64.9
pounds in 1998). Annual per capita pork consumption levels remained relatively constant over the same
period. Chicken and turkey consumption, however, increased dramatically: annual per capita chicken
consumption increased from-27.4 pounds in 1970 to 50.8 pounds in 1998 (an 85 percent increase). Turkey
consumption increased from 6.4 pounds per capita to 14.2 pounds over the same period (Putnam and
Allshouse, 1999).           •

       Changes in demand for meat and poultry are considered to be largely responsible for higher degrees
of concentration in the meat and poultry industry. In the case of beef, declining pec capita consumption
                                              2-40

-------
                                      Table 2-19
             Consumer Price Index for All Items, Food, and Meat, 1980-2000

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
All Items
Index
82.4
90.9
96.5
99.6
103.9
107.6
109.6
113.6
118.3
124.0
130.7
136.2
140.3
144.5
148.2
152.4
156.9
160.5
163.0
166.6
172.2

-------

-------
resulted in a corresponding decrease in the demand for cattle. Combined with increased fabrication by beef
slaughterers, the decreasing demand for cattle led to increased concentration in the cattle slaughter industry.
Per capita demand for pork has remained relatively constant through the years, so concentration in hog
slaughter has increased only slightly. Increasing per capita consumption of chicken and turkey
consumption, on the other hand, has tended to limit industry concentration in poultry production (see
Section 2.4 for a more detailed discussion on concentration in the meat products industry),

        There are several causes for these demand related changes in the volume and composition of meat
consumption. First, meat has become increasingly affordable to consumers in recent years. Even though per
capita consumption of meat has increased, the percentage of income spent on meat purchase has decreased
from 4.3 to 2.2 percent in the last quarter century (AMI, 2000a). Second, health and safety concerns
regarding red meat have been found to result in lower per capita beef consumption and increased per capita
poultry consumption (Flake and Patterson, 1999; Moon and Ward, 1999). Third, as noted above, the
relative price of chicken to beef has declined; combined with increased chicken production, these data are
 consistent with increased productivity in the industry, thus reinforcing the apparent increase in demand for
 poultry.

        In addition to the abovementioned overall changes in the markets for meat products, a change in
 marketing strategy by poultry producers shifted retail packaging of chicken from whole birds to a product
 mix of traypacks, parts, and other further processed products (Hetrick, 1994). The poultry industry also
 started the branding of processed products (Ollinger, 2000). The introduction of such retail marketing
 strategies has apparently increased consumer demand for reasonably priced and convenient value-added
 branded chicken products.

         Value-added products include case-ready and consumer-ready meats. Case-ready meats are
 trimmed, precut, processed, portion controlled, sealed directly by the processor, and sold to supermarkets
 ready for purchase. In  addition to traditional cuts, case-ready meats include whole muscle portions and
 even ground beef (Krizner, 1998). Consumer-ready meat products, also known as home-meal-replacement
 items, include microwavable, oven-ready, and other ready-to-cook items.11 The advantages of case-ready
          11 Meat producers have begun producing consumer-ready products in an attempt to regain some of the
  business that the increased popularity of eating out has cost them (Rice, 1998). These meals are catered to
  consumers seeking convenience and nutrition in meal preparation.
                                                 2-43

-------
 meats for retailers include extended shelf life, reduced labor costs, and fewer out-of-stocks. Consumers
 benefit from the consistency, unproved quality and packaging, and safety of a product untouched by human
 hands (Nunes, 1999; AMI, 2000b). Branded meat products, unlike private and store labels, are those
 processed by meat producers themselves, using the highest-quality animals, and sold directly to consumers.
        By further processing meat, integrated producers may be able to reduce the impact of price
 fluctuations in the related commodity markets (Standard & Poor's, 1999).12 Value-added products can
 benefit meat producers by potentially giving them-more control over the pricing of their products. Other
 benefits to packers include control over all aspects of the production process and, possibly, brand name
 recognition (Krizner, 1995). The branding of products allows a producer to differentiate its product from
 its competitors' and to certify the quality of its products.                         .             .

        As mentioned above, these product trends were pioneered by the chicken industry in the 1970s and
 emulated by the turkey industry in the 1980s. To emphasize the lower fat content in chicken, slaughterers
 produced further processed poultry such as traypacks, cut up and deboned chicken, nuggets, and luncheon
 nr'Us (.Ollinger, 2000). Pork production, on the other hand, has traditionally involved the furt'- ^r
 processing of the meat into hams, bacon, and sausages. Today, almost all poultry products and one-haif to
 two-thirds of pork products are consumer-ready (AMI, 2000b). Product branding of chicken and turkey,
 introduced in the 1960s, was met with positive consumer response; brand loyalty was achieved as
 consumers perceived certain branded products to be of higher quality. Beef producers are also adopting
 such retail strategies by establishing case-ready plants and branding their products.13
        12 This was an important reason for the sale of Iowa Beef Processors, Inc., or IBP (WSJ, 2000). In the past
DBP's stock price has been affected by price fluctuations in the commodity markets. IBP managers believed that
Stock market perceptions of the company did not account for IBP's shift away from commodity production toward
more value-added production. IBP managers felt that privatizing the company would better, insulate its corporate
valuation from commodity price fluctuations.
        13 In a study conducted by the National Cattlemen's Beef Association, consumers were indifferent to
consumer brand names. Nonetheless, beef processors are eager to differentiate their case-ready meats from others
(Nunes, 1999). The introduction of case-ready meats in the beef industry was attempted in the 1980s, but was not
successful until the late 1990s. Early problems included hesitation on the part of retailers and consumer concerns
with the appearance and packaging of the product (Krizner, 1998).
                                                2-44

-------
This is one of several strategic responses of beef producers and packers to decreasing red meat
consumption.14
2.4     INDUSTRY CONCENTRATION


        Another trend in the meat products industry is the growing concentration of industry output in a

handful of large companies. This trend is most dramatic in the beef slaughter industry, but is also evident in

the pork and poultry industries. Industry trends in all three meat industries give rise to two important

questions:

        •      What caused the increased concentration in each industry?

                Has this increased industry concentration led to market power on the part of the largest
                firms?

 The answers to both of these questions could have implications for the economic impact analysis.


         This report's discussion of trends in the beef and pork slaughter industries is based on statistics

 published by Packers and Stockyards Program (PSP) of the Grain Inspection, Packers, and Stockyards

 Administration (GIPSA). GIPSA was established in  1994 by USDA, but its roots lie in the U.S.

 Congress's Packers and Stockyards Act of 1921.


         PSP maintains time series data on slaughtering plants that purchase at least $500,000 worth of

 livestock in a fiscal year. Thus, many small slaughtering facilities are exempt from PSP reporting

 requirements, and the number of facilities reporting in any one year will fluctuate. However, plants meeting

 the PSP reporting requirements accounted for 97 percent of federally inspected slaughter and 95 percent of
          14 Another important response to declining red meat consumption has been a reduction in the fat content
  of beef and pork. The fat content in beef has declined 27 percent since the 1970s, and nutrient information on beef
  packaging now reflects this (NCBA, 2000b). This reduction in the fat content was made possible by the breeding of
  leaner animals and the use of leaner cuts in beef production. The beef industry has also been promoting the
  nutriuonal value of lean beef with research and marketing campaigns (Carpenter, 2000). In addition, the trend
  toward leaner meat is also seen in the pork slaughter industry, where there has been a 50 percent reduction m hog
  fat since the 1950s (NPPC, 1999a).
                                                 2-45

-------
commercial slaughter in the heifer and steer class and the hog class. For this profile, EPA mainly used the
PSP data to examine industry trends. As is described below, the trends in cattle and hog classes are
unmistakable, and small fluctuations in the number of plants reporting do not affect the conclusions drawn
from this data.

        Unfortunately, there is no publicly available source of data on trends in the poultry industry.
Discussion of trends in the poultry industry is therefore based on other researchers' analysis of the Census
Bureau's Longitudinal Research Database. Though it does not provide the same wealth of detail as PSP,
this source more than adequately documents the trend towards concentration in the poultry industry.

        Section 2.4.1 discusses, in turn, the trends toward concentration in the beef, pork, and broiler
industries. Section 2.4.2 describes the changes in these industries that may be responsible for the trend
toward concentration. Finally, Section 2.4.3 presents a summary of studies that have examined if the trend
toward concentration has given market power to large firms in these industries.
         2.4.1   Trends in Industry Concentration

         Beef

         Discussion of the beef industry will primarily focus on the heifer and steer industry segment. The
 generic term "cattle" applies to two distinct groups of animals: (1) heifers and steers and (2) cows and
 bulls. Heifers and steers are raised specifically for meat production and are corn fed prior to slaughter.
 Cows and bulls are generally culled from dairy and breeder herds and fed on grass and forage. Slaughter
 plants typically specialize in one of the two types due to differences in animal shapes and meat products
 from the two types (cows are generally used to make ground beef). Cow slaughter plants tend to be smaller
 and more geographically diversified than heifer and steer slaughter plants; sale lots of culled cows and bulls
 tend to be small, and the dairy industry is more geographically diversified than the beef industry
  (MacDonald et aL, 2000; Mathews et al., 1999). In 1998, GIPSA plants reported slaughter of 27.4 million
  steers and heifers, and 6.4 million cows and bulls.
                                                 2-46

-------
       Table 2-21 presents total annual steer and Heifer slaughter by head in slaughter plants reporting to
PSP from 1972 to 1998. Total steer and heifer slaughter increased by approximately 5 percent over that
period. More significant than the growth rate of steer and heifer slaughter is the distribution of that
slaughter among plants of various size. Plants that slaughtered fewer than 50,000 head per year accounted
for almost 21 percent of total slaughter in 1972 (5.4 million head); this had fallen to less than 3 percent of
total slaughter by 1998 (0.7 million head). Similarly, plants that slaughtered between 50,000 and 250,000
head per year accounted for almost 50 percent of total slaughter in 1972 (12.9 million head), but less than
5 percent in 1998 (L2 million head). Conversely, the largest plants—those slaughtering more than 250,000
head per year-increased their share of total slaughter from 30 percent in 1972 (7.8 million head) to almost
93 percent in 1998 (25.4 million head).

         The increased percentage of annual slaughter accounted for by the largest plants is much less than
 proportionate to the increased number of large plants in the industry, although the trend is similar. As
 Table 2 ฃ> shows, the number of plants with capacity in excess of 250,000 head increased from 20 :ti 1972
 (2.5 percent of plants reporting to PSP) to about 28 in 1998 (17 percent of plants reporting to PSP). The
 number of plants with the smallest capacity (below 50,000 head) fell from over 660 in 1972 to  130 in 1998
 (still 77.4 percent of PSP-reportmg plants), while intermediate plants (capacity between 50,000 and
 250,000 head) declined from 120 in 1972 to 10 in 1998.

         Thus, Tables 2-21 and 2-22 illustrate not only a shift from slaughtering by many small facilities to
  a handful of large facilities, but also a significant increase in the size of the largest facilities. In 1972, the
  largest-capacity facilities slaughtered an average of 389,000 head per plant; by 1998, that had  grown to an
  average of 909,000 head per plant.15

         The trend in the number of plants by size, as well as in slaughter by plant size, is mirrored by
  industry measures of concentration at the firm level. The percentage of annual commercial heifer and steer
  slaughter accounted for by the four largest firms in the industry (i.e., the four-firm concentration ratio, or
          15 Cow and bull slaughter shows a roughly similar pattern over the same period, but a much less extreme
  one. Slaughter was almost unchanged at about 6.4 million head in both 1972 and 1998. The number of plants
  declined for all but the largest-capacity plants (those with slaughter in excess of 100,000 head per year) which
  increased from 6 to 26 plants from 1972 to 1998. The average slaughter per facility rose from 133,500 to 192,500
  over the same period for the largest-capacity facilities.
                                                  2-47

-------
                                       Table 2-21
                                 Annual Heifer and Steer
                            Slaughter by Plant Size, 1972-1998


Year
1972
1973
1974
1975
1976
•1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997

Steers and Heifer Slaughter by Plant Size (Annual Slaughter by Head)
Less than 49,999
Head
(1,000s)
5,416
5,212
5,010
4,889
4,506
4,316
4,239
3,716
3,446
2723
2,436
2,238
2,141
1,947
1,623
1,264
1,257
1,156
987
860
711
684
717
627
686
61

Percent
20.7%
20.7%
19.7%
19.1%
16.7%
14.9%
14.9%
14.5%
14.1%
10.b%
9.6%
8.6%
8.2%
7.2%
6.1%
4.7%
4.6%
4.5%
3.8%
3.4%
2.8%
2.7%
2.7%
2.37
2.47
2.27

50,000-249,999
Head
(1,000s)
12,939
11,340
11,934
12,147
13,044
12,949
12,208
10,537
8,876
. 7,330
6,790
5,929
6,201
5,642
4,532
5,439
3,926
3,032
2,535
3,024
2,287
2,142
1,418
1,902
1,587
1,712
' L257
Percent
49.5%
45.0%
47.0%
47.5%
48.4%
44.6%
43.0%
41.1%
36.3%
29.1%
26.7%
22.8%
23.6%
20.9%
17.0%
20.0%
14.4%
11.7%
9.8%
11.9%
9.0%
8.4%
5.4%
7.0%
5.6%
6.27
4.67
More Than 249,999
Head
ouoobs)
7,778
8,657
8,457
8,536
9,408
11,785
11,930
11,359
12,157
15,171
16,250
17,879
17,897
19,433
20,482
20,443
21,992
21,698
22,238
21,495
22,293
22,725
23,992
24,820
26,062
25,490
25.439
• - - •
Percent
29.8%
34.3%
33.3%
33.4%
34.9%
40.6%
42.0%
44.4%
49.7%
60.1%
63.8%
68.6%
68.2%
71.9%
76.9%
75.3%
80.9%
83.8%
86.3%
84.7%
88.1%
88.97
91.87
90.87
92.07
91.67
92.97
Total Head
(1,000s)
26,133
25,209
25,401
25,572
26,958
29,050
28,377
25,612
24,479
25,224
25,476
26,046
26,239
27,022
26,637
27,146
27,175
25,886
25,760
25,379
25,291
25,551
26,127
27,349
28,335
27,813
27,396
Source: GIPSA, 1997; GIPSA, 2000.
                                            2-48

-------
          Table 2-22
Heifer and Steer Slaughter Plants
    by Plant Size, 1972-1998
1 — H

••w-
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
11983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Source: GI
Steer and Heifer Annual Slausshter (Number of Head)
Less than 50,000
Plants
666
660
615
597
591
542
552
529
520
442
422
423
397
359
315
314
309
262
257
237
222
218
191
173
165
ise
	 13C
>SA, 1997; GD
Percent ;:
82.5%
83.0%
81.8%
81.2%
80.3%
78.3%
80.8%
82.1%
83.1%
82.0%
82.3%
82.9%
82.9%
823%
82.0%
81.3%
82.6%
82.1%
82.9%
82.3%
82.8%
83.5%
83.0%
80.1%
78.2%
78.4%
77.4%
PSA, 2000.
50,000-249,999
Plants
121
112
115
116
123
123
105
91
8C
65
59
5^
51
46
39
43
33
25
20
2:
17
1
1
1^
\i
1
1
Percent
15.0%
14.1%
15.3%
15.8%
16.7%
17.8%
15.4%
14.1%
12.8%
12.1%
11.5%
10.6%
10.6%
10.6%
10.2%
11.1%
8.8%
7.87
• 6.57
. 7.37
6.37
5.77
4.87
6.57
6.67
6.57
6.07

More Than 249,999
"Plants;
20
23
22
22
22
27
26
24
26
32
32
33
31
3:
30
29
32
32
33
30
2
2
2
2
3
3
2
ฑ====
^Percent
2.5%
2.97o
2.97o
3.07o
3.07o
3.97o
3.8%
3.7%
4.2%
5.9%
6.27o
6.57o
6.57o
7.1%
7.87o
7.5%
8.67o
10.07o
10.67
10.47
10.87
10.77
12.27
13.47
15.27
15.17
16.77
Total
Plants
807
795
752
735
736
692
683
644
626
539
513
510
479
436
384
386
374
319
310
288
268
261
230
216
211
199
168

               2-49

-------
CR-4) increased from approximately 36 percent in 1980 to 80 percent in 1998 (Table 2-23).16 The CR-4
grew more rapidly than either the CR-8 or the CR-20. When the four largest firms in an industry account
for more than 50 percent of output, some economists argue, those firms may be starting to acquire
significant market power (Rogers and Sexton, 1994). The Herfindahl-Hirshman index (HHI) also
demonstrates increasing market concentration in the beef slaughtering industry, increasing from 561 in
1980 to 1,921 in 1998 (Table 2-23). The U.S. Department of Justice and the Federal Trade Commission
regard a market with an Hffl in excess of 1,000 to be moderately concentrated, and one with an Hffl above
1,800 to be highly concentrated (Mathews et ah, 1999)."

        However, increased firm-level concentration cannot be entirely attributed to die increased number
of large facilities and the growth in facility size. As Table 2-24 shows, the pattern of facility ownership
among the largest firms differs from that among smaller firms. The four largest firms each own, on
average, six slaughter facilities large enough to meet reporting requirements for PSP. The fifth through
eighth largest firms own, on  average, two slaughter facilities each (with a distinct downward trend, over
time, in facilities owned); smaller firms typically own one slaughter facility.
        Pork
        The pork industry displays many of the same trends as the beef industry. However, these trends
 have not been as strong as in the beef industry, and the pork industry has not reached the same degree of
 concentration as the beef industry.

        Table 2-25 presents total annual hog slaughter by head in slaughter plants reporting to PSP from
 1972 to 1998. Growth in total hog slaughter was double that of beef: slaughter increased by approximately
 10 percent between 1972 and 1998. The distribution of that slaughter among plants of various size was
         16 MacDonald et al. (2000) claim that no other industry has experienced as rapid an increase in
 concentration over any 15-year period as the cattle slaughter industry.
         17 One calculates HHI by taking the square of each firm's market share, then summing over all firms.
 Thus, a market consisting of 100 firms, each with a 1 percent market share, has an HHI of 100; a market
 consisting of one firm with a 100 percent market share has an Hffl of 10,000. The HHI is generally considered, a
 more reliable indicator of market concentration than the CR-4 (Mathews et al., 1999).
                                                2-50

-------
                      Table 2-23
                Concentration Ratios and
               Herfindahl-Hirshman Index
         for Steer and Heifer Slaughter, 1980-1998
                                             lerfindahl-
                                             Hirshman
                                               Index
Source: GIPSA, 2000.
                            2-51

-------
                                      Table 2-24
                       Firms Performing Steer and Heifer Slaughter
                     Number of Plants Owned by Firm Size, 1980-1998


Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Plants Owned by Steer and Heifer Slaughter Firms (Ranked by Size)
Rank 1-4
Total
23
23
20
22
23
20
21
28
27
25
26
29
26
28
25
27
28
27
25
Average
5.8
5.8
5.0
5.5
5.8
5.0
5.3
7.0
6.8
6.3
6.5
7.3
6.5
7.0
6.3
6.8
7.0
6.8
6.3
Rank 5-8
Total
24
19
18
8
9
9
9
10
12
12
10
6
10
8

t
6
6
7
Average
6.0
4.8
4.5
2.0
2.3
, 2.3
2.3
2.5
3.0
3.0
2.5
1.5
2.5
2.0
1.8
1.3
1.5
1.5
1.8
Rank 9-20
Total
19
18
17
22
20
21
19
18
18
15
16
12
12
12
12
12
12
13
13
Average
1.6
1.5
1.4
1.8
1.7
1.8
1.6
1.5
1.5
1.3
1.3
1.0
1.0
1.0
1.0
1.0
1.0
1.1
1.1
Rank 21-50
Total
37
, 40
37
41
37
41
41
39
40
32
32
32
37
36
35
37
38
34
32
Average
1.2
1.3
1.2
1.4
1.2
1.4
1.4
1.3
1.3
1.1
1.1
1.1
1.2
1.2
1.2
1.2
1.3
1-1
1.1
Source: GIPSA, 2000.
                                           2-52

-------
                                      Table 2-25
                                 Annual Hog Slaughter
                                By Plant Size, 1972-1998
1

Year
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
11984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Hog Slaughter by Plant Size (Animal Slaughter by Head)
Less Than 99,999
Head
(1,000s)
6,380
. 5,630
5,364
4,651
4,603
4,779
4,850
4,568
4822
5,134
4,748
4,536
4,301
3,977
3,841
3,714
3992
3,963
3,784
3,825
3,915
3,755
3,685
3,508
3,093
3,12f
|| 1998 1 2,764
Percent
7.6%
7.4%
6.9%
6.8%
6.7%
6.4%
6.5%
5.6%
5.2%
6.0%
' 5.8%
5.8%
5.2%
4.9%
4.8%
4.8%
4.8%
4.8%
4.7%
4.6%
4.3%
4.2%
4.1%
3.8%
3.7%
3.6%
1- 3.0%
100,000-299,999
Head
(1,000s)
9,410
9,970
8,153
8,748
9,216
7,754
8,073
6,446
5,601
4,666
5,359
6,402
5,859
4,540
3,930
2992
2,720
3,250
2,861
2,423
2,715
1,591
1,796
2,719
2,605
2,55C
, 2,27^
_L^______
Percent
11.2%
13.1%
10.5%
12.7%
13.4%
10.4%
10.8%
7.8%
6.0%
5.4%
6.5%
8:1%
7.1%
5.6%
4.9%
3.9%
3:3%
3.9%
3.6%
2.9%
3.0%
1.8%
2.0%
3.0%
3.1%
1 2.9%
300,000-999,999
Head
(1,000s)
37,894
35,933
38,452
38,961
36,169
34,132
30,137
22,970
23,998
24,95C
23,18(
20,279
23,522
17,920
17,589
14,946
13,826
12,287
9,798
5,249
6,66
7,74^
6,06
6,16
4,75
4,44^
' 2.5% 1 4,28
Percent
45.2%
47.2%
49.5%
56.6%
52.6%
45.6%
40.3%
27.9%
25.8%
29.0%
28.2%
25.8%
28.5%
22.3%
22.1%
19.3%
16.6%
14.8%
12.2%
6.3%
7.3%
8.77
6.87
6.77
5.77
5.17
4.77
More Than 999,999
Head
(1,000s)
30,120
24,661
25,646
16,418
18,828
28,219
31,787
48,~?6
58,504
51,151
48,788
47,491
48,937
53,979
54,398
55,900
62,952
63,687
63,651
71,632
78,258
76,053
77,663
79,222
73,08
77,680
82,469
Percent
35.97o
32.47o
33.07o
23.97o
27.47o
37.7%
42.57o
58.7%
63.0%
59.57o
59.4%
60.37o
59.27o
67.17o
68.2%
72.1%
75.4%
76.67o
79.57o
86.27o
85.57o
85.3%
87.1%
86.57o
87.57o
88.57o
89.87
Total,
Head
(1,000s)
83,804
76,194
77,615
68,778
68,816
74,884
74,847
82,220
92,925
85,901
82,075
78,708
82,619
80,416
79,758
77,552
83,490
83,187
80,094
83,129
91,549
89,143
89,209
91,611
83,529
87,799
91,798
Source: GIPSA, 1997; GIPSA, 2000.
                                           2-53

-------
similar the distribution in the beef industry. Both the number of hogs slaughtered and the percentage of
total hog slaughter declined among all but the largest plants over the time period. Plants that slaughtered
fewer than 300,000 head per year accounted for almost 19 percent of total slaughter in 1972 (15.8 million
head); this had fallen to less than 6 percent of total slaughter by 1998 (5.0 million head). Similarly, plants
that slaughtered between 300,000 and 1 million head per year accounted for 45 percent of total slaughter in
1972 (37.9 million head), but less than 5 percent in 1998 (4.3 million head). Conversely, the largest plants,
those slaughtering more than 1 million head per year, increased their share of total slaughter from 36
percent in 1972 (30.1 million head) to almost 90 percent in 1998 (82.5 million head).

       As with beef slaughter, the increased percentage of annual slaughter accounted for by the largest
plants is  less than proportionate to, the increase in the number of large plants. As Table 2-26 shows, the
number of plants with capacity in excess of 1 million head increased from 23 in 1972 (3.9 percent of plants
reporting to PSP) to about 30 in 1998 (16.5 percent of plants reporting to PSP). (Note that plants in this
capacity  range reached a high of 41 plants in 1980.) The absolute number of plants with lets capacity fell
significantly in all other ranges, but as a percentage of the number of plants in tae industry, the number of
plants with less capacity remained relatively stable.

       Thus, the hog industry mirrors the beef industry, with a shift from slaughtering performed by many
small facilities to slaughtering performed by a handful of large facilities, as well as a significant increase in
the size of the largest facilities. In 1972, the largest-capacity facilities slaughtered an average of 1.3 million
head per plant; by 1998, that had grown to an average of 2.8 million head per plant.

        However, although the hog industry has become much more highly concentrated over the last
quarter century, it has not, by standard economic measures, reached the degree of market concentration
found in the beef industry. The CR-4 for the hog industry (CR-4) increased from approximately 34 percent
in 1980 to 54 percent in 1998 (Table 2-27). This is much lower than the CR-4 for the beef industry in
1998. Also, the CR-8 for hog slaughter actually grew faster than the CR-4.  Similarly, the hog slaughtering
industry  Hffl increased from 436 hi 1980 to 960 in 1998 (see Table 2-27). This is below the benchmark set
by the U.S. Department of Justice and the Federal Trade Commission for moderate concentration (an Hffl
in excess of 1,000).
                                                2-54

-------
                                       Table 2-26
                                  Hog Slaughter Plants
                                 By Plant Size, 1972-1998


Year
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
iHog Annual Slauahter (Number of Head)
LessThan59,999
Plants
463
433
417
380
382
356
354
374
394
380
363
362
341
317
278
278
277
249
272
250
240
216
200
187
175
162
132
Percent
77.6%
76.8%
76.8%
75.7%
76.9%
75.9%
75.8%
77.1%
77.4%
78.4%
77.9%
78.5%
77.7%
78.7%
77.2%
79.2%
79.4%
78.1%
81. 2%
81.4%
80.0%
79.1%
78.7%
76.3%
75:47
74.37
72.57
100,000-499,999^
'Plants
47
51
43
45
45
39
40
34
32
25
27
31
31
23
20
16
15
19
16
14
16
10
11
17
17
16
13
Percent*
7.9%
9.0%
7.9%
9.070
9.1%
8.37o
8.67o
7.070
6.3%
5.2%
5.870
6.7%
7.1%
5.770
5.67o
4.6%
4.370
6.070
4.870
4.670
5.37o
3.77o
4.3%
6.97o
7.37
7.37
7.17
300,000-999,999
Plants
64
61
64
65
56
52
48
41
42
43
41
36
37
29
31
25
24
. 19
16
10
10
13
10
10
Q
(
(
*
Percent
10.7%
10.87o
11.8%
12.97o
11.3%
11.1%
10.3%
8.5%
8.37o
8.9%
8.8%
7.8%
8.4%
1.2%
8.6%
7.1%
6.9%
6.07o
4.8%
3.3%
3.3%
4.8%
3.9%
' 4.1%
3.470
4.1%
3.87
More Than 999,999
Plants
23
19
19
12
14
22
25
36
41
37
35
32
30
34
31
32
33
32
31
33
34
34
33
31
32
31
30
Percent
3.970
3.4%
3.5%
2.4%
2.8%
4.1%
5.4%
. 1.4%
8.1%
7.67o
7.5%
. 6.97o
6.87o
8.470
8.67o
9.1%
9.57o
10.070
9.370
10.7%
11.3%
12.570
13.0%
12.1%
13.87o
14.2%
16.57
Total
Plants
597
564
543
502
497
469
467
485
509
485
466
461
439
403
360
351
349
319
335
307
300
273
254
245
232
218
182
Source: GIPSA, 1997; GIPSA, 2000.
                                           2-55

-------
                      Table 2-27
                Concentration Ratios And
               Herfindahl-Hirshman Index
             . For Hog Slaughter, 1980-1998

Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Concentration Ratio
4 Firm
33.6%
33.3%
35.8%
29.1%
35.0%
32.2%
32.5%
36.6%
33.5%
34.0%
40.3%
41.9%
43.8%
43.5%
44.3%
45.5%
49.6%
54.3%
53.9%
8 Firm
50.9%
48.9%
53.2%
46.0%
53.1%
50.8%
53.6%
55.3%
52.2%
52.4%
58.1%
60.7%
62.6%
65.0%
67.4%
69.4%
69.2%
75.7%
75.4%
20 Finn
71.2%
69.0%
74.7%
68.8%
79.6%
80.5%
84.0%
81.2%
77.8%
78.3%
82.8%
84.4%
- 86.0%
86.1%
85.5%
87.3%
84.1%
89.6%
86.8%
Herfindafal-
Hirshmau
Index
436
411
479
363
487
456
481
516
456
470
593
649
689
704
734
754
797
969
960
Source: GIPSA, 2000.
                           2-56

-------
       Hog slaughter firms' pattern of ownership is similar to that of steer and heifer firms: the largest
firms own several slaughter facilities, and the number of facilities owned declines with firm size. Table 2-
28 presents hog slaughter facility ownership by firm size.
       Poultry

       Although sales of processed poultry products have grown much more rapidly than sales of beef and
pork products over the past 25 years, poultry processing has displayed many of the same trends toward
concentration as the red meat industry. Table 2-29 presents annual liveweight slaughter of young chickens
and turkeys from 1972 to 1995; both chicken and turkey slaughter by weight have more'than tripled since
1972.18 This compares to the 5 percent growth hi cattle slaughter and 10 percent growth in hog slaughter
over the same period.            .    •

        Oiilager et al. (1997) have published estimates of CR-4 for chicken and turkey slaughter based jn
the Census Bureau's Longitudinal Research Database; these estimates are summarized in Table 2-30.
Between 1963 and 1992, the value share of shipments accounted for by the four largest chicken slaughter
firms increased from 14 percent to 41 percent, and from 23 percent to 45 percent for the four largest turkey
slaughter firms. Although not directly comparable with the beef and pork data presented above (because it
is calculated on value shares, not slaughter shares), the data for the chicken and turkey markets show the
same unmistakable trend toward concentration.

        Even more marked than the growth in CR-4, the value share of shipments accounted for by large
facilities increased from 29 percent to 88 percent in chicken slaughtering, and from 16 percent to 83
percent in turkey slaughtering over the 1963 to 1992 period. Although the growing importance of large
facilities in the poultry industry is as striking as in the red meat industry, it should be noted that Ollinger et
al. define larger facilities as those employing more than 24 workers. However, the 7997 Economic Census
of the poultry processing NAICS code (U.S. Census, 1999d) also demonstrates the importance of very
        18 Because the average slaughter weight of both chickens and turkeys has increased over this period, the
 annual slaughter by head has increased at a somewhat slower rate.
                                               2-57

-------
                                     Table 2-28
                            Firms Performing Hog Slaughter
                    Number of Plants Owned by Firm Size, 1980-1998


Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
Plantซ Owned hy Hog Slaughter Firms (Ranked by Size)
Rank 1-4
Total
27
28
26
25
25
23
20
19
16
15
16
15
17
16
17
17
19
19
18
Average
6.8
7.0
6.5
6.3
6.3
5.8
5.0
4.8
4.0
3.8
4.0
3.8
4.3
4.0
4.3
4.3
4.8
4.8
4.5
Rank 5-8
Total
12
11
11
14
13
9
12
1C
9
8
8
8
8
1<
10
11
I
1
9
Average
3.0
2.8
2.8
3.5
3.3
2.3
3.C
2.5
2.2
2.C
2.C
2.(
2.0
3.5
2.5
2.5
2.(
2.:
2.3
Rank 9-20
Total
21
22
25
23
33
32
32
29
3C
25
2^
19
20
15
15
15
15
15
2
___ — — — —
Average
1.8
1.8
2.1
1.9
2.8
2.7
2.7
2.4
2.5
2.1
2.0
1.6
1.7
1.3
1.3
1.3
1.3
1.3
1.8
1
Rank:;
Total
42
43
43
44
41
41
39
41
41
41
40
41
• 37
35
38
39
37
37
3
51-50
Average
1.4
1.4
1.4
1.5
1.4
1.4
1.3
1.4
1.4
1 " 1.4
1.3
1.4
1.2
1.2
1.3
1.3
1:2
1.2
1.0
Source: GIPSA, 2000.
                                          2-58

-------
                                           Table 2-29     -
                              Annual Poultry Production, 19,72 -1995
Year
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
Young
Chicken
Liveweight
Slaughtered
(1,000 Ibs.)
10,957,278
10,858,806
10,999,837
10,982,560
12,407,838
12,740,714
13,656,047
15,111,418
15,530,601
16,349,889
. 16,456,531
16,893,860
17,800,956
18,622,787
19,675,636
21,339,550
22,207,755
23,881,618
25,549,697
27,170,780
28,997,878
30,474,243
32,765,941
34^352,980
Growth
Rate

-0.9%
1.3%
-0.2%
13.0%
2.7%
7.2%
10.7%
2.8%
5.3%
0.7%
2.7%
5.4%
4.6%
5.7%
8.5%
4.1%
7.5%
7.0%
6:3%
6.7%
5.1%
7.5%
4.8%
Ready-to-
Cook Chicken
(1,000 Ibs.)
7,823,383
7,786,095
7,916,834
7,966,103
8,987,270
9,227,289
9,883,206
10,915,517
11,272,385
11,905,743
12,039,023
12,388,980
12,998,613
13,569,204
14,265,627
15,502,464
16,124,400
17,334,190
18,554,511
19,727,657
21,052,418
22,178,143
23,846,169
25,020,790
Growth
Rate

-0.5%
1.7%
0.6%
12.8%
2.7%
7.1%
10.4%
3.3%
5.6%
1.1%
2.9%
4.9%
4.4%
5.1%
8.7%
4.0%
7.5%
7.0%
6.3%
6.7%
5.3%
7.5%
4.9%
Turkeys,
Liveweight
Slaughtered
(1,000 Ibs.)
2,140,783
2,123,718
2,173,898
2,031,627
2,324,808
2,285,685
' 2,418,733
2,643,203
2,823,335
2,060,006
3,003,980
3,156,641
3,187,169
3,444,031
3,879,405
4,609,521
4,876,206
5,191,490
5,684,400
5,798,849
6,040,376
6,075,032
6,279,731
6,456,579
Growth
Rate

-0.8%
2.4%
-6.5%
14.4%
-1.7%
5.8%
9.3%
6.8%
-27.0%
45.8%
5.1%
1.0%
8.1%
12.6%
18.8%
5.8%
6.5%
9.5%
2.0%
4.2%
0.6%
3.4%
2.8%
Ready-to-
Cook Turkey
(1,000 Ibs.)
1,796,505
1,787,912
1,835,821
1,716,053
1,950,111
1,892,479
1,983,476
2,181,794
2,332,381
2,509,107.
2,458,890
2,563,110
2,574,095
2,799,723
3,133,078
3,717,084
. 3,923,452
4,174,874
4,560,901
4,651,915
4,828,939
4,847,657
4,992,225
5,128,816
Growth
Rate

-0.5%
2.7%
-6.5%
13.6%
-3.0%
4.8%
10.0%
6.9%
7.6%
-2.0%
4.2%
0.4%
8.8%
•11.9%
18.6%
5.6%
6.4%
9.2%
2.0%
3.8%
0.4%
3.0%
2.7%
Source: USDA, 1997a.
                                              2-59

-------
                                  Table 2-30
                           Concentration Ratios For
                         Poultry Industry, 1963 -1992
—
Year
1963
1967
1972
1977
1982
1987
• 1992
'
==s=s===========!!=
Chicken Slaughter
Value Share of
Shipments by 4
Largest Finns
14.0%
23.0%
18.0%
22.0%
32.0%
42.0%
41.0%
1
Value Share of
Shipments by
Large Plants1
ND
29.0%
34.0%
45.0%
65.0%
76.0%
. 88.0%
=====
======== . .
Turkey Slaughter
Value Share of
Shipments by 4
Largest Firms
23.0%
28.0%
41.0%
41.0%
40.0%
38.0%
45.0%
*=====
Value Share of
Shipments by
Large Plants1
ND
16.0%
15.0%
29.0%
35.0%
64.0%
j 	 83.0%|
Source: OlUnger, et. al., 1997.
1 Large is defined as plants with more than 24 employees.
                                           2-60

-------
large poultry processing facilities (see Section 2.2.1.3 above). Therefore, the concentration trends in
poultry processing are quite similar to those in the beef and pork industries.
        2.4.2   Facility Size and Economies of Scale

        This section examines potential causes of increased concentration in the meat products industry;
the section that follows examines the results of research into the question of market power in the industry.

        Research into why industry concentration has increased focuses on economies of scale due not only
to changes in technology, but to changes in industry institutional arrangements as well.

        MacDonald et al. (2000) used the U.S. Census Bureau's Longitudinal Research Database to
 examine plant-specific slaughter costs, both over time and between plants. MacDonald et al. adjust the
 output of slaughter plants to account for the trend toward increased fabrication at both cattle and hog
 slaughter plants (e.g., increased production of boxed beef); not only has fabrication become .increasingly
 prevalent as a share of output, but it is correlated with larger plants as well.19 Increased fabrication
 increases production costs, so ignoring the change in product mix would obscure  evidence of technological
 economies of scale in large plants. MacDonald et al. found evidence of relatively small but statistically
 significant economies of scale: the largest facilities have a 3 to 5 percent per unit cost advantage over
 facilities with one-fourth the capacity.

         Although MacDonald et al. (2000) find similar trends in economies of scale in both cattle and hog
 slaughter, concentration in the beef slaughter industry is much more pronounced. The authors believe this is
 a result of the relative industry growth rates. Growth in demand for beef has been relatively flat, while the
 market for pork has grown faster. The higher growth rate has enabled older, smaller hog slaughter facilities
 to remain in business even as newer, larger, more efficient facilities have developed. In the beef industry,
          19 Using GIPSA statistics, about 43 percent of heifer and steer slaughter in 1979 was accounted for by
  boxed beef, and 47 percent of boxed beef was fabricated in plants that slaughtered more than 500,000 head per
  year  By 1998, almost 95 percent of heifer and steer slaughter in GlPSA-reporting plants was accounted for by
  boxed beef, and 88 percent of that boxed beef, was fabricated in plants that slaughtered more than 500,000 head per
  year.
                                                  2-61

-------
flat growth means that the larger, more efficient facilities have gained market share more rapidly, because
the less efficient firms have exited the market. Hence the faster rate of concentration in the beef industry
than in the pork industry.

        Ollinger et al.'s (1997) examination of trends in the entry/exit of facilities in beef, pork, and
poultry markets also tends to confirm the impact of demand growth in these markets. They found higher
entry rates for facilities in the poultry markets than in the red meat markets, especially hi recent years.. Just
as significantly, the exit rate of facilities was much lower in the poultry markets than in the red meat
markets; the higher rate of growth allows marginal facilities to remain in production. This slows the rate of
concentration. Also, the exit rate is, not surprisingly, higher for small facilities than for larger facilities.

        Plant-level factors were more important than market-level factors in plant exit decisions in a probit
analysis of plants that slaughtered cattle hi 1991 but no longer did in 1993 (Anderson et al., 1998). The age
of the plant, variety of animals slaughtered, and degree of downstream processing performed at the plant
were more relevant to the exit decision than the regional HHI. However, a plant was more likely to close if
it was already only a small player in its regional market. Regional supply and demand conditions, such as
wages, population, and income, had little effect on closure, indicating the national market for beef products.
The results could not discern whether plants were "forced out" or if normal competition was at work.
Either way, the authors concluded, the welfare losses from  industry consolidation are likely to be offset by
efficiency gains (Anderson et al., 1998).

        Azzam and Schroeter (1995) quantify the concentration/welfare trade-off. They showed that given
the elasticities and other parameters of the beef industry, a  50 percent consolidation of packing plants
would require only a 2.4 percent cost savings to be welfare neutral. They also estimate that the actual cost
savings from a 50 percent increase hi plant size would be about 4 percent. Paul (1999) concludes that
"Increasing concentration in the U.S. meat packing industry seems justifiably to have emerged from cost
economies, which appear hi turn to be primarily transmitted to suppliers and demanders of cattle and meat
products rather than generating excessive profits for the plants or firms. Thus, these cost economies and
resulting evidence of concentration seem better interpreted  in the context of social efficiency than
inefficiency" (p. 629).
                                                2-62

-------
       Findings of the GIPSA-sponsored detailed survey of pricing practices support Paul's argument that
the benefits of cost economies are passed on to suppliers and demanders (Texas Agricultural Market
Research Center, 1996). The survey found that the largest processors paid the highest prices for fed cattle
when the data were corrected for quality and uniformity. The largest packers operated the largest plants
and maintained the highest-capacity utilization ratios, presumably to take advantage of their economies of
scale in processing (Texas Agricultural Market Research Center, 1996).

       Hayenga (1997) has provided some anecdotal evidence of economies of scale. Double-shifting of
production lines in beef and pork slaughter plants has become much more prevalent in recent years.-
According to one industry expert interviewed by Hayenga, double-shifting a hog processing plant adds
approximately 20 percent to facility cost for additional cooler capacity and similar items, but expands the
volume of output by approximately 95 percent. This implies that per unit fixed costs in a double-shifted
plant are roughly 60 percent of those in a single-shift facility. Hayenga's interviewees also discussed the
importance of maintaining a relatively constant high utilization rate, thus providing some confirmation of
findings of both Paul (1999) and the Texas Agricultural Market Research Center (1996) discussed above.
        2.4.3   Industry Concentration and Market Power

        Industry concentration does necessarily lead to market power (firms selling at prices that differ
 from the competitive price in order to gain a profit).  If market power existed in the meat products
 industries, it could have substantial implications for the economic impact analysis. This section reviews the
 existing literature on market power in these industries.                         •
         Beef

         Beef packers' abuse of market power was one of the motivating forces for the Sherman Anti-Trust
 Act of 1890. New regulation and technological developments contributed to the breakup of the 19th century
 packers' trust, but once again in the 1990s a few packers have come to dominate the beef markets. A
 debate has developed over whether packers are extracting excess profits from their potential market power.
                                                2-63

-------
To detect the use of market power, economists look for indications that prices differ from those they expect
to see if markets are competitive. Most of this research has focused on lov/er than competitive cattle prices,
but some has considered the whole marketing chain.

        One method for detecting price deviations is the Structure-Conduct-Performance (SCP) approach.
SCP relates measures of market power to measures of market performance. If there are few firms and they
are exerting their market power, market prices should be lower than the competitive market equilibrium.
This empirical approach finds correlations between market-level variables. The SCP literature has
produced a robust empirical regularity of statistically significant negative correlations between buyer
concentration and prices of cattle and hogs; that is, when markets are more concentrated, prices are lower
(Azzam and Anderson, 1996). However, the SCP method offers no means to distinguish between market
power and other possible causes of a correlation, such as efficiencies of size. Without a theoretical
framework, a conclusion of market power is not warranted (Azzam and Anderson, 1996).

        The New Empirical Industrial Organization (NEIO) approach begins with microeconomic theory.
Under perfect competition, profit maximization requires that input price equal net value of marginal
product. Deviations from this equality are taken as evidence of market power. SCP involves only
 generalized relationships, but NEIO incorporates measures of the appropriate response to  other firms'
 decisions into a firm's optimal input choices. The degree of market power can be assessed through the
 deviation of observed prices from the net value of marginal product as predicted by an economic model
 (Rogers and Sexton; 1994; Muth and Wohlgenant, 1999). Incidental results can be used to.verify the
 models and support conclusions on market power. As with most applications of economic theory, '
 assumptions about the appropriate economic model and the firms' optimal choices influence NEIO results.

         Schroeter (1988) applied NEIO methods to beef packing in the years 1955 to 1983. He found both
 the cattle and beef markets significantly distorted, but with only small price effects: price  distortions
 amounted to 1 percent in the cattle market and 3 percent in the beef market: Increasing concentration in the
 industry during the 1970s had no observed effect on the price distortion. Schroeter attributed the lack of
 larger distortions to a threshold effect—a CR-4 might need to exceed 68 percent before substantial
 distortions would be observed. That level was exceeded in 1988.
                                                2-64

-------
        Many later authors applied less restrictive NEIO models to similar data and obtained similar
'results. NEIO studies often find a statistically significant but economically small deviation in prices from
 the competitive ideal. Stiegert et al. (1993) point out that 1 to 4 percent sounds modest, but may represent
 10 to 40 percent of the packers' marketing margin. Schroeter and Azzam (1990) found the most significant
 exercise of market power in the beef and pork markets. They claimed that 55 percent of the farm-retail
 price margin for beef and 47 percent of the margin for pork were attributable to market power. This result
 appears to be an outlier, as other studies found market power distortions to be from 1 to 3 percent of prices
 (Stiegert et al., 1993; Weliwita and Azzam, 1996). Azzam and Anderson (1996) note that all of the studies
 they reviewed vary in methods, data sources, geographic coverage, and temporal coverage and therefore are
 not additive in their conclusions. They also conclude "there is not a definitive analysis in the lot" (Azzam
 and Anderson, 1996, p. 110).

         NEIO studies have come under increasing criticism. Hunnicutt and Weninger (1999) cite three
 aspects of the NEIO approach that may lead to inaccurate assessments. First, most NEIO studies ignore the
 dynamic nature of oligopoly relationships. When there are few competitois, all of the firms involved are
  aware of each other's reputation, as are growers and regulators. Thus current behavior will reflect on
  future relationships. In this situation, the NEIO assumption of short-term profit maximization may not be
  reasonable. Second, NEIO approaches ignore other strategic variables that could be influencing a firm's
  price decision. As Hayenga (1997) and others observed, firms may be more interested in stable supplies
  than high current profits, and may accept prices accordingly. Finally, NEIO generally relies on empirical
  supply and demand elasticities. These are notoriously difficult to measure and so are unlikely to be an
  adequate reflection of market conditions.

          To relax some of the traditional NEIO assumptions, Muth and Wohlgenant (1999) implemented a
  flexible model of imperfect competition that did not rely on empirically estimated demand and supply
  elasticities and allowed the market power terms to vary over time. They found that the conclusion of
  whether imperfect competition was present was highly sensitive to the constraint that the market power
  terms were constant. Whenever they were constrained, the data supported imperfect competition. Whenever
  they were allowed to vary through time, the results indicated perfectly competitive markets. As many prior
   studies constrained these parameters, the conclusions of the earlier studies may have been merely artifacts
   of the constraint, not accurate observations of market power.
                                                 2-65

-------
       Clark and Reed (2000) implemented an even more flexible model with very little structure imposed
on the data before developing test statistics for competition and oligopsony. The changing structure of the
agricultural industries over the period they, considered, 1960-1997, was subsumed in the model. Clark and
Reed (2000) could not refute the competitive model for beef, dairy, eggs, or poultry.

       Just as SCP studies were methodologically challenged but lead to a robust observation, NEIO
studies consistently find a gap between the price of cattle and its net marginal value product. Azzam and
Anderson (1996) conclude that "the body of empirical evidence from both SCP and NEIO studies is not
persuasive enough to conclude that the industry (red meat packing) is not competitive" (p. 122). They note
that failure to show that the industry is not competitive is not evidence that it is competitive.

       More recent studies have taken different directions in response to the perceived weaknesses of the
SCP and NEIO approaches. Rogers and Sexton (1994) suggest that the nature of agricultural goods
(specialized, difficult to transport, perishable) encourages the exercise of market power in regional markets,
though it might not be detectable in aggregated national markets. Several studies have tried to define the
regional markets and detect whether market power is important in them. Bailey et al. (1995) used spatial
statistical techniques to identify eight  market areas for feeder- cattle. Though the market areas were large  •
(all of them included several states), irregular, and largely overlapping, Bailey et al. found that where only
one feeder market area  dominated a county, prices for feeder cattle were lower than elsewhere by $8 to $9
for a 700- to 800-pound steer. Where  feeding areas overlapped, sellers received a premium price.

        Koontz et al. (1993) have developed a non-cooperative game theoretic model of short run beef
packer behavior for four regions. They suggest that packers may switch between cooperative and non-
cooperative pricing regimes based on  observation of their own margin between boxed beef and fed cattle
prices. In econometric estimations using daily price data, the researchers found that the difference between
the cooperative and non-cooperative price was only 0.5 to 0.8 percent and that the difference varied over
time and across regional markets.

        Meat packers sell to a concentrated network of wholesalers and large retail food chains, which may
also have market power. The observed deviations from competitive  conditions  could originate at either
market level. Azzam (1992) used an approach in which the prices of inputs to the production process, along
                                               2-66

-------
with the price spread from farm, output to retail product, indicate the presence of monopsony or monopoly
power on the part of packers in either the farm or wholesale markets. The study found that packers
exercised monopsony power in the cattle market but did not have market power in the wholesale markets.
Schroeter et al. (2000) tested an econometric model for bilateral oligopoly power between meat packers and
retailers. The model encompassed the possibilities that packers were price takers, retail chains were price
takers, or both were price takers. The model that fit the data best implied that packers were price takers in
the face of retailers' market power. The authors point out that less complete models of the same data would
have indicated that all levels were price takers and rejected any market power.

        In summary, the literature on beef marketing indicates that beef packers exercise little market
power at the national level. There has been a consistent finding that packer concentration has resulted in
statistically  significant but economically small reductions in the prices received by farmers. What evidence
there is for deviation of national prices from the competitive level can also be explained by cost efficiencies
and methodological errors. In a region dominated by a single packer, the evidence indicates, that packer
may exercise some market power within a limited range.
        Pork
        Though it is often included under the rubric of "red meat" packing, pork processing has attracted
 less academic attention than beef processing when it comes to market power. Paarlberg and Haley (2000)
 state they are "unaware of any estimates of market power for the swine industry" (p. 6).

      "  While hog processing has not increased in concentration as rapidly as beef production, the industry
 has integrated vertically to a much greater extent as it has expanded into new production regions. It has
 been argued that hog production may follow the path of broiler production to large, vertically integrated
 producer/processor firms that control the whole production chain (Martinez, 1999). This arrangement can
 benefit both the integrator and the contract grower. The integrator gets an assured supply of the desired
 quality of input, ensuring a high rate of return on investments in plant and research. The contract grower
 gains an assured price and access to the latest genetic technology. In a  sense, the benefits to contract
                                                2-67

-------
growers come at the expense of independent growers, who face thinner, more volatile auction markets, and
cannot keep pace with current technology (Paarlberg et al., 1999).

        The findings of Hayenga (1997) may help explain the combination of high levels of concentration
found in the pork industry with the relatively insignificant degrees of market power. Hayenga found that
pork plant capacities are typically determined by the peak seasonal demand for pork products. However,
even during offpeak seasons, the plant runs most efficiently at high- rather than low-capacity utilization.
Thus, the pork industry suffers from overcapacity during the offpeak season.  Competition engendered by
industry overcapacity may prevent the largest firms from gaining market power.
        Poultry
        Analyses of the beef and pork industry has focused on the issue of market concentration and the
 potential for market power with respect to both farmers and consumers. This is probably a result of both
 academic and government concern with the industry dating back over 100 years; the behavior of the meat
 packing industry probably played a key role in the passage of the Sherman Act—the foundation for
 antitrust law in the United States—in 1890 (Azzam and Anderson, 1996). In contrast, concern about
 competition and market power in the "poultry industry is relatively recent. Analysis of the poultry industry
 has focused in particular on the nature of the production contracts between growers and processors. That
 focus is reflected hi the studies summarized below.

        Almost all broiler production is carried out under contract to large integrators (e.g., poultry
 slaughtering and processing firms or feed mills) that process chicken for retail sale. This arrangement
 evolved as farmers sought price stability and processors sought assured supplies of a standard quality.
 Technological and genetic improvements allowed both growers and processors to develop larger and more
 efficient facilities. Contract terms are not made public and there is no market-determined price for live
 broilers, so the methods used to discover market power in the beef sector are inappropriate. Several studies,
 however, make it clear that integrators wield market power over growers.
                                               2-68

-------
       Most poultry producers are compensated through a two-part piece rate tournament contract that
compares their production performance with that of the group of growers delivering birds to the processor
at about the same time (Tsoulouhas and Vukina, 2000). Farmers opt for this type of contract knowing that
they may receive a lower price but that their stream of income will be less volatile than if they faced a
fluctuating retail market price. Whether the farmer is made better or worse off by this type of contractual
arrangement depends on the farmer's risk preferences. Tsoulouhas and Vukina (1999) argue that the form
of the contract is a result of the volatility of prices, growth of the industry, and bankruptcy risk of
integrators. As integrators are large, risk-neutral corporations, unlikely to go bankrupt and national in
scope, they can accept all of the risk-averse growers' price risk and the growers need not be concerned
about integrator bankruptcy risk. The tournament contract protects the integrator from unproductive
growers by basing payments on performance relative to others. Using their market power, the integrators
are able to drive growers' payments down to the grower's average reservation utility (the point at which the
grower is indifferent between raising chickens and leaving the industry) and extract all rents from the farm
process.

        In contrast, Lewin-Solomons (1999) assumes that growers may still earn rent, but that an
integrator may require specific capital improvements adapted to its production system in order to commit
growers to sell exclusively to them. Required assets screen out low-ability growers and leverage
integrators' capital resources (Tsoulouhas and Vukina, 2000), but unreasonable asset specificity ties the
grower to one integrator in a franchise relationship.  With no alternative use for the specific assets, the
grower must continue to satisfy this integrator in order to recover its sunk costs (Lewin-Solomons, 1999).
Integrators are able to maintain the franchise relationship because there are relatively few competing outlets
for poultry producers.

        Bernard and Willett (1996) established that wholesale broiler prices "cause" farm and retail prices
in all regions of the United States: "concentration and power of the integrators have allowed the wholesale
price to become the center, causal price in the [broiler] market"  (p. 288). Downward movements in
wholesale prices are passed on more fully to broiler growers than upward movements, suggesting that
integrators use price decreases to adjust the contract arrangement. Consumers do not suffer greatly from
the asymmetry between upward and downward movements in retail prices. Only in the North Central
United States can integrators pass on a larger portion of price increases than decreases.
                                               2-69

-------
       Research has tended to focus on the evolving relationship between poultry processors and poultry
growers and its implications for market power in the industry. These studies leave little doubt that large
broiler integrators have, and use, market power in their relationships with growers and retailers. There is
less evidence that large poultry processors have significant market power in then: relationship with
consumers.
2.5     PROFILE OF INDUSTRY LEADERS

        Section 2.5 provides a profile of key industry leaders in the meat products industry. The
information contained in these profiles is based on a large number of publicly available information sources
such as industry trade journals, company internet sites, and company 10-K reports on file with the
Securities Exchange Commission.(SEC) EDGAR database.

        The information provided by these various sources is not always consistent. The information may
vary for a number of reasons. The most reliable source, the EDGAR database, often does not provide detail
on many meat product entities because they are frequently divisions or subsidiaries of much larger
companies. Privately-owned companies do not file 10-K reports with the SEC; much information on
privately-held companies had to be estimated. In addition, sources may often make undisclosed
assumptions concerning the processes, subsidiaries, divisions,  and brands included in their estimates. For
example, three different sources provide estimates of 11, 17, and 48 individual plants owned by Sara Lee.
 Sara Lee owns a large number of subsidiaries, some of which  perform slaughtering operations, others of
 which perform processing operations; the differences in these estimates are most likely due to which
 subsidiaries and processes were included. EPA used its judgement to reconcile differences between various
 sources. However, the uncertainty concerning the reliability of sources means that the figures cited in the
 tables and profiles below should be used with care. They should be considered no more than order of
 magnitude estimates that provide useful information about the relative size of various companies
 operations.

         Also, in the summary tables EPA included some entities that are subsidiaries, divisions, or even
 brands of a larger business entity. EPA presented the material in this manner for a. number of reasons.
                                               2-70

-------
First, many of the subsidiary and division names (e.g., Oscar Mayer, Butterball, Bryan Foods, John
Morrell) are widely used throughout the industry. EPA thought that knowledge of the industry would be
improved by using the "familiar name" and linking it to the parent entity through the use of footnotes.
Second, many of these subsidiaries and-divisions are significant entities; presenting information at this
level, when appropriate, provides additional detail of company operations (e.g., information is presented
separately about ConAgra's red meat operations under the ConAgra name, and poultry operations under
the Butterball name). Third, some subsidiaries apparently have considerable autonomy within the main
corporate structure (e.g., in 1995 John Morrell was acquired by Smithfield Foods; in 1998, John Morrell is
listed by GJPSA (2000) as the purchaser of Mohawk Packing).  .,                                  .

        Section 2.5 is organized as follows. Sections 2.5.1 and 2.5.2 present the largest beef and pork
 slaughtering operations in the U.S. ranked by 1999 slaughter. Section 2.5.3 profiles the largest poultry
 slaughtering operations ranked by 1999 liveweight slaughter. Finally, in order to place the different type of
 meat operations in perspective, Section 2.5.4 provides a listing of U.S. meat producers with 1999 revenues
 in excess of $250 million, regardless of type of meat produced or operations performed.

         It is important to note that most of the company information presented below is specific to the year
  1999.  This year is the base year of the analysis because it is the latest year for which financial data will be
  available from the Section 308 survey. As appropriate, mergers and other significant events that occurred
  after 1999 have been added to the profiles for completeness.
         2.5.1   Beef Slaughtering Operations

         Table 2-31 lists beef firms ranked by 1999 slaughter performed.  The revenues of the top four
  slaughterers demonstrates the high degree of concentration in the industry discussed in Section 2.4.1 above
  (although note that revenues reflect all operations, -not just beef slaughter operations). Revenues of the fifth
  largest operation in 1999, Packerland Packing, are roughly one third those of the fourth largest company,
  and one tenth those of the largest company.
                                                 2-71

-------
                                             Table 2-31
                                  Firms With Beef Slaughter Plants
                                     Ranked by 1999 Slaughter
Company
EBP1
ConAgra 2
Cargill Red Meat Group (Excel) 3
Farmland Refrigerated Foods
Packerland Packing
GFI America
Moyer Packing
American Foods Group
Emmpak Foods
Taylor Packing
Sam Kane Beef Processors
Washington Beef
PM Holdings
Harris Ranch Beef
Shamrock Beef Processors
Agri-Processors
Caviness Packing
Simplot Meat Products
Vienna Sausage Manufacturing
Abbyland Foods
Beef
Slaughter
Rank
1
' 2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Pork
Slaughter
Rank
1
2
3
5















17
1999 Fiscal
Year Sales
($ millions)
$14,100
$12,500
$9,000
$3,800
$1,300
$600
$560
$530
$490
$380
$320
$280
$250
$180
$130
$100
• $100
$100
$100
$90
1999
Employment
49,000
48,000
20,000
12,000
4,000
1,200
1,600
1,500
1,800
1,000
600
620
800
530
300
300
200
250
500
350
Plants
60
72
18
14
4
4
2
2
3
1
1
• 3
4
1
1
1
2
1
1
2
Source: Meat Marketing & Technology, 2000; Meat Processing, 2000; Meat&Poultry, 2000a.
1 In fiscal 2000 IBP acquired Corporate Brand Foods America, a processor of beef, pork, chicken, and turkey. '
2 ConAgra Poultry, a subsidiary of ConAgra Inc. is the 5* largest broiler processor. ConAgra also acquired
Seaboard Corporation's poultry division which is the 10th largest broiler company. ConAgra's turkey operations
are carried out under the name Butterball Turkey, the 2nd largest turkey producer in the U.S.
3 Cargill is also the 3ri largest turkey processor.
                                                2-72

-------
       Two things in particular are worth noting ill Table 2-3 li First, the rapid decline in relative size
from the largest firm to the 20th largest firm as measured by revenues, employment, and the number of
plants owned. Second, the four largest beef slaughter operations also rank among the top five hog
slaughter operations. With a single exception, Abbyland Food, all other companies appear to specialize in
beef slaughter. Only the very largest companies are involved in both types of livestock slaughter
operations, and they are heavily involved in both.

       'Brief profiles of the ten largest beef slaughter companies are provided below. It is worth noting
that four of the ten largest U.S. beef slaughter companies in 1999 were also among the top ten largest meat
and poultry companies in the world (Meat&Poultry, 2000c).
        Iowa Beef Processors, Inc. (IBP)

        In 1999, IBP, Inc. was the largest meat and poultry company in the U.S. and the world, with
revenues of $14.1 billion (Meat&Poultry, 2000a; Meat&Poultry, 2000c).  IBP was also the largest beef
packer and the largest pork packer in the U.S. (Hughes, 2000). IBP's operations are conducted by two
segments. The Fresh Meats segment produces fresh and boxed beef and pork, while the  Foodbrands
segment produces value-added food products.  The Foodbrands segment consists of three subsidiaries:
Foodbrands America, The Bruss Company, and IBP Foods, Inc.

        IBP owned a total of 45 meat plants in the U.S. in 1999, 13 of which were beef plants with a total
slaughter capacity of 38,800 head per day (Meat&Poultry, 2000a; Hughes, 2000).  IBP also owned ten
beef carcass production facilities, eight of which produced boxed beef.  In 1999, IBP's processing facilities
operated at 84 percent of production capacity.  The company also had one ground beef processing facility.

        EBP was among the most aggressive meat product companies in acquiring smaller operations; IBP
purchased at least 12 other meat product companies from 1995 to 1999 (GIPSA, 1997; GIPSA, 2000).
Among IBP's  acquisitions in 1999  are H&M Food Systems Company, Inc., Russer Foods, Wilton Foods,
and Thorn Apple Valley, Inc. IBP also acquired Corporate Brand Foods America in fiscal 2000.
Corporate Brand Foods was a processed meat company with 11 plants whose products included deli meats,
                                              2-73

-------
ground beef, and roast beef (IBP, 2000). ffiP's acquisitions reflect the company's active role in the
expansion of its case-ready and value-added meats segments. With one producing case-ready facility, IBP
was set to acquire two more in 2001 (Meat&Poultry, 2000b). The company was also planning to venture
into the cooked-beef products market by 2002.

       Since 1999, IBP has further expanded its case-ready and cooked meats sectors. The company
entered into a partnership with Cameco Holdings in early 2001 to share the operations of its ground beef
processing plant (Meat&Poultry, 2001a). EBP's fiscal 2000 sales were $17 billion, lower than ConAgra's,
making IBP the second largest meat and poultry company in the U.S. and the world (Meat&Poultry, 2001g
and2001j).

       In January 2001, IBP agreed to an acquisition by Tyson Foods, Inc. This merger, valued at $4.7
billion, was approved by ffiP stockholders and completed in September 2001 (IBP, 2001). Tyson Foods is
now believed to hold 28 percent of the U.S. beef market, 23 percent of the chicken market, and 18 percent
of the pork market (IBP, 2001). The sale of EBP is apparently a result of the industry trend towards
emphasis on higher value-added products such as case-ready meats.  IBP management felt its company was
undervalued by the stock market, which had failed to perceive its move away from commodity meat
production (WSJ, 2000). This motivated the managerial decisions that eventually resulted in the sale of
IBP to Tyson.
        ConAgra, Inc.

        ConAgra, Inc. was the second largest meat and poultry company in the U.S. and the world in 1999
 (Meat&Poultry, 2000c). A largely diversified food company, ConAgra's operations fall into three
 segments: Packaged Foods, Refrigerated Foods, and Agricultural Products. Con Agra's beef operations are
 conducted by ConAgra Beef Companies under the Refrigerated Foods segment. ConAgra Beef Companies
 included Armour Fresh Meats, E.A.Miller, Monfort, Northern States Beef, and Signature Ground Beef.
 Together, these companies made ConAgra Beef the second largest beef packer in the U.S. in 1999 (Hughes,
 2000). ConAgra's meat and poultry related sales for fiscal 1999 were $12.5 billion (Meat Processing,
 2000).
                                              2-74

-------
       ConAgra had a total of 83 meat and poultry plants in 1999 (Meat&Poultry, 2000a). Seven of
them were beef slaughtering plants with a combined daily capacity of 23,000 head (Hughes, 2000).
ConAgra's annual sales for beef were $6.7 billion in 1998 (Hughes, 2000). The company and its divisions
own several brands of cooked and refrigerated convenience meals.

       At the end of 1999 ConAgra announced a major restructuring process whereby the above
mentioned beef companies were integrated under one unit, called the ConAgra Beef Company (ConAgra,
2000). The ConAgra Beef company had projected sales of $5.8 billion in 1999 from its eight plants
(Meat&Poultry, 2000a). The company's restructuring plan also emphasized customer focus and value-
added products (Meat&Poultry, 2000b). As of 1999, among ConAgra's subsidiaries in other segments
identifiably involved in beef processing are Goodmark Foods, and Decker Food Company. ConAgra has
been aggressive in acquiring subsidiaries since 1995.

       ConAgra's post-1999 acquisitions includes Marburger Foods,  a bacon producer (Meat&Poultry,
2000d). The company also entered into a joint venture with Sigma Alimentos to market frozen foods in the
U.S., Canada, and Mexico (Meat&Poultry, 2001i). ConAgra's fiscal 2000 sales amounted to $20 billion
making it the largest meat and poultry company in the U.S. and the  world in 2001, a position enjoyed by  •
IBP in 2000 (Meat&Poultry, 2001g and 2001J).
        Cargill Red Meat Group (Excel Corporation)

        Excel is a wholly owned subsidiary of Cargill, Inc., an international marketer, processor and
 distributor of agricultural, food, and industrial products. In 1999 Forbes ranked Cargill, Inc. as the largest
 privately-owned company in the U.S. (Hoover's, 2000). Cargill was the third largest meat and poultry
 company in the U.S. and the world in 1999, with annual sales of $9 billion for meat and poultry related
 operations (Meat&Poultry, 2000a; Meat&Poultry, 2000c). Cargill's beef operations are carried out under
 the umbrella of Excel Corporation.

        Excel owned five beef plants and four beef and pork further processing plants in 1999 (Excel,
 2000). At the time; Excel's beef plants had a total daily capacity of 22,500 cattle, earning the company
                                              2-75

-------
$6.4 billion in beef sales annually (Hughes, 2000).  In addition, the company also had six branded value-
added lines and three case ready plants in the U.S. (Meat&Poultry, 2000b).

       CargiU's position as the third largest meat and poultry producer in the U.S. and the world in 1999
remains unchanged hi 2001 (Meat&Poultry, 200Ij). The company's 2000 sales equaled $10 billion
(Meat&Poultry, 200 Ig).  In an effort to expand its value-added lines, during the end of 2000 Excel formed
a joint venture with Advance Food Company (Meat&Poultry, 2000d). In 2001, Excel also announced plans
to acquire Emmpak Foods, Inc., a value-added meat producer with three processing plants and sales of
$570 million in fiscal 2000 (Meat&Poultry, 2001f). This acquisition will give Excel the capability of
producing 180 million pounds of cooked meat a year (Meat&Poultry, 20011). Furthermore, Excel will
acquire Taylor Packing in early 2002 (Meat&Poultry, 2001n).
        Farmland National Beef

        Farmland National Beef was the fourth largest beef processor in the U.S. in 1999 supplying 10
percent of the country's beef (Farmland, 2000b).  Farmland National Beef Company, L.P. is owned jointly
between Farmland Industries and U.S. Premium Beef, both agricultural cooperatives.  As of August 1999,
Farmland owned 71.2 percent of National Beef. Farmland Industries is an agricultural farm supply,
processing, and marketing cooperative. In 1999, Farmland was owned by 1,400 local cooperatives; its
membership was expected to grow even larger with its planned merger with Cenex Harvest States to create
United Country Brands.  Farmland National Beef earned total revenues of $3.8 billion in fiscal year 1999
(Meat&Poultry, 2000a).

        Farmland National Beef had two beef processing facilities with daily capacities of 9,000 head in
 1999 (Hughes, 2000). In 1999, these two facilities slaughtered an aggregate of 2.6 million cattle. Sales
from beef processing and marketing increased $223 million in 1999 compared to 1998.  Farmland National
Beefs processing capacities increased 50 percent from 1992 to 1999 (Farmland, 2000b). In addition,
Farmland was also involved in the production of branded case-ready beef and cooked beef meals
(Meat&Poultry, 2000b). Farmland Industries' members provided 38 percent of the beef cattle processed in
                                              2-76

-------
1999 while U.S. Premium Beef members also supplied cattle. Farmland also had a feed business segment
which provided cattle producer members with feed:

       Since 1999, Farmland's position has fallen from the fourth largest to the sixth largest meat and
poultry company in the U.S. with annual,sales of $4.4 billion in fiscal 2000 (Meat&Poultry, 2001g).  The
company has been expanding its case-ready operations and plans to open a third case-ready plant in 2002
(Meat&Poultry, 200 li).
        Packerland Packing Company, Inc.

        Packer-land was the fifth largest beef processing company in 1999. This privately held company
had four beef plants and fiscal 1999 sales of $1.3 billion (Meat Processing, 2000).  Packerland's daily
slaughter was about 5,200 head of cattle (Hoover's, ?,000). During the fall of 2001, Smithfield Foods
acquired Packerland at a price of $250 million (Meat&Poultry, 200Ij).
        GF1 America, Inc.

        GFI America was the sixth largest beef slaughter company in the U.S. with annual sales of $600
 million in 1999 (Meat Marketing & Technology, 2000).  A private company, GFI is owned and operated
 by its founding family. GFI's products include ground beef, cooked beef, value-added beef, and custom cut
 fresh beef.

        GFI owned four plants in 1999* two of which were slaughtering and rendering plants, while the
 other two were custom processing plants (GFI, 2000). GFI's vertically integrated beef operations also
 include a special procurement team to select and purchase cattle, and a strategically designed feeding
 program (GFI, 2000). Federal Beef Processors and GFI Premium Foods are among GFI's subsidiaries.
 GFI currently owns three facilities, including one which performs slaughtering, boning, and rendering
 operations, as well as two custom processing facilities (GFI, 2001).
                                               2-77

-------
       Moyer Packing                                                              .

       Moyer Packing Company, a beef processing and rendering business, was the seventh largest beef
slaughtering company in the U.S. in 1999 (Meat Marketing & Technology, 2000).  Moyer's annual sales in
1999 were $560 million (Meat&Poultry, 2000a).

       Moyer processed 1,850 cattle head per day in 1999 to produce 330 million pounds of boxed beef,
ground beef, and variety meats (Moyer, 2000). The company owned, in addition to its beef processing
plant, two rendering plants and one protein blending plant (Moyer, 2000). The company exported as much
as 20 percent of its annual production in 1999.

       Smithfield Foods acquired Moyer in 2001, thus entering the beef case-ready market
(Meat&Poultry, 2001d). Moyer's fiscal 2000 sales were $6 billion and this ninth largest beef processor
processed 360 million pounds of beef in fiscal 2000 (Meat&Poiiltry, 2001g ar.J. 2001i).
       American Foods Group

       American Foods Group is a processor of beef and pork, producing fresh and ground beef, among
other products. American Foods had annual sales of $530 million in 1999 (Meat&Poultry, 2000a). The
company owned two beef slaughtering and processing plants at the time, where it processed 1,800 cattle
per day (American Foods Group, 2000). Subsidiaries known to be involved in meat processing include:
Green Bay Dressed Beef, Huron Dressed Beef, and Dawson-Baker Packing Company. American Foods'
fiscal 2000 sales amounted to $580 million (Meat&Poultry, 2001 g).  Smithfield Foods announced plans to
purchase American Foods Group, but canceled the transaction in late 2001 (Meat&Poultry, 2001m).
        Emmpak Foods

        Emmpak Foods, Inc. is a meat processor with three subsidiaries, Emmber Foods, Peck Meat
 Packing, and Wis-Pak Foods (Meat&Poultry, 2000a). Emmpak's annual sales in 1999 were $490 million
                                             2-78

-------
(Meat&Poultry, 2000a). At the time, the company owned three beef plants with an estimated daily
slaughter capacity of 1,800 head in. In 1999 Emmpak announced an alliance with Titan Corporation,.
whereby Emmpak's beef will be pasteurized in Titan's irradiation facility (Salvage, 1999).  In 2001,
Emmpak was acquired by Excel Corporation, a subsidiary of Cargill, Inc. (Meat&Poultry, 2001f).
        Taylor Packing Company, Inc.

        Taylor Packing Co., Inc., with annual sales of $380 million in 1999, produces fresh and value-
added beef products (Meat&Poultry, 2000a). This company owned one processing plant at the tune,
capable of processing 1,900 cattle per day (Taylor, 2000).  In addition, the company's subsidiary, Taylor
By-Products, Inc., operates a rendering plant (Taylor, 2000).  Taylor Packing was the tenth largest beef
slaughter firm in the U.S. in 1999 (Meat Marketing & Technology, 2000). Since 1999, the company's
sales have grown to fiscal 2000 sales of $455 million. Cargill announced plans to purchase Tayloi Packing
in late 2001 (Meat&Poultry, 2000n).
        2.5.2   Hog Slaughtering Operations

        Table 2-32 lists the 20 largest entities in 1999 performing hog slaughtering operations in the U.S.
 As observed in Section 2.4.1 above, the concentration ratios in the hog slaughtering industry are lower than
 the in the beef industry; in Table 2-32, it can also be observed that the relative size of companies owning
 hog slaughtering facilities does not decline as rapidly as in the beef industry; Hormel Foods, ranked seventh
 in pork packing in 1999, had similar 1999 reve'nues as .the fourth ranked beef packer, Farmland. Also, note
 that the list of top twenty pork packers contains a number of firms much smaller than the list of top 20 beef
 packers. Finally, Table 2-32 again displays the tendency for firms to specialize in one meat type or the
 other; with one exception, only four out of the five largest pork packers are also significant beef packers.
                                               2-79

-------
                                             Table 2-32
                                 Firms With Pork Slaughter Plants
                                     Ranked by 1999 Slaughter
Company
IBP
ConAgra 2
Cargill Red Meat Group (Excel) 3
Sara Lee 4
Farmland Refrigerated Foods
Smithfield Foods 6
Hormel Foods-7
John Morrell 6
Seaboard Corporation 2'8
Bryan Foods 4
Indiana Packers
Clougherty Packing
Premium Standard Farms
Bob Evans Farms
Simeus Foods International
T.H. Routh Packing
Abbyland Foods
Sioux-Preme Packing
Cloverdale Foods
Leidy's
Pork
Slaughter
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Beef
Slaughter
Rank
1
2
3

4











20



1999 Fiscal
Year Sales
($ millions)
$14,100
$12,500
$9,000
• $4,100
$3,800
$3,800
$3,400
$1,600
$820
$640
$360
$320
$300
$260
$150
$120
$90
$70
$50
$40
1999
Employment
49,000
48,000
20,000
15,000 5
12,000
25,000
12,000
6,000
4,100
' 2,100
1,200
1,300
800
350
700
340
350
250
300
260
Plants
60
72
18
30s
14
31
12
8
1
2
1
2
1
6
2
1
2
2
2
1
Source: Meat Marketing & Technology, 2000; Meat Processing, 2000; Meat&Poultry, 2000a.
1 In fiscal 2000 IBP acquired Corporate Brand Foods America, a processor of beef, pork, chicken, and turkey.
2 ConAgra Poultry, a subsidiary of ConAgra Inc. is the 5* largest broiler company. ConAgra also acquired
Seaboard Corporation's poultry division which ranks as the 10th largest broiler producer.  ConAgra's turkey
operations are carried out under the name Butterball Turkey, the 2nd largest turkey processor.
3 Cargill is also the 3ri largest turkey producer in the U.S.
4 Sara Lee's turkey operations are carried out through its subsidiary Bil Mar Foods. Sara Lee's other subsidiary,
Bryan Foods, also a pork processor, is ranked 10th on this list.
sSara Lee employment and number of plants differed between sources by roughly 100 percent.
6 Smithfield Foods is also a producer of turkey. It's subsidiary, Carolina Turkeys, is the 5th largest turkey
processor. John Morrell, another subsidiary of Smithfield Foods, is the 8th largest pork slaughterer on this list.
7In addition to pork, Hormel also produces turkey. It's subsidiary Jennie-O Foods is the largest turkey slaughterer
in the country.
8 Seaboard Corporation's 1999 fiscal year sales and employment numbers do not include its poultry business
(Seaboard Farms) which was recently acquired by ConAgra.
                                                 2-80

-------
       Brief profiles of the 10 largest pork packers in 1999 follow.
       IBP, Inc.
       IBP had six pork carcass production facilities in 1999, which together had a daily slaughter
capacity of 69,500 hogs, and operated at 64 percent of their daily capacity (NPPC, 1999a). In addition,
IBP also had seven processing facilities.  At the time, IBP did not have facilities of its own to raise cattle or
hogs in the U.S. IBP's main supply of live cattle and hogs was purchased by IBP buyers trained to select
high quality animals that would.be candidates for higher yields.  In 1999, IBP completed its acquisitions of
Thorn Apple Valley, Inc., a further processor of pork and poultry with five processing facilities. For
further information on IBP see Section 2.5.1 above.
        ConAgra, Inc.

        ConAgra's Refrigerated Foods Division operates its fresh pork business through its subsidiary
Swift & Company. In 1987, ConAgra purchased Monfort and Swift Independent Packing Company which
were merged, and eventually renamed Swift & Company (Swift & Co., 2000). Swift owned three pork
processing plants in 1999 with a total daily slaughter capacity of 39,400 hogs (NPPC, 1999a). Swift also
operated three further processing plants. In 1998, Swift acquired Zoll Foods, a processor of custom pork
ribs and other pork products. For further information on ConAgra, see Section 2.5.1 above.
         Cargill Red Meat Group (Excel Corporation)

        Cargill's pork operations are also carried out under the umbrella of Excel Corporation, one of the
 two wholly owned subsidiaries of Cargill, Inc. Excel produces fresh, frozen, and processed pork products.
 Its three pork slaughter plants had a total daily capacity of 38,700 hogs in 1999 (NPPC, 1999a). In
 addition, Excel operated four beef and pork further processing plants and two case ready plants in 1999
                                               2-81

-------
(Excel, 2000). Due to private ownership and Cargill's extensive operations, little information was readily
available regarding its pork operations. For more information on Cargill, see Section 2.5.1 above.
       Sara Lee
        Sara Lee Corporation is engaged in pork and poultry slaughter, processing, and further processing
(Meat Processing, 1999).  Sara Lee's meat and poultry operations are conducted under the Sara Lee
Packaged Meats segment, which had revenues of $4.1 billion in fiscal 1999 (Meat&Poultry, 2000a).  Sara
Lee Packaged Meats ranked as the fifth largest meat and poultry company in the U.S. and the seventh
largest meat and poultry company in the world in 1999 (Meat&Poultry, 2000a and 2000c). Among Sara
Lee's subsidiaries engaged in pork slaughter and processing are Bryan Foods, Inc., Hillshire Farm &
Kahn's, and Jimmy Dean Foods.  Bryan Foods ranked as the tenth largest pork slaughterer in the U.S. in
 1Q99. Sara Lee is also involved in turkey processing through its Bil Mar Foods subsidiary,

        In 1999, Sara Lee owned two slaughter facilities for pork. .These facilities had a daily slaughter
capacity of 9,000 hogs (NPPC, 1999a).  The company completed the construction of another pork
processing facility in 1999. .SaraLee's involvement in the value-added meat segment can be illustrated
through the example of Hillshire Farm, one of the above mentioned subsidiaries. In 1997, Hillshire was
 actively involved in the production of gourmet sausage convenience meals, trying  to gain a niche for
 sausage products in the home meal replacement market (Nunes,  1997).

        Sara Lee's fiscal 2000 sales were no different from the 1999 sales and the company lost its rank as
 the fifth largest meat and poultry company.  It is now the seventh largest meat and poultry company in the
 U.S., as well as the world (Meat&Poultry, 200 Ig and 2001J). Sara Lee and its brands are believed to have
 the largest share of the hot dog, smoked sausage, breakfast sausage, breakfast sandwich, cocktail sausage,
 and com dog markets (Meat&Poultry, 2001i). Due to Sara Lee's extensive operations, further information
 through its 10-K, annual report, or website about its pork production and processing operations was not
 available.
                                               2-82

-------
       Farmland Foods, Inc.                                                           ,  .

       Farmland Foods, Inc. is a 99 percent owned subsidiary of Farmland Industries, Inc. Farmland is a
processor of both beef and pork.  Farmland's sales from pork processing and marketing decreased $130.4
million in 1999 compared with 1998.  In 1999, Farmland Foods, Inc. operated 11 processing facilities
across the country.  The company is at least partially vertically integrated, producing swine through
contract growers. In addition, the Livestock Production Group is another business segment of Farmland
producing market hogs for processing. Farmland is also involved in the production of case-ready pork
products.

       Farmland Industries formed Triumph Pork Group, LLC in 1999. Triumph was a joint venture
with The Hanor Company and Pork Technologies L.C.  Triumph provided Farmland's pork producers with
customized genetic lines, safety and environmental welfare programs, and brand alignment (Farmland,
2000a)  More information on Farmland can be found in Section 2.5.1 above.
       Smithfield Foods, Inc.

       Smithfield Foods, Inc. was the nation's largest vertically integrated hog-grower and pork processor
and the world's ninth largest meat and poultry company in 1999 (Meat&Poultry, 2000c). Smithfield
conducts its business through the Hog Production Group and the Meat Processing Group, including various
subsidiaries under each segment. The company produced 2.4 billion pounds of fresh pork and 16 billion
pounds of processed meat products in the U.S. in fiscal 2000. Smithfield's revenues for fiscal 1999 were
$3.8 billion (Meat Marketing & Technology, 2000).

       In 1999, the Meat Processing Group consisted of six domestic pork producing subsidiaries
including: John Morrell and Company, Smithfield Packing Company, Inc., Gwaltney of Smithfield, Ltd.,
Lykes Meat Group, Inc., Patrick Cudahy, Inc., and North Side Foods, Corp.  Along with IBP and
ConAgra, Smithfield has been the most aggressive U.S. meat packer at acquiring new  firms since 1995.
John Morrell was the largest of these subsidiaries in 1999, and ranked as the eighth largest pork slaughterer
hi the U.S.  John Morrell owned eight meat plants, and had fiscal 1999 sales of $1.6 billion (Meat
                                              2-83

-------
Processing, 2000).  Collectively, the above subsidiaries and four foreign subsidiaries operated 48
slaughtering and further processing plants in 1999. The five slaughter plants in the U.S. had an aggregate
daily slaughter capacity of 78,300 hogs. Smithfield has been increasing volumes of case-ready pork
products and opened four new case-ready facilities in 2000. Together with John Morrell, Smithfield
Packing was expected to produce 75 million pounds of case ready products in fiscal 2001 (Meat&Poultry,
2000b). The Meat Processing Group purchased approximately 50 percent of its live hog requirements
from the Hog Production Group in 1999.                       -

        Since 1999, Smithfield has continued its aggressive acquisitions of companies.  In 2001, the   .
company stepped into the beef processing sector acquiring two companies: Moyer Packing Company and
Packerland Packing Company (Meat&Poultry, 2001d and 20011). The company also expanded its case-
ready sectors by acquiring a stake in Pinnacle Foods, Inc. (Meat&Poultry,.2000g).

        Smithfield's pork related acquisitions include Gorges/Quik-to-Fix Foods (a producer of value-
 added beef, pork, and poultry products for $34 million), Stadlers Country Hams, Inc. (a processor of pre-
 cooked beef and pork entrees), The Smithfield Companies (a producer of ham, previously unrelated to the
 company), and PvMH Foods (Meat&Poultry, 2001h and 2001k). Combined, all these acquisitions are
 likely to make Smithfield one of the largest meat and poultry companies in the future. It is currently the
 sixth largest meat and poultry company in the world (Meat&Poultry, 2001k). Its fiscal 2000 sales were
 $5.1 billion (Meat&Poultry, 2001g).
         Hormel Foods Corporation

         Hormel Foods Corporation and its subsidiary, Rochelle Foods, Inc., are involved in both the
  processing of fresh meat and the manufacture of branded consumer products. Hormel also produces turkey
  products under the Jennie-O name. With revenues of $3.4 billion in 1999 for all operations, Hormel ranked
  as the tenth largest meat and poultry company in the world (Meat&Poultry, 2000c).

         The company owned three hog slaughter plants with a total daily slaughter capacity of 31,600 in
  1999 (NPPC, 1999a). One of these plants was leased to Quality Pork Processors of Dallas, Texas. In
                                                2-84

-------
addition, Hormel owned eleven processing plants for the production of manufactured food items. The
company had already moved into the case-ready segment by 1999 and planned to diversify into the cooked
meals and ethnic foods markets as well (Meat&Poultry, 2000b). Hormel's 2000 sales equaled $3.7 billion
and it ranked as the eighth largest meat and poultry company in the U.S. (Meat&Poultry, 2001g). The
company acquired The Turkey Store in 2001 (see Section 2.5.3 for more detail).
        John Morrell, Inc.

        John Morrell is a subsidiary of Smithfield foods and its pork slaughter operations are discussed
under that name.

        Seaboard Corporation

        Seaboard Corporation is a diversified international agribusiness and transportation company.  As
part of its primary domestic operations, the company produces and processes pork and poultry. Early in
2000 Seaboard sold its poultry operations to ConAgra and started to expand its vertically integrated pork-
segment.  Seaboard's pork revenues in fiscal 1999 were $820 million (Meat Marketing & Technology,
2000).

        Seaboard owned a hog processing plant with double-shift capacity of approximately four million
hogs in 1999. At the time, Seaboard was planning the construction of a second integrated pork operation
with a capacity to process over four million hogs annually. Seaboard's fiscal 2000 sales were $725
million, down from $1 billion in 1999 (Meat&Poultry, 2001g).
         Bryan Foods, Inc.

        Bryan Foods, the tenth largest hog slaughter operation in the U.S. in 1999, is a subsidiary of Sara
 Lee and its operations are discussed under that name.
                                               2-85

-------
       2.5.3   Poultry Slaughtering Operations

       Tables 2-33 through 2-35 present summary information on the largest poultry slaughter companies
in the U.S. in 1999. Table 2-33 lists the 25 largest broiler companies ranked by estimated annual
liveweight slaughter. Table 2-34 summarizes the 20 largest turkey slaughter operations, again, ranked by
annual liveweight slaughter.20 Finally, Table 2-35 combines the information in Tables 2-33 and 2-34 to
provide a ranking of the 30 largest poultry slaughter entities in 1999. The purpose of Table 2-35 is to
provide a sense of the relative size of broiler operations to turkey operations.

       Table 2-33 shows that Tyson Foods clearly dominated the industry in 1999, processing 2.6 times
more broilers by weight than the second largest company; Tyson alone accounted for 24 percent of 1999
industry broiler slaughter by weight. Due to incomplete data, exact concentration ratios cannot be
calculated from this data.  However, the percentage of live animal slaughter accounted for by the largest
companies is highly suggestive of the degree of concentration in this industry. The four largest broiler
companies, Tyson, Gold Kist, Perdue, and Pilgrim's Pride, slaughtered an estimated 20.1 billion pounds of
broilers in 1999,47 percent of the 46.2 billion pound industry total. Adding the next four largest
companies, ConAgra Poultry, Wayne Farms, Sanderson Farms, and Cagle's, to the total means that the
eight largest broiler companies in the U.S. produced 63 percent (26.7 billion pounds liveweight slaughter)
of the national production in 1999.

        In the turkey sector, the industry is  not as dominated by a single firm as the broiler sector (Table 2-
34).  The largest turkey producer in 1999, Jennie-O Foods (a wholly owned subsidiary of Hormel),
accounted for 13 percent of the U.S. total (860 million pounds of turkeys out of 6.7 billion pounds, live
slaughter weight). Production by the top four and eight turkey processors, however, is roughly as
concentrated as the broiler industry. The four largest turkey producers in 1999, Jennie-O, Butterball
Turkey (a subsidiary of ConAgra), Cargill,  and Wampler Foods, produced 44 percent of U.S. turkey (2.9
        20 Sources for this memorandum cite the average number of birds slaughtered weekly, and average bird
 weight. Thus, EPA estimated annual slaughter by multiplying the average number of birds slaughtered weekly Dy
 the average weight, then multiplied by 52.  Slaughter was converted to an estimated annual rate in order to
 facilitate a comparison between broiler operations and turkey operations.  Turkey slaughter data was akeady
 expressed in pounds of annual liveweight slaughter.
                                                2-86

-------
                                          Table 2-33
                              Firms with Broiler Slaughter Plants
                                  Ranked by 1999 Slaughter
                          Rank
    Broiler
  Slaughter1
   (millions
  of pounds)
                                          1999 FY
                                            Sales
                                         ($ millions)
                                                           1999
                                                       Employment
                                                                      Primary
                                                                         Processing Plants
Further
Total
Tyson Foods
 1
                                  10,338
                                                $7,400
                                                              65,000
                                             42
                                                                                   14
                                                                                             56
Gold Kist
       3,963
                                                $1,800
                                                            18,000
                                              12
                                                                                             14
Perdue Farms3
       3,200
                                              $2,500
                                                              19,000
                                                                            14
                                                       19
                                                                                           33
Pilgrim's Pride
                                      2.595
                  $1,400
                                                            15,000
                                                                                              15
ConAgra Poultry 4
        2,420
                                                            10,800
Wayne Farms5
                                      1,582
                    $830
                                                             9,100
                                                                                              12
(Sanderson Farms
        1,372
                                                  $560
                                                             7,700
 Cagle's
        1,268
                                                $310
                                                               7,000
 Foster Farms 6
        1,099
                                              $1,100
                                                               8,900
                                                                                              10
 Seaboard Farms
10
                                    1,006
                                                   $460
                                                               5,000
 Townsends
11
                                        955
                                                $520
                                                               4,400
 Fieldale Farms
12
                                        889
                                                $450
                                                               4,800
                                                                                               4
tempi
       er Foods
13
                                        881
                                                $890
                                                               7,100
 O.K. Foods
14
                                        836
                                          $250 - $499
                                                               4,300
 Allen Family Foods
15
                                      713
                                                   $300
                                                               2,40(
 Mountaire Farms
16
                                        701
                                                 $300
                                                               2,900
 Choctaw Maid Farms
17
                                        667
                                                 $250
                                                               3,20(
 Peco poods
18
                                        658
                                                 $300
                                                               3,90(
  Simmons Foods
 19
                                        622
                                                 $420
                                                               4,3(K
 Case Foods
20
                                        508
                                                 $200
                                                               2,00(
 George's
21
                                        471
 Marshall Durbin
                                        464
                     $200
                                                              1,80
 B.C. Rogers Poultry
                                        456
                     $330
                                                              3,40
 House of Raeford Farms 8
          451
                                                 $480
                                                                5,00
 IlKoch Foods
 25
                                      401
                                                   $530
                                  4,40
 Source: Meat&Poultry, 2000a; Meat Processing, 2000; Thornton, 2000a; Thornton, 2000b; Thorntpn, 2000c.
 11999 average weekly estimated slaughter x average slaughter weight x 52.
 2 For companies producing both broilers and turkeys, plants estimated to adjust for double-counting.
 3 Perdue Farms is also the 12th largest turkey producer.
 4 ConAgra Poultry is a subsidiary of ConAgra, Inc. The company also recently acquired Seaboard Farms,
 (Seaboard Corporation' poultry division) ranked 10th on this list. Seaboard Farms' 1999 fiscal year sales and
 employment numbers do not include Seaboard Corporation's pork business.  ConAgra is also engaged in turkey
 slaughter through  Butterball Turkey. This division of ConAgra is the 2nd largest turkey producer. ConAgra is
 also the 2nd largest pork processor and 2nd largest beef processor in the U.S.
 5 Wayne Farms is a division of ContiGroup Companies. ContiGroup also produces beef, but does not slaughter or
 process it; ContiGroup does process pork.
 6 Foster Farms is also the 13th largest turkey processor.
 7 Wampler Foods, a subsidiary of WLR Foods, Inc. is also the 4th largest turkey processor:
 8 House of Raeford is also the 10th largest turkey producer in the country.
                                               2-87

-------
                                                   Table 2-34
                                    Firms With Turkey Slaughter Plants
                                          Ranked by 1999 Slaughter
Company
Jennie-O Foods *
Butterball Turkey 3
Cargill4
Wampler Foods 5
Carolina Turkeys 6
Rocco Enterprises 7
The Turkey Store
Louis Rich Brand 8
BilMar9
House of Raeford Farms 10
Willowbrook Foods
Perdue Farms u
Foster Farms I2
Norbest
Farbest
Zacky Foods 13
Cooper Farms
West Liberty Foods
Iowa Turkey Products
Empire Kosher Poultry 14
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Turkey
Slaughter
(millions
of pounds)
859
790
715
579
460
427
375
350
260
245
227
224
173
150
146
144
143
138
85
50
1999 FY
Sales
($ millions)



$890
$350
$550
$250 - $499


$480

$2,500
$1,100
$145

$330
$150



1999
Employment



7,100
2,300
3,600
2,700


5,000

19,000
8,900
1,300

3,000
850


1,100
Processing Plants x
Primary
4
4
4
7
1
3
2
1
1
4
2
14
5
2
1
2
1
1
1
. 1
Further
4


1

1


1
4
1
19
5
4

1
1


1
Total
8
4
4
8
1
4
2
1
2
8
3
33
10
6
1
3
2
1
1
2
Source: Heffeman, 2000; Meat&Poultry, 2000a; Meat Processing, 2000.
1 For companies producing both broilers and turkeys, plants estimated to adjust for double-counting.
2 Jcnnie-O is a subsidiary of Hormel Foods. Hormel Foods also produces pork and is the 7th largest pork slaughter company.
3 Butterball Turkey is a division of ConAgra, Inc. The parent company is also engaged in broiler processing. Its subsidiary, ConAgra
Poultry, is the 5* largest broiler producer.  ConAgra also recently acquired Seaboard Farms' poultry division. Seaboard is the 10th
largest broiler company. ConAgra is also the 2nd largest pork processor and 2"d largest beef processor in the U.S.

4 Cargill also produces beef and pork through its subsidiary Excel Corporation.  Cargill is the 3rd largest beef processor and 5th largest
pork processor in the U.S.
5 Wampler Foods, a subsidiary of WLR Foods, Inc. is also the 13th largest broiler processor.
'Smithfield Foods, the parent company of Carolina Turkeys also produces pork. Smithfield is the 6th largest pork producer and its
subsidiary John Morrell, also engaged in pork slaughter, is the 8* largest pork producer.
7 Rocco Enterprises' turkey production is carried out through its subsidiary, Shady Brook Farms. Rocco is also the 30th largest broiler
producer in the U.S.
'Louis Rich Brand is a brand name of Kraft Foods, a subsidiary of Philip Morris Companies, Inc.
' Bil Mar Foods is a subsidiary of Sara Lee Corporation. Sara Lee is also the 4th largest pork producer, and its subsidiary, Bryan
Foods, is the 10* largest pork producer.
10 House of Raeford is also the 24* largest broiler company in the U.S.
11 Perdue Farms is also the 3rt largest broiler company.
n In addition to turkey Foster Farms also producers broilers and is ranked as the 9th largest broiler producer in the country.
15 Zacky Farms also processes broilers, beef, and pork.
14 Empire Kosher Poultry is also engaged in broiler slaughter and processing.
                                                       2-88

-------
billion pounds, live slaughter weight).  The next four largest producers, Carolina Turkeys, Shady Brook
Farms, The Turkey Store, and Louis Rich Brand (Kraft Foods), added 1.6 billion pounds to the total; thus
the eight largest turkey producers accounted for 68 percent of U.S. production by liveweight slaughter in
1999. For both broiler and turkey processing, the concentration ratio estimated from this data are quite
consistent with those cited in Section 2.4.1 above.

        Also, note that turkey operations are much more likely to be subsidiaries or divisions of larger meat
product firms than any of the other types of meat slaughtering operations examined in this profile.  Of the
 10 largest turkey slaughterers in 1999, only two, The Turkey Store (ranked seventh among turkey
 operations), and House of Raeford (ranked tenth) were independent companies at the time.21

        Table 2-35 provides a comparison of the size of turkey operations relative to broiler operations as
 of 1999. No turkey slaughterer ranks among the 10'largest poultry operations; only the three largest turkey
 slaughterers rank among the 20 largest poultry slaughterers. Thus, turkey operations are, in general, much
 smaller than broiler operations. It should be remembered, however, that turkey demand is much more
 seasonal than broiler demand, thus the peak capacity of a turkey slaughter plant may be closer to that of a
 broiler plant than indicated by this comparison.

         Finally, note that, as in the case of both beef and pork  slaughter operations, business entities tend
 to specialize in either broiler or turkey production, but not both.  Of the 30 largest poultry companies listed
 in Table 2-35, only six produced both broilers and turkeys in 1999.  Perdue Farms, Wampler Foods, and
 Foster Farms are the only top ten broiler companies that also produced turkeys, and of those three, only  for
 Wampler Foods was a  large percentage of overall output attributable to turkey operations in 1999.
         21 Since then, The Turkey Store has been acquired by Hormel Foods (Meat&Poultry, 2001b).
                                                 2-89

-------
I,
" a>
Ul -M


ft
  "



N
42

S
W
fl
1



1
1 f-
s
-.85
ON
TH

1
1
• (millions of
i
. 0
53












1
1

_
o
H


O)
f
I1
•s

1-1 "5
"3 'w
•3.1
11

>ซ^;



ซ
3
Is
o>
c
t.
P5

"
• 0
C

>
I
I

-fi





3
V)



Tfr


CN
s
n
vo
=


OO
o


OO
en
f
O


,-H






*•






r— <



CO


CN
t~~*
00

oc


i



i
e'


ex






w




r

en



ON



R
ON

I


i


e>
o
c
en


en





en
ft






^



vo


ON
0
In

1


ON
V



\f
CN


^






C.

•


p.

*""*



r-


ON






C



c
CN


VI



-<


r
CN
^"*



^~


OO
o
ON

X


X



CN
OO
V
"""*


VO





in
t




H
OO




I— (



^
(C

ON
X


VC


ON
o
ot
OO



r-





"S
t




*





^



^
P-T

vo


CN
r-
en



CN
e^
• '


OO





i
P

Tl


cy
O
^



VI



ง
OO

^


CN
o


en
ON
c
1 '


ON






I
tl

'



OO




en


VI
^
t~:

=^


OO
vo
o



OO
VC
'


o






W3

>i


u
vo




CN



(T5
VI

VO
n-






VO
^


<-<



CO


(




C/l
V)




^



o
^f

CN
•f.


V)
ON



VI
ON




-------


M
•S
a
3
HH
'I
*


, 1
•ฃ•
^•K
^•T*
T-


'I
O
eol
1
fS
t*
i
"io
s
ฃ
CC




-






"i
H


I*
.2
1
-CS
1
ON S
ON g>
ซ 1
"ซ*.2
1X2 S
^ '3
• ซs> S
V
,
^^
1


f
u
H
ป-<
2
j.
1
ซ

^
c
.9
M

c


cn



j



CN
CN
cn
o




r-~
VO
VO






vo
VC



^
c-



c/
|
c
	 " •
Choctaw Maid F
VO



-sj




S
ON
en
o
O
en
•E/3-




OO
^S






OO
VO



en
CN






VI
1
1
f-



_!.



cn
o
o
en

S
•&0-




CN
\c






c<
^D



^
CN






Simmons Foods
cn







en
o
o
o
CN
O




OO
O







OO
If



ID
CN






1
a.
a
en



T— (



CN
O
cn
o
en
cn
•&0-




ON
ON
^^


2



ป0
c^



^o
c^






V
C!
N
cn



!











1 — 1
f-
^f






r-
^t"



^
CN






George's
CN
5 "






CN
OO
f— f
o
m




s
••^






s
"*



oc







Marshall Durbin
"*







ป — i
cn
CN"
,o
cn




o
*^"


o
"*







ON
CN




^
rfi
Carolina Turkeys
^



en




O
cn
o
tr
•&O




Jo
^t*






vo
^



0





ฃ

PB.C. Rogers Pou
Ch





-------
Below are brief profiles of the 10 largest U.S. broiler producers, and the five largest U.S. turkey producers
in 1999.

       2.5.3.1 Broiler Companies

       Tyson Foods, Inc.

       Tyson Foods, Inc. was the nation's largest producer of broiler chickens in 1999. The company
was also the nation's largest poultry-based food company and the world's fifth largest meat and poultry
company (Meat&Poultry, 2000c). Tyson Foods is also involved in hog production and processing. A fully
integrated company, Tyson breeds, rears, feeds,, processes, further processes, markets, and distributes its
value enhanced chicken products. Company revenues for the fiscal year 1999 ending in September were
$7.4 billion (Thornton, 2000b).

        The principal poultry operations of the company consisted of 56 processing plants in 1999,
involved with various phases of slaughtering, dressing, cutting, packaging, deboning or further processing.
Together, these plants had a capacity of 47.6 million head per week. The average weekly production of
ready-to-cook chicken in 1999 was 154.3 million pounds (Thornton, 2000b).  Tyson completed several
 plant expansions in 1999 and planned to expand operations at two processing plants in 2000 (Thornton,
 2000b).  Tyson also began focusing on its value-added line of products and market testing convenience
 chicken products (Meat&Poultry, 2000b).

        Tyson's acquisition of the beef and pork giant IBP took place in 2001 making it the largest protein
 provider in the world (IBP, 2001).  The company also expanded its international operations in China,
 Mexico, and Central America (Meat&Poultry, 2001e).  Tyson is now the fourth largest meat and poultry
 company in the U.S. with sales of $7^2 billion in fiscal 2000 (Meat&Poultry, 200 Ig). It is still the fifth
 largest meat and poultry company in the world (Meat&Poultry, 2001k).
                                              '2-92

-------
       Gold Kist, Inc..

       Gold Kist, Inc., the second largest poultry processor in 1999, is a diversified agricultural
cooperative broken into two segments: Poultry and Agri-Services.  The Poultry segment maintains broiler,
pullet, and breeder flocks, and operates hatcheries, feed mills, and processing plants. Broiler production in
1999 was 14.8 million head per week (Thornton, 2000a). Gold Kist's sales in fiscal 1999 amounted to
$1.8 billion (Meat Processing, 2000).

        Gold Kist's integrated facilities included twelve processing plants, two further processing plants,
and three rendering plants in 1999. Gold Kist also operated nineteen hatcheries, twelve feed mills, ten
distribution plants, and six wastewater treatment plants (Thornton, 2000b).  In 1997, Gold Kist acquired
Golden Poultry with four integrated complexes, and Carolina Golden, with one complex (Meat&Poultry,
1999).

        Perdue Farms, Inc.

        The third largest broiler company in the U.S. in 1999, Perdue Farms, Inc. is a vertically integrated
agribusiness producing chicken, turkey, and grain. This company produced 47.8 million pounds of ready-
to-cook chicken weekly in 1999 (Thornton, 2000b).  Perdue's revenues for the fiscal year 1999 were $2.5
billion (Thornton, 2000b).

        Perdue's integrated operations included thirteen processing plants, nineteen further processing
plants, and three rendering plants in 1999.  In addition, the company also owned eighteen hatcheries, eleven
 feed mills, and four distribution centers. In 1998, Perdue acquired Gol-Pak Corporation, a producer of
 value-added chicken specialties, and Advantage Foods, a breast deboning operation (Perdue, 2000).
 Perdue's new products as of 1999 included precooked and cooked chicken meals.

         The company's fiscal 2000 sales were slightly more than $2.5 billion and Perdue is currently the
 ninth largest meat and poultry company in the U.S. (Meat&Poultry, 2001g). As of 2001, the company had
 a total of 21 processing plants in the U.S. producing 50 million pounds of poultry on a weekly basis
 (Meat&Poultry, 2001i).
                                                2-93

-------
       Pilgrim's Pride Corporation

       Pilgrim's Pride Corporation is a vertically integrated company producing fresh and frozen chicken.
The company's operations include hatcheries, grow-out farms, feed mills, processing and further
processing plants, distribution centers, rendering plants, and wastewater treatment plants.  Revenues for
1999 were $1.4 billion (Thornton, 2000b).

       The company's six processing plants had the capacity to produce 8.2 million head of chicken per
week as of 1999. Pilgrim's Pride also had three prepared foods plants, one of them purchased in 1998 .
from Plantation Foods, Inc. These prepared foods plants, located in Texas, operated two shifts in a six day
week. In 1999 the company produced 38.2 million pounds  of ready-to-cook chicken per week (Thornton,
2000b).

        The company's fiscal 2000 sales were $1.5 billion (Meat&Poultry, 2001g). In early 2001,
Pilgrims Pride acquired WLR Food, Inc., which owns Wampler Farms, a major turkey producer, for a total
of $280 million (Meat&Poultry, 2001g; Pilgrims Pride, 2001).

        ConAgra Poultry Companies

        ConAgra Poultry Companies ranked as the fifth  largest broiler processor in the country in 1999.
ConAgra is a large diversified company, whose operations  fall into three segments: Packaged Foods,
Refrigerated Foods, and Agricultural Products. ConAgra Poultry Companies include ConAgra Broiler
Company and ConAgra Frozen Foods which together produced approximately 34.9 pounds of chicken per
week in 1999 (Thornton 2000b).

        ConAgra Poultry Companies operated nine processing plants, seven further processing plants, and
four rendering plants in 1999.  Moreover, the company had ten hatcheries, nine feed mills, nineteen
distribution centers, and five wastewater treatment plants (Thornton, 2000b). ConAgra Poultry's fiscal
 1999 sales were $1.4 billion.
                                              2-94

-------
       In January 2000 ConAgra, Inc. acquired Seaboard Farms, Seaboard Corporation's poultry
division; this acquisition was expected to make CoriAgra the third largest broiler processor in the U.S. in
2000. In addition, ConAgra operates Butferball Turkey as a division specializing in turkey production
(second largest turkey producer in 1999), and was the second largest pork and beef processor in the U.S.
in 1999.  For more information on ConAgra see Section 2.5.1.
        Wayne Farms LLC

        Wayne Farms LLC is a division of ContiGroup Companies (CGC). CGC is a largely diversified
entity and was one of the leading poultry and pork processors in the U.S. in 1999 (Hoover's, 2000).
Wayne Farms was the sixth largest broiler company and produced 25.3 million pounds of ready-to-cook
chicken in 1999 (Thornton, 2000b).

        Wayne Farms' operations at the time included eight processing plants and four further processing
plants. The company has one subsidiary, Southland Foods, which is a poultry processing facility
(Meat&Poultry, 2000a).  Wayne Farm's complexes also included eight hatcheries and seven feed mills.  In
1999, the company slaughtered 4.74 million birds per week and had revenues amounting to $830 million
(Thornton, 2000b).

         Sanderson Farms, Inc.

        Sanderson Farms, Inc. is a fully integrated poultry processing company. The company produces,
processes, markets, and  distributes fresh and frozen chicken products. Sanderson Farms, Inc. has three
divisions: Production, Processing, and Foods. The Production Division produces broilers, while the
Processing Division processes, sells, and distributes the product. In addition, the Foods Division processes,
markets, and distributes  prepared food items. The company's sales topped $560 million in 1999, producing
almost 5.0 million head  per week (Thornton, 2000b). The company owned six processing plants, a further
processing plant, a rendering plant, and five wastewater treatment plants in 1999 (Thornton, 2000b).
                                               2-95

-------
        Cagle's, Inc.   '          '

       Cagle's, Inc. and its wholly owned subsidiary Cagle's Farms, Inc., raise broiler chickens to
produce fresh and frozen poultry products. The company's vertically integrated Operations include
breeding, hatching, and growing chickens, as well as feed milling, processing, further processing, and
marketing. In 1999 Cagle's weekly production was 19.7 million pounds of ready-to-cook chicken and its
fiscal 1999 sales were $310 million (Thornton, 2000b).

        Cagle's processed approximately 2.2 million birds per week in three processing plants in 1999, two
of which operated in double shifts, and two further processing plants. Cagle?s expected to begin operation
of its new Perry, GA, processing facility in September 2000 with a capacity of 1.2 million head of broilers
per week.

        In 1999 Cagle's owned a 50 percent interest in a joint venture fully integrated poultry company
located in Camilla, GA. As of 1999, this facility was growing and processing approximately 1.3 million
birds per week. Cagle's also formed another joint venture partnership with Executive Holdings, L.P. called
Cagle's-Keystone Foods LLC which was expected to construct an integrated poultry complex in Kentucky
(Daily Edition, 1997; Hoover's, 2000).22  Keystone Foods is a privately-owned meat processor of frozen
meat products made from purchased beef.  Meatnewsicom estimated that Keystone was the eleventh largest
 meat processor in the U.S. in 1999 (Meat Processing, 2000).
         Foster Poultry Farms

         Foster Poultry Farms is a vertically integrated company producing quality chicken and turkey
 products.  In 1999 Foster Farms slaughtered 4.1 million birds weekly, producing 15=4 million pounds of
 ready-to-cook chicken per week (Thornton, 2000b).  The ninth largest broiler company in the U.S. as of
 1999, Foster Farms was also the largest poultry farm in the Western U.S. (Foster Farms, 2000).  Sales
 for Foster Farms in 1999 were $1.1 billion (Meat&Poultry, 2000a).
         22 This is EPA's interpretation of the relationship between these three entities as of 1999, based on very
  limited information.
                                                2-96

-------
       As of 1999, Foster Farms' broiler operations included four processing plants, four further
processing plants, two rendering plants, and three wastewater treatment plants in addition to hatcheries,
grow-out ranches, feed mills, and distribution centers (Thornton, 2000b). Foster Farms completed an
expansion project in its Fresno chicken plant in fiscal. 2000, adding 45,000 square feet to the 150,000
square feet facility  (Meat&Poultry, 2000a).

        Since then, Foster Farms also announced plans to acquire Zacky Farms' chicken operations
(Meat&Poultry, 2001c). Zacky Farms' operations include a processing plant, feed mill, hatchery, and 35
ranches and its inclusion in the Foster family is expected to increase Foster's chicken production by 25
percent (Meat&Poultry, 2001c).

        Seaboard  Farms

        Seaboard  Corp. is a diversified international agribusiness and transportation company. Through
 1999, poultry production took place through its wholly-owned subsidiary Seaboard Farms.  As part of its
 domestic operations, the company also produces and processes pork.                  -

         As of January 2000, Seaboard Farms was acquired by ConAgra, Inc. for $375 million. The
 facilities sold included four processing plants and two further processing plants. In 1999 the company
 produced 14.5 million pounds of ready-to-cook chicken per week (Thornton, 2000b). Having completed
 several capital improvements to increase capacity prior to the acquisition, Seaboard had hoped to increase
 production by two million pounds per week in 2000 (Thornton, 2000b). Seaboard Farms earned $460
 million in fiscal 1999 sales (Thornton, 2000b).
         2.5.3.2 Turkey Companies
          Jennie-O Foods, Inc.
          Jennie-0 Foods, Inc. was the nation's largest turkey processor in 1999, based on live pounds
  processed (Heffernan, 2000). The company produced approximately 859 million live pounds of turkey in
                                                2-97

-------
that year (Heffeman, 2000). Jennie-O Foods is a wholly owned subsidiary of Hormel Foods Corporation,
which was ranked as the seventh largest meat processor, and seventh largest pork slaughterer in the U.S. in
1999.

        A vertically integrated turkey operation, Jenhie-O apparently owned four processing and four
further processing plants in 1999.  Capital improvements, including the expansion of a plant and new
processing equipment, were expected to increase Jennie-O's output by 40 million pounds in the year 2000
(Hormel, 2000).  In 2001, Hormel Foods purchased The Turkey Store Company, the sixth largest turkey
producer in the U.S. Combined with Jennie-O's turkey production, Hormel is expected to produce more
than 1.2 billion pounds of turkey annually (Meat&Poultry, 2001b).
       Butterball Turkey Co.

       ConAgra Poultry's operations include the integrated production of turkeys under the Butterball
Turkey Company label. Butterball Turkey Company, the second largest U.S. turkey processor in 1999,
operates in the Refrigerated Foods Division (Heffeman, 2000).

       Butterball Turkey operated four processing plants in 1999; a fifth processing plant in California
was sold to Foster Farms in July 1999. Heffeman (2000) estimated that this decreased Butterball's
slaughter by 40 million pounds in 1999.

       As of 1999, ConAgra had not announced any restructuring plans associated with its acquisition of
Seaboard Corporation's broiler operations.  Assuming ConAgra closes none of Seaboard's plants,
ConAgra could produce a total of 4.2 billion pounds of poultry between its ConAgra, Seaboard, and
Butterball facilities, which could make it the second largest poultry producer in the U.S. (including turkey
production). For more information on ConAgra, see Section 2.5.1.
                                              2-98

-------
       Cargill. North American Turkey Operations

       Cargill, Inc. is an international marketer, processor and distributor of agricultural, food, and
industrial products.  One of its two subsidiaries, Cargill North American Turkey Operations is a turkey
processor. Cargill's revenues for all meat and poultry related operations in 1999 were estimated at $9
billion (Meat&Poultry, 2000a). It ranked as the largest privately-owned company in the U.S. in 1999
(Hoover's, 2000).                                                                               •

       According to WATT PoultryUSA, Cargill was the country's third largest turkey processor,
slaughtering 715 million pounds of turkeys by live weight in 1999 (Heffernan, 2000).  At the time,
Cargill's four processing plants had the capacity to handle 23,000 birds a day (Cargill, 2000). Cargill
acquired Plantation  Foods  in September 1998 (Heffernan, 2000).

        Cargill acquired Rocco Enterprises' turkey opeiations in 2001 (Meat&Poultry, 2001h). This
acquisition is expected to increase Cargill's turkey sales to $1 billion.  For more information on Cargill, see
Section 2.5.1.
        Wampler Foods, Inc.

        Wampler Foods, Inc., a subsidiary of WLR Foods, Inc., produces, processes, and markets fresh,
 frozen, and further processed chicken and turkey. Wampler was the thirteenth largest broiler processor and
 the fourth largest turkey processor in 1999 as measured by live slaughter weight. Its combined turkey and
 broiler operations made it the seventh largest overall poultry processor.  WLR Foods had sales of $890
 million in fiscal 1999 (Meat&Poultry, 2000a). A vertically integrated company, Wampler Foods' primary
 operations include the breeding, hatching, grow-out and processing of chickens and turkeys.

        The company owned four chicken processing plants with a double-shift capacity of 3.7 million
 chickens per week in 1999 (Thornton, 2000c).  Wampler had three turkey processing plants with a
 slaughter capacity of 450,000 turkeys per week on a single shift  as of 1999 (Heffernan, 2000). In 2001,
 WLR Foods, Inc. was purchased by the Pilgrims Pride Corporation (Meat&Poultry, 2001g).
                                               2-99,

-------
        Carolina Turkeys

       Carolina Turkeys is jointly owned by Carroll's Foods, Inc. and Goldsboro Milling Company
(Carroll's Foods, 2000).  In May 2000, Smithfield Foods, Inc. acquired Carroll's Foods and its 49 percent
interest in Carolina Turkeys. Carolina Turkeys was the fifth largest turkey producer in the U.S. with an
annual production of 460 million live pounds in 1999 (Heffernan, 2000). Carolina Turkey's 1999 fiscal
sales amounted to $350 million (Meat&Poultry, 2000a).

       Carolina Turkeys is an integrated producer and had the largest processing plant in the United
States in  1999 (Carroll's  Foods, 2000). The company processed 22 million turkeys in 1999 and production
took place round the clock  (Carolina Turkeys, 2000).  Carolina Turkeys also had its own hatcheries,
breeding farms, feed mills, growing farms, research farms, and diagnostic labs (Carolina Turkeys, 2000).
        2.5.4   Overall Ranking of Meat Processing Companies

        Table 2-36 presents summary information for all meat product industry companies with 1999
revenues in excess of $250 million.  Although most of the companies perform slaughter operations, and
have appeared already in Tables 2-31 through 2-35, a number of companies that primarily perform
processing operations do appear in Table 2-36. The companies meeting the revenue cutoff for Table 2-36
are predominantly companies that perform at least some slaughter operations. Of the 71 companies listed,
only 12 were confirmed as having minimal slaughter operations. Among the top 15 companies listed in
Table 2-36, only three (Oscar Mayer, Keystone Foods, and OSI International Foods) apparently do not
perform significant slaughtering operations. It is interesting to note that these three companies all employ
significantly fewer workers than the slaughter companies with similar 1999 revenues. This is consistent
with the census data in Sections 2.1  above, which showed that processing plants tended to be smaller than
slaughter plants, but have a relatively greater value added.
                                              2-100

-------
            Table 2-36
     Meat Processing Firms with
1999 Revenues Exceeding $250 Million
1: ;-:;: ;:-:*vr:^f
Company
EBP1
ConAgra 2 	
Carsill Red Meat Group (Excel)
Tyson Foods 	
Sara Lee 3
Farmland Refrigerated Foods
jSmithfield Foods5 	
HHormel Foods6
Oscar Mayer 7
Perdue Farms
Keystone Foods 	 	
OSI Int'l Foods
Gold Kist
John Morrell 5
Pilgrim's Pride 	
Packerland Packing 	 .__ 	
Foster Farms
Wampler Foods 8
I 	 t. 	 . 	 ; 	
Wavne Farms 9
Seaboard Corporation (pork) 10
Corporate Food Brands America '
Empire Beef
Colorado Boxed Beef
Brvan Foods 3
Rosen's Diversified
GFI America
Mover Packing
Sanderson Farms
Wolverine Packing 	
Rocco Enterprises n
American Foods Group
Greater Omaha Packing 	
Koch Foods
Townsends 	
Rank
1
2
3
4
5
6
6
8
9
9
11
11
13
, 14
15
16
17
18
19
20
21
22
• 23
24
25
26
27
27
27
30
31
31
31
34
Type
M
M
M
P
M
M
M
M
M
P
M
M
P
M
P
M
P
P
P
M
M
M
M
M
M
M
M
P
M
P
M
M
P
P
Slaughter
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
N
Y
Y
Y
Y
Y
Y
Y
Y
N
N
N
Y
Y
Y
Y
Y

Y
Y
Y
Y
Y
1999 Fiscal
Year Sales
($ millions)
$14,100
$12,500
$9,000
$7,400
$4,100
$3,800
$3,800
$3,400
$2,500
$2,500
$2,200
$2,200
$1,800
$1,600
$1,400
$1,300
$1,100
$890
$830
$820
$800
$720
. $650
$640
$620
$600
$560
$560
$560
$550
$53C
$53C
$53C
$52C
1999
Employment
49,000
48,000
20,000
65,000
15,000 4
12,000
25,000
12,000
9,000
19,000
2,500
2,000
18,000
6,000
15,000
4,000
8,900
7,100
9,100
4,100
3,600
230
450
2,100
1,000
1,200
1,600
7,700
250
3,600
1,500
65C
) 4,40C
i 4,40C
Plants
60
- 72
18
56
30"
14
31
12
8
33
15
14
14
8
15
4
10
8
12
1
11
3
1
2
3
4
2
7
3
4
2
1
9
5||
                2-101

-------
         Table 2-36 (cont.)
     Meat Processing Firms with
1999 Revenues Exceeding $250 Million
Company
Nebraska Beef
Sherwood Food Distributors
Enimpak Foods
House of Raeford Farms
Seaboard Farms (poultry) 2
Fieldale Farms
Simmons Foods
Taylor Packing
Indiana Packers
Carolina Turkeys 5
Hatfield Quality Meats
Sysco Corp.
B.C. Rogers Poultry
Zacky Foods
Clougherty Packing
Sam Kane Beef Processors
Cagle's
Bar-S Foods
Allen Family Foods
Mountaire Farms
Peco Foods
Premium Standard Farms
Freshmark
Washington Beef
Bob Evans Farms
Buckhead Beef
Choctaw Maid Farms
International Trading
Lundy Packing
Omaha Steaks Int'l
PM Holdings
United Food Group
Barker's/Lombard! Bros.
JAO Long Island Beef
O.K. Foods
Rank
35
35
37
38
39
40
41
42
43
44
45
45
47
47
49
49
51
51
53
53
53
53
57
57
59
60
60
60
60
60
60
60
60
60
60
Type
M
M
M
P
P
P
P
M
M
P
M
M
P
P
M
M
P
M
P.
P
P
M
M
M
M
M
P
M
M
M
M
M
M
M
P
Slaughter
Y

Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
N
Y
Y

Y

Y
N
Y
N
N
N
Y
1999 Fiscal
Year Sales
($ millions)
$500 - $865
$500 - $865
$490
$480
$460
$450
$420
$380
$360
$350
$340
$340
$330
$330
$320
$320
$310
$310
$300
$300
$300
$300
$280
$280
$260
$250
$250
$250
$250
$250
$250
$250
$250 - $499
$250 - $499
$250 - $499
1999
Employment
1,200
550
1,800
5,000
5,000
4,800
4,300
1,000
1,200
2,300
1,600

3,400
3,000
1,300
600
7,000
1,500
2,400
2,900
3,900
800
1,500
620
350
430
3,200
1,000
900
1,500
800
380
650
350
4,300
Plants
1
3
3
8
6
4
7
1
1
1
4
3
4
3
2
1
8
i
.
*
6
1
;
*.
6
1
:
f.
:
t
ฃ.

:
i.
'
                2-102

-------
                                         Table 2-36 (cont.)
                                    Meat Processing Firms with
                               1999 Revenues Exceeding $2fp Million


Company
Randall Farms



Rank
60
60


Type
P



Slaughter
Y
Y
1999 Fiscal
Year Sales
($ millions) •
$250 - $499
$250 - $499

1999
Employment
600
2,700

'['.- '' • ":t .-•_;• - '•'• '
^Plants
3
2
Thornton, 2000a; Thornton, 2000b; Thornton, 2000c.                                            .
1 IBP purchased Corporate-Brand Foods America in fiscal 2000.
2 ConAgra's significant divisions include Butterball Turkey and ConAgra Poultry. ConAgra also recently acquired
Seaboard Corporation's poultry division, Seaboard Farms. The 1999 fiscal year sales and employment numbers for
Seaboard Farms (poultry) do not include Seaboard Corporation's pork business.
3 Sara Lee's significant subsidiaries include Bil Mar (turkey) and Bryan Foods (pork).
4 Sara Lee employment and number of plants differed between sources by roughly 100 percent.
5 Smithfield's significant subsidiaries include Carolina Turkeys and John Morrell (pork).
6 Hormel's major subsidiary is Jennie-O Foods, the largest turkey processor in the U.S.
7 Oscar Mayer is a brand name of Kraft Foods, a subsidiary of Philip Morris Companies, Inc.
8 Wampler Farms is a subsidiary of WLR Foods, Inc.
9 Wayne Farms is a division of ContiGroup Companies.
 10 Seaboard Corporation's 1999 fiscal year sales and employment numbers do not include Seaboard Farms
 (Seaboard's poultry division recently acquired by ConAgra).
 11 Rocco Enterprises carries out its turkey operations through its subsidiary Shady Brook Farms.
                                                 2-103

-------
       Oscar Mayer             '

       Oscar Mayer is a brand of Kraft Foods North America, in itself, a wholly owned subsidiary of
Philip Morris Companies, Inc. As of 1999, Oscar Mayer's meat production took place in its nine
slaughtering and processing plants. Thus Oscar Mayer did perform slaughter operations, however, they
were apparently not large enough to rank.among the top 20 slaughter operations for either beef or pork.
Presumably then, its high ranking in the meat product industry must have been due primarily to its
processing operations. Oscar Mayer's business growth can be attributed to its focus on "quick-to-fix"
products (Meat&Poultry, 2000b). The company had 1999 fiscal sales' of $2.5 billion (Meat Processing,
2000). Kraft Foods North America's operating revenues in fiscal 1999 were $17.5  billion.  In June 2001,
Kraft Foods became a publicly traded company (Meat&Poultry, 200 li). Owing to the large, diversified
business interests of both Philip Morris and Kraft, EPA could not find additional information on Oscar
Mayer from the 10-K, annual report, or the company's website.

        Keystone Foods

        Keystone Foods is a privately held meat and poultry processor operating 15 processing plants as of
 1999.  Keystone apparently performed little or no livestock slaughter, and thus, its revenues were
presumably from its processing operations.  This company had fiscal 1999 sales amounting to $2.2. billion
(Meat Processing, 2000). Keystone owns a joint venture partner with Cagle's, a producer of broilers
 (Hoover's, 2000).  Little public information is available on Keystone.
        OSI International Foods

        OSI Group of Companies, previously known as Glenmark, is the parent company of OSI
 International Foods (OSI, 2000). OSI processed beef, pork, and poultry in its 14 meat plants in 1999; OSI
 apparently performed little or-no livestock slaughter, and its revenues were apparently from its processing
 operations. A privately held company, OSI had estimated fiscal 1999 sales of $2.2 billion (Meat
 Processing, 2000).
                                              2-104

-------
2.6    REFERENCES


AMI. 2000a. Meat Consumption in the U.S. American Meat Institute, http://www.meatami.org.
       Downloaded on October 18, 2000.

AMI. 2000b. What Is in the Meat Case of the 1990s and Beyond. American Meat Institute.
       http://www.meatami.org. Downloaded on October 19, 2000.

Anderson, Donald W., Brian C. Murray, Jacqueline L. Teague, and Richard C Lindrooth. 1998. Exit from
       the Meatpacking Industry: A Microdata Analysis. American Journal of Agricultural Economics.
       80(1 February):96- 106.                         .

 Azzam, Azzeddine. 1992. Testing the Competitiveness of Food Price Spreads. Journal of Agricultural
       Economics- '43:248-256.
 Azzam, Azzeddine M., and Dale G. Anderson. 1996. Assessing Competition in
        History, Theory, and Evidence. Report GIPSA-RR 96-6. Prepared for the U.S. Department of
        Agriculture, Grain Inspection, Packers, and Stockyard Administration.

 Azzam Azzeddine M., and John R. Schroeter. 1995. The Tradeoff Between Oligopsony Power and Cost
        Efficiency in Horizontal Consolidation: An Example From Beef Packing. American Journal of
        Agricultural Economics. 77(3 August):825-836.               '                            .

 Bailey DeeVon, B. Wade Brorsen, and Michael R. Thomsen. 1995. Identifying Buyer Market Areas^and
        the Impact of Buyer Concentration in Feeder Cattle Markets Using Mapping and Spatial Statistics.
        American Journal of Agricultural Economics. 77(2 May):309-3 18.

 Bernard, John C., and Lois Schertz Willett. 1996. Asymmetric Price Relationships in the US. Broiler
         Industry. Journal of Agricultural and Applied Economics. 28(2 December):279-289.
  Carpenter, Dave. 2000. Beefing Up-U.S. Meat Consumption Ends Long Slurry Fa* '-Market
         March 13. http://www.foxmarketwire.com/031300/beef.sml. Downloaded on November 1, 2000.

  Clark J. Stephen, and Albert J. Reed. 2000. Structural Change and Tests of Market Power in the U.S.
         Foodlndustry. Presented at The American Consumer and the Changing Structure of the Food
         System Conference, Sponsored by U.S. Department of Agriculture, Economic Research Service,
         Washington D.C.

  Daily Edition. 1997. Cagle's Forms Joint Venture. Atlanta Business Chronicle. November 17
         httP.y/www.biZjournals.co^atlanta/storieS/1997/n/17/daily7.htmL

         2000.

  Flake Oliver L., and Paul M. Patterson. 1999. Health, Food Safety and Meat Demand. Presented at the
       '  1999 Annual Meeting of the American Agricultural Economics Association, Nashville, Tennessee.
                                              2-105

-------
GBPSA. 1997. Packers and Stockyards Statistical Report: 1995 Reporting Year. U.S. Department of
       Agriculture.

GIPSA. 2000. Packers and Stockyards Statistical Report: 1998 Reportmg Year. U.S. Department of
       Agriculture.

Hayenga, Marvin L. 1997. Cost Structures of Pork Slaughter and Processing Firms: Behavioral and
       Performance Implications. Review of Agricultural Economics. 20(2):574-583.

Heffeman, Bernard E. 2000. Top Turkey Companies in the USA: Jennie-O Solidifies Top Ranking. WATT
       PoultryUSA. January: 19-23. http://www.wattnet.com.

Hetrick, Ron L. 1994. Why Did Employment Expand in Poultry Processing Plants? Monthly Labor
       Revie\v. June:31-34.

Hoover's Online. 2000. Austin, Texas: Hoover's Inc. http://www.hoovers.com. Company profiles
        downloaded in July, August, and September.

Hughes, Harlan. 2000. Market Share of Top Packers. North Dakota State University.
        http://w\v\v.ag.ndsu.nodak.edu/aginfo/lsmktadv/topics/marketshare.htm. Downloaded on August
        2,2000.

Hunnicutt, Lynn and Quinn Weninger. 1999. Testing for Market Power in Beef Packing: Where Are We
        and WJiat's Next? Virginia Tech (sic), Research Institute on Livestock Pricing, Research Bulletin
        7-99.

Koontz, Stephen R., Philip Garcia, and Michael A. Hudson. 1993. Meatpacker Conduct in Fed Cattle
        Pricing: An Investigation of Oligopsony Power. American Journal of Agricultural Economics.
        75(3 August):537-548.

Krizner, Ken.  1995. Packers Shed Their Case-Ready Skepticism. Meat Marketing & Technology. April.
        http://www.mtgplace.com. Downloaded on October 19,2000.

Krizner, Ken.  1998. Turning a Vital Comer. Meat Marketing & Technology. March.
        http://wmv.mtgplace.com. Downloaded on October 19, 2000.

Lewin-Solomons, Shira B. 1999. Asset Specificity and Hold-up in Franchising and Grower Contracts: A
        Theoretical Rationale for Government Regulation? Iowa State University Working Paper.

MacDonald, James M., Michael E. Ollinger, Kenneth E. Nelson, and Charles R. Handy. 2000.
        Consolidation in U.S. Meatpacking. Agriculture Economic Report No. 785. Washington, D.C.:
        U.S. Department of Agriculture, Economic Research Service. February.

Mathews, Kenneth H., Jr., William F. Hahn, Kenneth E. Nelson, Lawrence A. Duewer, and Ronald A.
        Gustafson. 1999. U.S. Beef Industry: Cattle Cycles, Price Spreads, and Packer Concentration.
        Technical Bulletin 1874. Washington, D.C.: U.S. Department of Agriculture, Economic Research
        Service. April.

                                             2-106

-------
Martinez, Steve W. 1999. Vertical Coordination in the Pork and Broiler Industries: Implications for
       Pork and Chicken Products. Agricultural Economic Report No. 777. Washington, D.C.: U.S.
       Department of Agriculture, Economic .Research Service.

Meat&Poultry. 1999. Japan's Evolving Marketplace. Meat&Poultry. December.
       http://www.meatpoultry.com/sitesearch/siteSearchJeatures_art.asp?ArticlelD=25135.
       Downloaded on November 2, 2000.                                        '

Meat&Poultry. 2000a. 22nd Annual Ranking of Meat and Poultry Companies: The Top 100.
       Meat&Poultry. July: 45-72.

Meat&Poultry. 2000b. Corporate Profiles. Meat&Poultry. August: 30-56.

Meat&Poultry. 2000c. Ranking the World's Largest Meat and Poultry Companies. Meat&Poultry.
       October: 21-36.

Meat&Poultry. 2000d. The Front Page. Meat&Poultry. November: 3-4.

Meat&Poultry. 2001a. The Front Page. Meat&Poultry. January: 3-4.

Meat&Pouliry. 2001b. The Front Page. Meat&Poultry. February: 3-4.

Meat&Poultry. 200 Ic. The Front Page. Meat&Poultry. April: 3-4.

Meat&Poultry. 2001d. The Front Page. Meat&Poultry. May: 3-4.

Meat&Poultry. 2001e. The Front Page. Meat&Poultry. June: 3-4.

 Meat&Poultry. 2001f. The Front Page. Meat&Poultry. July: 3-4.

 Meat&Poultry. 2001g. Money Makers: How the Top 50 Stack Up. Meat&Poultry. July: 25-43.

 Meat&Poultry. 2001h. The Front Page. Meat&Poultry. August: 3-4.                          ,

 Meat&Poultry. 2001L The Billionaires' Club. Meat&Poultry^  August: 19-47.

 Meat&Poultry. 200 Ij. The Front Page. Meat&Poultry. September: 3-4.

 Meat&Poultry. 2001k. The Front Page. Meat&Poultry. October: 3-4.

 Meat&Poultry. 20011. The Front Page. Meat&Poultry. November: 3-4.

 Meat&Poultry. 2000m. The Front Page. Meat&Poultry. December: 3-4.

 Meat&Poultry. 2000n, Business Nots: Excel Adds Taylor Packing to its Menu. Meat&Poultry. December:
         14.              '    -  •
                                             2-107

-------
Meat Marketing & Technology. 2000. The MEATing Place 2000 MMT Industry Elite. Meat Marketing &
       Technology. http://www.mtgplace.com/SpecialFeatures/MMTElite2000/Index.asp. Downloaded
       on July 24,2000.

Meat Processing. 1999. Top 200 Companies: Meat Packers, Processors, and Poultry Companies. Meat
       Processing. June. http://w\vw.wattnet.com/. Downloaded on August 12, 1999.

Meat Processing. 2000. The Top 200 in 2000. Meat Processing. June: 39-46.

Moon, Wanki, and Ronald W. Ward. 1999. Effects of Health Concerns and Consumer Characteristics on
       U.S. Meat Consumption. Presented at the 1999 Annual Meeting of the American Agricultural
       Economics Association, Nashville, Tennessee.

Muth, Mary K., and Michael K. Wohlgenant. 1999. A Test for Market Power Using Marginal Input and
       Output Prices With Application to the U.S. Beef Processing Industry. American Journal of
       Agricultural Economics. 81(3 August):638-643.

NCSA. 2000a. International Markets: Position Paper—2000. National Cattlemen's Beef Association.
       http://www.beef.org/library/factsheets/pp_int_markets.htm. Downloaded on October 19, 2000.

NCSA. 2000b. Diet and Health; Position Paper—2000. National Cattlemen's Beef Association.
       http://www.beef.org/library/factsheets/pp_diet.htm. Downloaded on October 19,2000.

NPPC. 1999a. Pork Industry Statistics & Information. National'Pork Producers Council.
http:/Awv\v.nppc.org/PorkFacts/pfindex.html.             •                    •

NPPC. 1999b. 1999/2000 Pork Issues Handbook. National Pork Producers Council.
       hUp://\v\vw.nppc.org/IssueHandbook/1999_2000_Handbook.html.

Nunes, Keith. 1997. Whats for Dinner? Sausage Processors Are Setting Their Sights on the Occasion.
       Meat&Poultry. http://www.meatpoultry.com. Downloaded on October 18,2000.

Nunes, Keith. 1999. Consumers and Case Ready: Do They Care? Meat&Poultry.
        http://www.meatpoultry.com. Downloaded on November 2, 2000.

Ollinger, Michael, James M. MacDonald, Charles R. Handy, and Kenneth E. Nelson. 1997. Structural
        Change in the U.S. Meat and Poultry Industries. Paper Presented at NE-165 Conference; Strategy
        and Policy in the Food System: Emerging Issues. Food Marketing Policy Center, University of
        Connecticut.

 Ollinger, Michael, James MacDonald, and Milton Madison. 2000. Structural Change in U.S. Chicken and
        Turkey Slaughter. Agricultural Economic Report No. 787. Washington, D.C.: U.S. Department of
        Agriculture, Economic Research Service.

 Outlook. 12/27/00 and 8/29/00. Livestock, Dairy and Poultry Situation and Outlook. Washington, D.C.:
        U.S. Department of Agriculture, Economic Research Service.
                                             2-108

-------
Paarlberg, Philip L., and Mildred M. Haley. 2000. Market Concentration and Vertical Coordination in
       the Pork Industry: Implications for Public Policy Analysis. Presented at The American Consumer
       and the Changing Structure of the Food System Conference, Sponsored by U. S. Department of
       Agriculture, Economic Research Service, Washington D.C.

Paarlberg, Philip L., Michael Boehlje, Kenneth Foster, Otto Doering, and Wallace Tyner. 1999. Structural
       Change and Market Performance in Agriculture: Critical Issues and Concerns About
       Concentration in the Pork Industry. Purdue University, Department of Agricultural Economics,
       Staff Paper No. 99-14.

Paul, Catherine J. Morrison. 1999. Aggregation and the Measurement of Technological and Market
       Structure: The Case of the U.S. Meatpacking Industry. American Journal of Agricultural
       Economics. 81(3 August):624-629.

Putnam, Judith J., and Jane E. Allshouse. 1999. Food Consumption, Prices, and Expenditures, 1970-97.
       Statistical Bulletin Number 965. Washington, D.C.: U.S. Department of Agriculture, Food and
       Rural Economics Division, Economic Research Service.

Rice, Judy. 1998. Formulation Source; Bringing Beef Back. Meat&Poultry. http://www.meatpoultry.com/.
       Downloaded  on November 2,2000.

Rogers, Richard T., and Richard J. Sexton. 1994. Assessing the Importance of Oligopsony Power in
        Agricultural Markets. American Journal of Agricultural Economics. 76(5 December): 1143-1150.

 Salvage, Bryan. 1999. Emmpdk To Test Ground Beef Irradiation. Ohio State University Meat Science.
        http://www.ag.ohio-state.edu/~meatsci/news/6_15_emmpak.html. Downloaded on August 17,
        2000.

 Schroeter, John R. 1988. Estimating the Degree of Market Power in the Beef Packing Industry. Review of
        Economics and Statistics. 70(February):158-162.

 Schroeter, John R., and Azzeddine Azzam. 1990. Measuring Market Power in Multiproduct Oligopolies:
        The U.S. Meat Industry. Applied Economics. 22:1365-1376.

 Schroeter, John R., Azzeddine Azzam, and Mingxia Zhang. 2000. Measuring Market Power in Bilateral
        Oligopoly: The Wholesale Market for Beef. Southern Economic Journal. 66(3 January):526-547.

 Standard & Poor's. 1999. Agribusiness. Standard & Poor's Industry Surveys. 167(30).

 Stiegert, Kyle W., Azzeddine Azzam, and B. Wade Brorsen.  1993. Markdown Pricing and Cattle Supply in
        the Beef Packing Industry. American Journal of Agricultural Economics. 75(3 August):549-558.

 Texas Agricultural Market Research Center. 1996. Price Determination in Slaughter Cattle
        Procurement. Report GIPSA-RR 96-2. Prepared for the U.S. Department of Agriculture, Grain
        Inspection, Packers and Stockyard Administration.
                                              2-109

-------
Thornton, Gary. 2000a. Directory of Broiler Plants: Processing, Further Processing Plants. WATT
       PoultryUSA. January: 82-91. http://www.wattnet.com.

Thornton, Gary. 2000b. Top 10 Broiler Companies in the USA: Tyson Reigns at the Top. WATT
       PoultryUSA. January: 30-34. http://www.wattnet.com.

Thornton, Gary. 2000c. U.S. Broiler Companies: A to Z Profiles. WATT PoultryUSA. January: 42-56.
       http://\vw\v.wattnet.com.

Tsoulouhas, Theofanis and Tomislav Vukina. 1999. Integrator  Contracts with Many Agents and
       Bankruptcy. American Journal of Agricultural Economics. 81(1 February):61-74.

Tsoulouhas, Theofanis and Tomislav Vukina, 2000. Regulating Broiler Contacts: Tournaments versus
       Fixed Performance Standards. North Carolina State University, Department of Agricultural and
       Resource Economics Working Paper.                       .

U.S. Census Bureau. 1999a. Animal (Except Poultry) Slaughtering. EC97M-3116A. 1997 Economic
       Census: Manufacturing Industry Series. Washington, B.C.: U.S. Department of Commerce.
       November.

U.S. Census Bureau. 1999b. Meat Processed From Carcasses. EC97M-3116B. 1997 Economic Census:
       Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.

U.S. Census Bureau. 1999c. Poultry Processing. EC97M-3116D. 1997 Economic Census: Manufacturing
       Industry Series. Washington, D.C.: U.S. Department of Commerce. November.

U.S. Census Bureau. 1999d. Rendering and Meat Byproduct Processing. EC97M-3116C. 1997 Economic
       Census: Manufacturing Industry Series. Washington,  D.C.: U.S. Department of Commerce.
       December.

U.S. Census Bureau. 2000. Annual Survey of Manufacturers: Geographic Area Statistics, 1998.
       Washington, D.C.: U.S. Department of Commerce. October.

USDA. 1997a. U.S. Department of Agriculture. Poultry Yearbook. Washington, D.C.: U.S. Department of
       Agriculture, Economic Research Service. March.

USDA. 1997b. U.S. Department of Agriculture. World Beef & Cattle Trade: Evolving and Expanding.
       Agricultural Outlook. Washington, D.C.: U.S. Department of Agriculture, Economic Research
       Service.

USITC. 1998. Industry & Trade Summary. USITC Publication 3148. Washington, D.C.: U.S.
       International Trade Commission,, Office of Industries.  December.

Weliwita, Ananda, and Azzeddine Azzam. 1996. Identifying Implicit Collusion Under Declining Output
       Demand. Journal of Agricultural and Resource Economics. 21(2 December):235-246.
                                            2-110

-------
WSJ. 2000. IBP Accepts
       C27.
$24 Billion Offer for a Buyout by DLJ Fund. Wall Street Journal. 8 October:
COMPANY WEB SITES REFERENCED

American Foods Group. 2000. http://www.american-foods.com. Downloaded on August 17,2000.

Cargffl. 2000. http://www.cargillfoods.com. Downloaded on August 11, 2000.

Carolina Turkeys. 2000. http://www.carolinaturkeys.com. Downloaded on August 14, 2000.

Carroll's Foods. 2000. http://www.carrollsfaods.com. Downloaded on August 14, 2000.

 ConAgra. 2000. http://www.conagra.com. Downloaded on August 16, 2000.

 Excel. 2000. http://www.excelmeats.com. Downloaded on July 20,2000.

 Farmland. 2000a. http://www.farmland.com. .Downloaded on August 7, 2000.

 Farmland. 2000b. http://www.nationdlbeef.com. Downloaded on August 7, 2000.

 Foster Farms. 2000; http://www.fosterfarms.com. Downloaded on August 11, 2000.

 GFI. 2000. http://www.gfiamerica.com. Downloaded on August 17, 2000.

 GFI. 2001. http://www.gfiamerica.com. Downloaded on November 27, 2001.

  Hormel.  2000. http://www.hormel.com. Downloaded on August 16, 2000.

  IBP. 2000. http://www.ibpinc.com. Downloaded on August 16, 2000.

  IBP. 2001. http://www.ibpinc.com. Downloaded on November 27,2001.

  Moyer. 2000. http://www.mopac.com. Downloaded on August 17, 2000.

  OSI. 2000. http://www.osigroup.com. Downloaded on September 27, 2000.

  Perdue. 2000. http://www.perdue.com. Downloaded on August 11,2000.

  Pilgrims Pride. 2001. http://www.pilgrimspride.com. Downloaded on November 8, 2001.

   Swift & Company. 2000. http://www.freshpork.com/swifi/conagra.htm. Downloaded on August 14, 2000.

   Taylor. 2000. http://www.taylorpacking.com. Downloaded on August 18, 2000.
                                             2-111

-------

-------
                                       CHAPTERS
                      ECONOMIC IMPACT METHODOLOGY
       This section provides a brief overview of the methodology used in the economic impact,
regulatory flexibility, and environmental justice analyses. EPA will use two methodologies to evaluate
economic impacts of the effluent limitations and guidelines (ELGs) on the meat products industry. For
the proposed rule, EPA evaluated impacts based on models developed from publicly available
information obtained from the U.S. Census Bureau, the U.S. Department of Agriculture, and other   -
sources. For the final rule, EPA will examine impacts based on data collected in the Section 308 Meat
Products Industry Survey. (This survey is the reason.why EPA chose to use two approaches: the detailed
survey could not be completed in time for EPA to incorporate its data into the economic impact
analysis.) Section 3.1 presents the methodology used to evaluate the impacts of the proposed rule.
Section 3.2 presents the methodology that will be used to evaluate the impacts of the final rule.

        The discussion in Section 3.1 works from the smallest scale (costs for specific configurations of
option, subcategory, and site) up to the largest scale  (market analysis). The section presents the economic
impact methodology as follows:

        •      Cost annualization model, Section 3.1.1
        •      Facility-level impacts model, Section 3.1.2
        •      Financial ratio analysis, Section,3.1.3
        •      Market model, Section 3.1.4
         •      National impacts, Section 3.1.5

 The results of these analyses are presented in Chapter 5.

         In general, the methodologies that will be used for the final rule are the same as those used for
 the proposed rule. However, for the final rule the analysis will primarily be based on survey data.  For
                                               3-1

-------
the proposed rule, most analysis is based on publicly available data. Section 3.2 will discuss the
differences between the two methodologies.
3.1    METHODOLOGY FOR THE PROPOSED RULE
       3.1.1   Cost Annualization Model
       The beginning point for any analysis is the cost annualization model (see Figure 3-1). Inputs to
the cost annualization model come from EPA's engineering staff and secondary data.

       EPA's engineering staff developed capital and operating and maintenance (O&M) costs for
incremental pollution control. The capital cost, a one-tune cost, is the initial investment needed to
purchase and install equipment involved in pollution control. The O&M cost is the annual cost of
operating and maintaining that equipment; a site incurs its O&M cost each year. For this proposal, EPA
estimated average compliance costs for a series of model facilities based on subcategory, size, and  •
discharge type (for details, see the Development Document, U.S. EPA, 2002).

        Annualized costs are calculated as the equal annual payments of an annuity that has the same
present value as the stream of cash outflow over the project life and includes the opportunity cost of
money or interest. An annualized cost is analogous to a mortgage payment that spreads the one-time
investment of a home over a series of constant monthly payments. There are two reasons to annualize
capital and O&M costs. First, the capital cost is incurred only once in the equipment's lifetime;
therefore, initial investment should be expended over the life of the equipment.  Second, money has a
time-based value, so expenditures incurred at the end of the equipment's lifetime or O&M expenses in
the future are not the same as expenses paid today.

        All other inputs into the cost annualization model are from secondary data sources. The
depreciation method used in the cost annualization model is the Modified Accelerated Cost Recovery
System (MACRS). MACRS can model businesses as depreciating a higher percentage of an investment
in the early years and a lower percentage in the later years. A real discount rate of 7 percent, as
recommended by OMB, was used to represent the opportunity cost of capital (OMB,  1996).

                                               3-2

-------
Data Sources      Inputs
Engineering
Incremental
Pollution Control
Costs
 Secondary
 Sources,
                   Capital Costs
O&M Costs
Cost Deflator to  	
$1999,

Depreciation Method-
(MACRS)
                    Federal Tax Rate

                     State Tax Rate
                     Discount Rate
                     (OMB)
                     Taxes Paid
                     (Limitation on Tax
                     Shield; Modeled from
                     Census data)

                     Tax Status
                     (by Assumption)
                                               Cost Annualization
                                                     Model
                                                                           Outputs
                                                                             Pre-tax
                                                         After-tax
 Present Value
of Expenditures
                                                                        Annualized
                                                                           Cost
                                         Figure 3-1

                                  Cost Annualization Model
                                           3-3

-------
       The Internal Revenue Code Section 168 classifies an investment with a lifetime of at least 20
years but less than 25 years as 15-year property. Therefore, the cost annualization model uses a 15-year
depreciable lifetime for the capital cost. A mid-year depreciation convention is used; that is, EPA
assumes that a 6-month period elapses between purchase of equipment and time of operation. As such,
the model covers a 16-year period, with a 6-month period in the first year and a 6-month period in the
sixteenth year.

       Tax rates are determined by the national average state tax rate plus the federal tax rate. Taxable
income—earnings before interest and taxes (EBIT)—was derived from Census data. EPA used the value
of each site's EBIT to determine the tax bracket for that site. Derivation of EPA's estimate of EBIT is
discussed in Section 3.1.2 below, and in more detail in Appendix B. EPA assumed that all model
facilities pay federal and state taxes at the corporate rate. EPA used its estimates of taxes to ensure that a
facility's tax shield could not be greater than the taxes it paid.

       A sample cost annualization spreadsheet is located in Appendix A of this document. Section A.3
of Appendix A details the calculations used to determine annualized costs (before and after taxes) and
present value of costs (before and after taxes).

       The cost annualization model calculates the present value of the pre- and posttax cost streams.
Then it calculates the annualized cost based on the site-specific discount rate. Thus, the model calculates
four types of compliance costs for each site: present value of expenditures (pre- and posttax) and
annualized cost (pre- and posttax). The latest year for which financial data will be available from the
detailed survey is 1999, so the model uses 1999 dollars.

       The cost annualization model's outputs feed into the other economic analyses. Pretax annualized
costs are used to project economic impacts in:

       •       The market model
       •       Facility-level income
       •       Financial ratio analysis
                                               3-4

-------
       •      Corporate financial distress analysis
              The cost-effectiveness analysis (see Appendix F; not part of economic achievability)

Posttax annualized costs are used to project economic impacts at the site level based on estimated net
income and cash flow.


        3.1.2   Facility-Level Impact Analysis

        EPA used publicly available information to project facility-level impacts under the proposed
effluent guidelines. This section briefly outlines the primary features of the methodology used for this
facility-level analysis; Appendix B provides a detailed explanation of data sources, methodology, and
assumptions used to develop the analysis. Section 3.2.2 discusses the facility-level methodology for the
final rule.

        EPA based its facility-level analysis on the U.S. Census Bureau's 1997 Economic Census of the
 following four industries: Animal (Except Poultry) Slaughtering (NAICS 311611), Meat Processed From
 Carcasses (NAICS 311612), Rendering and Meat Byproduct Processing (NAICS 311613), and Poultry
 Processing  (NAICS 311615). The Census provides detailed revenue and cost information by
 employment class, which EPA used to build model facilities.

        To analyze facility-level Impacts based on the Economic Census data, EPA compared estimated
 compliance costs with four types of income measures:

         •       Average establishment revenues
         •       Average establishment EBIT
         •       Average establishment net income
         •       Average establishment cash flow

  Each level of analysis more closely approaches the goal of using estimated compliance costs to draw
  strong inferences about definable impacts on the establishment, but each level of analysis requires
                                                3-5

-------
additional assumptions to generate the test data. Thus, each level of analysis presents a tradeoff. For
example, the relationship between facility cash flow and the impact of compliance costs is much more
clearly defined than the relationship between facility revenues and compliance cost impacts. Estimating
average facility cash flow requires more assumptions than estimating average facility revenues, however,
and that increases the uncertainty about the baseline benchmark against which impacts are measured.

        Section 3.1.2.1 provides an overview of the basic strategy EPA followed to develop its facility-
level analysis. Section 3.1.2.2 explains how EPA measured model facility income. Section 3.1.2.3
describes how EPA estimated the distribution of income for each model facility. Section 3.1.2.4 presents
a simple example of how EPA used the model to assess potential impacts of the regulation. In Section
3.1.2.5 negative baseline facility income and its implications are discussed.  Section 3.1.2.6 discusses
how EPA matched its economic model facilities to the engineering models used to estimate compliance
costs.
        3.1.2.1 Overview of Basic Model Framework

        The microeconomic basis for the model framework is that a profit-maximizing firm will shut
 down when average variable costs exceed average revenues. Economic theory states that sunk costs
 (i.e., costs attributable to past capital purchases) are irrelevant to a firm's current decision making; only
 variable costs matter in the short run. This basic microeconomic principle can be observed in modern
 corporate finance where a firm is expected to close if its cash flow (i.e., net income plus depreciation)
 turns negative. Accounting cash flow, which is primarily composed of operating costs and revenues, is
 analogous to measuring short-run variable costs. By excluding depreciation (the accounting charge for
 the utilizatiqn of previously purchased capital equipment) from the cash flow calculation, cash flow
 essentially measures current operating revenues net of current operating costs.1  Negative cash flow is
 equivalent to average variable costs exceeding average revenues where the firm is expected to close.
         1 The cash flow calculation includes interest payments. Some may argue that interest payments also reflect
 costs associated with past capital purchases and therefore should be excluded from consideration in the shut down
 analysis. However, interest payments cannot be excluded from the analysis; if the facility cannot meet its interest
 payments, it will be in default on its loan and the bank will foreclose.
                                                3-6

-------
       The model developed assesses when and to what extent a facility is impacted by regulatory costs
by measuring the facility's pre- and post-regulatory cash flows. If cash flow becomes negative after
regulatory costs are subtracted from a facility's pre-regulatory cash flow, it can be reasonably inferred
that facility closure was a result of the regulatory cost burden. Impacts of the regulation then would
include closure of a facility along with its lost output and employment. The model framework also
evaluates impacts utilizing three alternative income measures,, revenues and net income. Although cash
flow is the most appropriate measure of short-run variable costs, and hence the best predictor of facility
survival, the additional income measures act as sensitivity analyses to check for consistency in model
results.

        The basic model framework is composed of the following stages:

        •      Develop model facility income measures, including revenues, earnings before interest
               and taxes (EBIT), net income, and cash flow, for establishments of different sizes,
        •      Estimate the frequency distribution of different income measures for. the class of
               facilities represented by each model facility,
        •       Estimate the percentage of facilities with income less than estimated compliance costs
                within each model facility class, which forms the basis for employment and output
                impacts of the regulation.

 A detailed discussion of each of the above stages is provided in the following sections.
        3.1.2.2 Development of Model Facility Income Measures

        In the first step of the modeling procedure, EPA developed ,a series of model facilities for the
 industry to be analyzed. The model facilities represent establishments of different sizes within the
 industry, where facility size is measured by facility employment. The 1997 Economic Census:
 Manufacturing-Industry Series data provide detailed revenue and cost information by employment
 class that EPA primarily used to build model facilities. EPA also utilized other data sources, such as the
 Annual Survey of Manufactures (ASM), and Federal and state corporate tax rates, to estimate interest
 payments and relevant tax rates (see Appendix B for a detailed discussion of the various data sources).
 For each model facility, EPA estimates the following income measures:

               .   .•                             3-7

-------
       •      Revenues,
       •      Earnings before interest and taxes (EBIT) - used to estimate net income and cash flow,
       •      Net income, and
       •      Cash flow.

The following sections describe in more detail how model facility income measures are constructed
from the various data sources.


       Model Facility Revenues

       The Census Bureau publishes the value of total shipments by employment size for each NAICS
code, along with the number of facilities in that size class. The value of total shipments includes the
value of primary and secondary shipments as well as resale, contract, and other miscellaneous receipts.
This makes the value of total shipments a reasonable proxy for total revenues. EPA calculated average
model facility revenues by employment class within each industry as:

        •       revenues = value of total shipments / number of establishments


        Model Facility EBIT

        In order to calculate model facility net income and cash flow from model facility revenues, EPA
first estimated model facility EBIT. EPA calculated EBIT by employment class data, then estimated net
income and cash flow from EBIT using additional assumptions.

        Census provides most of the significant categories of operating costs that would be included in
EBIT. For each of the four meat product NAICS industries, facility revenues were estimated by value of
shipments. Census also provides:
                                              3-8

-------
              payroll and material costs directly attributed to the employment class level2
       •      benefits, depreciation, rent, and purchased services attributed at the industry level

EPA used a few reasonable assumptions to distribute industry-level costs to the employment class level.
For example, EPA assumed that employment benefits are proportionate to payrolFand that depreciation
is proportionate to capital expenditures. See Appendix B for more detail on similar assumptions.

       EPA calculated model facility EBIT as:

               EBIT = (Value of Shipments-Operating Costs) / Number of Facilities

where:
               Operating Costs = Payroll + Material Costs + Benefits + Depreciation + Rent +
               Purchased Services

 Because revenues, payroll, and cost of materials are the most significant components of EBIT, the
 relative error introduced by distributing industry-level data among employment classes should be small.
 For NAICS 311613 (rendering), payroll and material costs make up over 86 percent of estimated costs
 (where estimated costs equal the sum of payroll, material costs, benefits, depreciation, rent, and
 purchased services). For NAICS 311611 (slaughter), 311612 (processing),  and 311615 (poultry), payroll
 and material costs exceed 90 percent of estimated costs.
         Model FacUity Net Income and Cash Flow

         EPA then calculated net income for each employment class model facility in each industry from
 EBIT, using additional assumptions to estimate tax and interest payments. Data for these two additional
 components of net income were derived from two Census Bureau publications, Annual Survey of
 Manufactures (ASM) and Economic Census, along with the Internal Revenue Service code. Because one
         2 In addition Census provides capital expenditures and value added directly attributed to the employment
  class level.  These are not direct components of operating costs, but are used to attribute industry level components of
  cost to the employment class level.
                                                3-9

-------
must use an additional layer of assumptions, albeit reasonable ones, to estimate net income from EBIT,
the uncertainty associated with the net income estimate is greater than that for EBIT.

       Estimating tax payments is relatively straightforward. EPA assumed that establishment EBIT is
equal to business entity EBIT as the basis for calculating faxes. To estimate facility tax payments, EPA
multiplied the model facility's EBIT by the sum of the relevant federal corporate income tax rate and the
average state corporate income tax. To estimate net income, EPA subtracted the estimated tax payment
from EBIT for each model facility.

        EPA estimated interest payments using a combination of ASM data on past investment by
industry, Census data on relative investment in buildings and equipment, and assumptions about
investment behavior. EPA first scaled ASM time series data on industry investment, which is based on
Standard Industrial Classification  (SIC) codes, to represent the current NAICS meat product industries.
EPA then used the average percentages of meat product industry investment in equipment and  structures,
 as presented in the Economic Census, to divide the ASM investment time series into those two
 components.

        In estimating interest payments from the time series of past investment in equipment and
 structures, EPA assumed:

               all investment in  each year was funded through bank loans,
                the interest rate on those loans was equal to the nominal prime rate for that year plus 1
               percent, and
                the average loan period was 7 years  for equipment and 25 years for structures.

 Using these assumptions, EPA developed a time series estimate of loan payments made by the industry,
 and of the portion of each year's  loan payments accounted for by interest (e.g., using the Lotus
  ฉIPAYMT function). Total interest payments in the'baseline year equals the sum of this year's interest
 payments on the stream of past years' investment.3 Interest payments were then attributed to each
         3 For example, interest payments on equipment investment for the year 1997 would equal the sum of interest
  paid in year 25 of loans from 1973 plus the interest paid in year 24 of loans from 1974, and so on.
                                               3-10

-------
employment class based on the percentage of industry investment accounted for by that employment
class in the 1997 Gensus.

       EPA calculated net income as:

               Net Income = EBIT x (1 - Tax Rate) - Interest Payments

Then, based on net income above, cash flow is computed as

               Cash Flow = Net Income + Depreciation

 where depreciation was estimated for the calculation of model facility EBIT.

        The link between impacts measured by comparing cash flow with compliance costs is much
 stronger than the link between either EBIT or revenues and compliance costs: when post-compliance
 cash flow is negative, the facility can be reasonably projected to close. Because the estimate of cash flow
 is dependent upon a series of assumptions, however, the uncertainty concerning the accuracy of the cash
 flow measure is much greater than for revenues or EBIT. Thus, this analytic approach presents a
 tradeoff between the accuracy of the income measure and the certainty of the impacts based on that
  measure.
         3.1.2.3 Distribution of Income Represented by Model Facilities

         The objective of the model framework is not simply to examine the revenues, costs, and impacts
  on a series of model meat products facilities. The model facility reflects the average of a group of
  facilities, not a group of identical facilities. Thus, income for a given group of facilities will lie in a
  distribution around the average; some facilities will have smaller and some will have larger incomes.
  Ignoring this distribution of facility income will result in impact estimation errors. If the model facility is
  projected to remain open after incurring regulatory costs, then some facilities that it represents with
  smaller than average income may, in fact, close due to the regulation despite the model results.
  Conversely, if the model facility is projected to close as a result of regulatory costs, then some larger
                                                3-11

-------
than average facilities that it represents may in reality remain open despite the regulatory costs. To
incorporate this concept into the model framework, EPA estimated the distribution of income
represented by model facilities. By modeling a facility income distribution with known mean and,
variance, the model can project how compliance costs impact not just the model facility, but the facilities
represented by it as well.

       To estimate the distribution of income, EPA obtained special tabulations of the variances and
covariances of relevant income components for each employment class (i.e., model facility) from the
Census Bureau (U.S. Census Bureau, 2001). Combining these data along with the assumption that these
observations are normally distributed around their mean, EPA constructed cumulative probability
distributions for the four income measures, revenues, EBIT, net income, and cash flow. The following
sections describe the cumulative probability distribution constructs for the individual income measures
in further detail.
       Distribution of Revenues

       For each sector of the four NAICS codes representing the meat products industry, EPA directly
obtained the variance of the value of shipments, OR, around its mean, liR, for each model facility to
estimate the cumulative probability function of revenues. Based on the assumption of normality (i.e., XR
~N(xR, oR)), the model evaluates impacts.as the number and percentage of facilities in an employment
class for which compliance costs exceed 1 percent and 3 percent of revenues.
       Distribution of EBIT

       Although the variance of revenues (value of shipments) is directly provided by the Census
special tabulation, the variance of EBIT needs to be estimated. EBIT is a linear function of its revenue
and cost components. Thus, the variance of EBIT can be estimated using the standard statistical
                                             3-12

-------
relationship where the variance of a linear function is itself a linear function of the variance and.
covariance of its constituents.

       To estimate the distribution of EBIT for each model facility, EPA used the variance and
covariance of the value of shipments (R), payroll (P) and material costs (M) for each employment class
provided by Census. Given that mean EBIT, "XE, for an employment class is:
                                      XE   XR   XP
 where "x"; denotes the mean value of revenues, R, payroll, P, and material costs, M. EPA computed the
 variance of EBIT, oE2, as:
                          Op =  On
OM  -
                                                                2o
                                                                  pM
 where of and oy represent the variance and covariance of revenues, payroll, and material costs
 respectively (Mendenhall et al., 1990). Although payroll and material cost do not comprise all operating
 expenses included in EBIT, they do comprise the vast majority of EBIT. Hence, excluding the variance
 for the remaining components should not cause a significant error in the variance estimate.
        Distribution of Net Income and Cash Flow

        EPA estimates the variance of net income and cash flow for each model facility from its
 estimated variance for EBIT. If some scalar, a, is added to the mean of a distribution, the variance of
 that distribution will be unchanged. However, if the mean of the distribution is multiplied by some
 scalar, k, then the variance of that distribution increases by the square of k.
                                              3-13

-------
Cash flow, for example, is estimated by: (1) multiplying EDIT by (1 - tax rate), (2) subtracting interest
payments, then (3) adding depreciation. If the scalar k represents (1- tax rate), and the scalar a
represents depreciation less interest payments, then mean cash flow for a model facility is equal to:
                                        XCF  =  a + kxE
The variance of facility cash flow will be equal to:
                                           -'CF
 (Harnett, 1982). EPA used these relationships to derive the variances for net income and cash flow
 from the variance for EBIT.

 Table 3-1 presents the mean and variance EPA estimated for each model facility and income measure.
        3.1.2.4  Use of Model Facility and Distribution to Project Closure Impacts

        As discussed above, both economic and corporate finance theory predict that a firm will close if
 cash flow becomes negative. EPA's strategy for assessing facility closure impacts therefore compares
 pre-regulatory cash flow with post-regulatory cash flow; post-regulatory cash flow is calculated by
 subtracting post-tax annualized compliance costs from pre-regulatory cash flow. EPA estimated cash
 flow for a series of model facilities from Census data; moreover, EPA estimated the  distribution of cash
 flow for facilities represented by each model facility. This section provides an intuitive example of how
 EPA uses this information to project facility closures.
                                               3-14

-------
                            Table 3-1
Model Facility Income Mean and Standard Deviation by Employment Class
NAICS ••' .-•''"'•
Establishment
Employment, Size
INCOME MEASURE
Revenues
x $1,000
Net Income
x $1,000
GashFIow
x $1,000
STANDARD DEVIATION
Revenues
Netlncbme
311611
Emp. 1 to 4
Emp. 5 to 9
Emp. 10 to 19
Emp. 20 to 49
Emp. 50 to 99
Emp. 100 to 249
Emp. 250 to 499
Emp. 500 to 999
Emp. 1,000 to 2,499
Emp. >= 2,500
$439.6
$1,265.2
$2,654.6
$8,412.6
$22,489.8
$69,474.3
$160,913.7
$262,734.0
$677,948.1
$1,426,054.3
$27.7
$46.3
$64.1
$336.3
$1,303.0
$2,696.1
$4,004.8
$4,982.8
$29,321.4
$9.933.5
$32.6
$55.2
$85.6
$382.4
$1,437.6
$3,248.2
$4,713.6
$6,924.2
$33,489.1
$18,501.2
292.5
841.8
1766.4
5597.6
14964.3
46227.0
107069.3
174818.7
451095.1
948872.3
56.2
89.1
147.1
617.2
2259.7
5210.8
8024.0
10402.7
53662.4
31988.4
311612
Emp. 1 to 4
Emp. 5 to 9
Emp. 10 to 19
Emp. 20 to 49
Emp. 50 to 99
Emp. 100 to 249
Imp. 250 to 499
mp. 500 to 999 '
tnp. 1,000 to 2,499
mp.>= 2,500
$412.6
$1,393.5
$2,844.8
$7,451.6
$19,048.8
$52,075.1
$105,065.6
$172,089.3
NA
NA
$29.6
$152.2
$160.3
$462.3
$1,823.4
$4,510.3
$6,308.4
$14,363.6
. NA
NA
$40.2
$181.5
$204.4
$562.4
$2,044.6
$5,449.7
$7,555.0
$16,840.2
NA
NA
380.8
1286.1
2625.5
6877.4
17580.9
48062.0
96968.9
158827.5
NA
NA
81.4
320.5
367.4
1079.2
3819.5
9935.8
13265.6
31591.3
NA
NA
11613
mp. 1 to 4
mp. 5 to 9
mp. 10 to 19
mp. 20 to 49
mp.50to992
mp. 100 to 249
mp. 250 to 499
mp. 500 to 999
mp. 1,000 to 2,499
mp. >= 2,500
$859.9
$3,818.0
$6,475.8
$11,680.8
$17,107.8
NA
NA
NA
NA
NA
$14.1
$509.8
$608.3
$1,879.1
$2,406.5
NA
NA
NA
NA
NA
$39.9
$571.7
$730.5
$2,244.0
$3,069.3
NA
NA
NA
NA
NA
1154.9
5127.6
8697.2
15687.5
22976.2
NA
NA
NA
NA
NA
310.5
793.9
1047.2
3198.6
4476.2
NA
NA
NA
• NA
NA
Cash Flow

56.2
89.1
147.1
• 617.2
2259.7
5210.8
8024.0
10402.7
53662.4
31988.4

81.4
320.5
367.4
1079.2
3819.5
9935.8
13265.6
31591.3
NA
NA

310.5
793.9
1047.2
3198.6
4476.2
NA
NA
NA
NA
NA
                                3-15

-------
                                       Table 3-1 (cont.)
          Model Facility Income Mean and Standard Deviation by Employment Class
NAICS
Establishment
Employment Size
INCOME MEASURE
Revenues
x$l,000
Net Income
x $1,000
Cash Flow
x $1,000
STANDARD DEVIATION ,
Revenues
Net Income
311615
Emp. 1 to 4
Emp. 5 to 9
Emp. 10 to 19
Emp. 20 to 49
Emp. 50 to 99
Emp. 100 to 249
Emp. 250 to 499
Emp. 500 to 999
Emp. 1,000 to 2,499
Emp. >= 2,500
$257.9
$759.4
$3,291.5
$11,721.5
$14,880.7
$29,999.3
$71,300.2
$117,768.1
$182,579.1
$321,884.5
$6.5
$23.2
$452.9
$2,428.2
$1,462.6
$2,323.7
$3,466.3
$13,361.8
$17,044.9
$1,072.1
$18.1
$39.9
$484.5
$2,564.0
$1,618.4
$2,744.6
$4,602.5
$14,783.8
$20,179.0
$7,855.7
158.1
465.4
2017.3
7183.8
9120.0
18385.8
43698.1
72177.1
111898.1
197274.9
28.3
69.5
631.3
3265.5
' 2224.7
3966.2
5955.6
20657.6
29094.2
4551.3
Cash Flow

28.3
69.5
631.3
3265.5
2224.7
3966.2
5955.6
20657.6
29094.2
4551.3
1 Due to disclosure issues, data for 2 facilities with 1,000 < employment < 2,499, and 1 facility with 2,500
employment combined in lower category for NAICS 311612.
2 Due to disclosure issues, data for 10 facilities with 100 < employment < 249, and 1 facility with 250 <
employment < 499 combined in lower category for NAICS 311613.
Data for combined size class calculated as (total minus sum of all other size classes).
                                               3-16

-------
       .Figure 3-2 presents graphically the cumulative normal distribution function for the cash flow of
all facilities in a model class with a mean of $100,000 and a standard deviation of 100,000.4 If EPA
estimates that annualized compliance costs equal $40,000 for the model facility, then in this example the
model facility itself would not be projected to close.. However, any facilities in the same model class
with cash flow of $40,000 or less would be projected to close. Given the mean and variance of cash flow
for that model class, the probability of facilities in that class earning less than $40,000 can be readily
calculated; about 27.4 percent of facilities in this class earn cash flow less than $40,000 per year.
Multiplying that probability by the number of facilities in the class results in the projected number of
closures for that class. Multiplying that same probability by the number of employees in the employment
class estimates the projected employment impacts of those closures.

        Note that EPA actually calculates the  incremental probability of closure. That is, EPA calculates
the probability that facilities have cash flow of less than $40,000 minus the probability that facilities have
negative cash flow. EPA's methodology compares positive pre-regulatory cash flow with post-regulatory
cash flow; if pre-regulatory cash flow is positive and post-regulatory cash flow is negative, then the
 facility is projected to close. If the facility's pre-regulatory cash flow is negative, EPA cannot evaluate it.
 In the above example, the incremental probability of closure would be equal to the probability a facility's
 pre-regulatory cash flow is less than $40,000 (27.4 percent) minus the probability a facility's pre-
 regulatory cash flow is less than zero (15.9 percent), or 11.5 percent.  The issue of negative pre-
 regulatory cash flow is discussed in more detail in section 3.1.2.5 and in Appendix B.

         Similarly, EPA constructs distributions for revenues, EBIT, and net income. Although there are
 not well-defined thresholds for these facility income measures that EPA can use to project facility
 closure if exceeded by compliance costs, EPA can use these distributions to estimate, for example, the
 probability that facilities incur pre-tax annualized compliance costs exceeding 3 percent of revenues.
 This provides useful information concerning the magnitude of impacts on facilities not projected to
 close. EPA also measures impacts by estimating pre-tax compliance costs as a percentage of model
 facility EBIT.
         4 Standard deviation is equal to the square root of the variance, and provides an equivalent means of
  characterizing a distribution.
                                                3-17

-------
         ^
    •2   I
     fl
     O
a)
*-<
S3


     CO
     cซ
    PQ
00
rH
 1
CO

-------
       In summary, EPA first estimates measures of average income for a series of model facilities.
EPA then estimates the distribution of income around the average for all establishments represented by
the model.  The distribution allows EPA to project the probability that compliance costs exceed some
specified percentage of facility income for that class of facilities, and use that probability to estimate the
number of establishments incurring that impact. Without use of the probability distribution, EPA would
be required to project "all or nothing" impacts based on the average income measure for the
representative model facilities. That is, either all facilities in a class would be projected to close if
annualized compliance costs exceed model facility cash flow, or all facilities in that class would be
projected to remain open if annualized compliance costs were less than model facility cash flow.
        3.1.2.5 Negative Baseline Facility Income

        The estimated means and variances for the distribution of each model facility's income results in
 some probability greater than zero that facilities in each employment class earn negative income. There
 are three primary reasons that these distributions do show some probability of negative establishment
 income:                 .

        •       Actual establishment income is less than zero.

        •       EPA assumed the distribution of income around the model facility mean is normally
                distributed when, in fact, it may be positively skewed.

        •       EPA could not directly measure the variance of the income distributions, but instead had
                to estimate it from incomplete data.

 This section discusses these reasons, and their implications for the model.5
        5 Table B-7 in Appendix B presents the model facility mean and standard deviation for each income measure
 by employment class and NAICS code, as well as the probability that income is less than zero (based on that mean and.
 standard deviation, and assuming income is normally distributed).
                                               3-19

-------
       Actual Establishment Income is Less Than Zero

       The actual establishment financial data collected by Census on which the estimated distribution
is based might reveal negative income for two reasons:

               the parent company that owns the establishment does not assign costs and revenues that
               reflect the true financial health of the establishment. Two important examples are cost
               centers and captive sites, which exist primarily to serve other facilities under the :
                  lership.6
                                                                       : same
own
        •       the establishment is in financial trouble; that is, true costs exceed revenues.

To the extent that these types of establishments are contained in an employment class, the projection of
negative baseline income is accurate.  In either case, EPA would be unable, even with the use of facility
specific survey data, to evaluate impacts to these establishments as a result of the rule.

        Skewed Distributions

        EPA assumed the distribution of income around the model facility mean is normally distributed
 based on the "law of large numbers." However, establishment income may be positively skewed.  The
 use of a normal distribution instead of a positively skewed distribution would result in a model with a
 higher percentage of establishments having negative baseline income than would actually occur in the
 industry.
        The effects of a positively skewed income distribution can be most apparent when considering
 the distribution of establishment revenues. While it is possible, even probable, that some establishments
 earn negative income - whether measured by EDIT, net income, or cash flow - they will not earn
 negative revenues (although they may earn zero revenues). Thus, the distribution of establishment
         6 Captive sites may show revenues, but the revenues are set approximately equal to the costs of the
 operation.  Cost centers have no revenues assigned to them.
                                               3-20

-------
revenues for an employment class should show zero facilities earning negative revenues.7 If, however,
some facilities earn atypically large revenues, then the distribution may be positively skewed (e.g., the
cumulative distribution function in Figure 3-2 would initially rise more steeply than the normal
distribution, but would then flatten out - reaching a probability of 1.0 at a higher level of income).

        Using a normal, symmetric distribution to approximate a skewed distribution would likely result
in an over estimate of the percentage of establishments earning negative income. Census confirmed that
in general, the distributions of revenues, payroll, and material costs in an employment class tend to be
positively skewed (Quash, 2001). However, even if the distribution of a variable such as revenues,  .
payroll, or material costs is positively skewed, the distribution of a function of those variables (e.g.,
revenues minus payroll and material costs) will not necessarily be skewed. Thus, while there is intuitive
reason to believe the distribution of establishment income measures is skewed, the degree of skewness is
difficult to determine.

               Adjustments to Variance

        EPA used the Census special tabulation to directly calculate the variance for:

         •       value of shipments - (payroll + material costs)

 in each NAICS code and employment class.  However, the actual measures of facility income used in the
 facility level economic impact model are:

         •       EBIT = value of shipments - (payroll -4- material costs + benefits + all other costs)

         •       Net Income = EBIT x (1 - tax rate) - estimated interest payments
         7 EPA estimated the percentage of establishments that would earn negative revenues assuming revenues are
 normally distributed with mean and variance determined by the Census special tabulation. As presented in Table B-7,
 from 5 percent to 23 percent of establishments are estimated to earn negative revenues based on these assumptions.
                                                3-21

-------
       •      Cash Flow = Net Income + depreciation


Because the actual income measures (EBIT, net income and cash flow) differed from the approximate
income measure  [value of shipments - (payroll + material costs)] on which variance was estimated, EPA

adjusted the variance of [value of shipments - (payroll + material costs)] associated with each of the

actual income measures used in the model.


       To adjust income variance, EPA used standard rules concerning the expected value of mean and

variance. Intuitively, if one multiplies the mean of a distribution by some scalar k, the variance of that

distribution expands or shrinks by the square of that scalar value.  However, if instead of scaling the
mean, its value is changed by adding'or subtracting some constant, then the distribution shifts to the right

or left on its x-axis, but its variance does not change.


        Applying these rules to the mean and variance for the various measures of income for Meat

Products model facilities yields the following results (see Appendix B for details):


        •       mean EBIT is smaller than the mean of [value of shipments - (payroll + material costs)],
               but the variance for EBIT equals the variance for [value of shipments - (payroll +
               material costs)]; the probability of negative EBIT is larger than the probability of
               negative [value of shipments - (payroll + material costs)].

        •      both the mean and the variance of net income are proportionate to the mean and
               variance of EBIT  ; the probability of negative net income equals the probability of
               negative net income.

        •      mean cash  flow is larger than mean net income, but the variance of cash flow equals the
               variance of net income; the probability of negative cash flow is smaller than the
               probability of negative net income.


 The probability that the [value of shipments - (payroll + material costs)] is less than zero in the four Meat

 Products NAICS codes ranges from 22 percent to 26 percent, while the probability EBIT and net income

 are less than zero generally ranges, in most cases, from 26 percent to 30 percent.  The probability that •
                                              3-22

-------
cash flow is less than zero tends to be about 3 percent to 5 percent smaller than the probability that net
income is less than zero.

       Effect on Modeling Impacts

       The effects of this issue on EPA's projection of economic impacts cannot be generalized. In any
model class with a given mean income, incremental closure impacts may be overestimated or
underestimated by a cumulative distribution function with too high a probability of negative baseline
income.  If compliance costs are small relative to mean income, the model will tend to overestimate
closures, as compliance costs increase relative to mean income the model will, at some point, start to
underestimate incremental closure impacts. Within the range of income and estimated complia
relevant for this analysis, the difference in incremental closures tends to be small.8
iiance costs
        3.1.2.6 Matching Economic Model Facilities to Engineering Model Facilities

        In addition to economic model facilities, EPA developed engineering model facilities in order to
 estimate compliance costs.  EPA estimated engineering model facility effluent loads based on data such
 as: wastewater samples, the type and level of facility production, and wastewater treatment typical
 facilities currently have in place. EPA then estimated the cost of technologies that, if purchased and
 installed, would enable the model facility to meet specified effluent guidelines (see the Development
 Document, U.S. EPA, 2002, for details).               •                                     '

         Because data to develop engineering model facilities and economic model, facilities had to be
  drawn from diverse sources, EPA then had to match its engineering model facilities with its economic
         8 EPA performed a sensitivity analysis to determine the importance of this issue. EPA projected closure
  impacts using the variance of model establishment income as estimated above and compared the results to those from
  a model with an identical mean income, but a smaller variance and a much smaller probability of negative baseline
  income (about 7 percent). For the relevant range of income and compliance costs, the difference between the two
  results was not significant. Details of this sensitivity analysis are included in Appendix E.
                                                 3-23

-------
model facilities in order to project the financial impacts of the proposed effluent guidelines.9 This

section describes how EPA matched the economic model facilities to the engineering model facilities in

order to project economic and financial impacts.



        Basis for Engineering Model Facility  Classes



        In order to develop a comprehensive series of representative engineering model facilities, EPA

classified the meat products industry based on the type of meat produced at the facility:



               •       Red meat (primarily beef and pork),


               •       Poultry (primarily chicken and turkey),


                •       Mixed (both red meat and poultry), or


                •       Rendering products or meat byproducts (either red meat or poultry);


 the type of processes performed at facilities:



                •       First processing (slaughter),


                •       Further processing, and/or


                •       Rendering (the process resulting in meat byproducts), and


 facility size (small, medium, large or very large) based on production and wastewater flow.
         9 For the economic analysis of the final regulation, EPA will be able to use Section 308 survey data. This
  data enables EPA to directly determine the financial characteristics of the facilities used to estimate engineering
  compliance costs.
                                                 3-24

-------
       This results in a set of model facility classes reflecting different combinations of the
characteristics listed above. For example, a model facility might be classified as a large poultry facility
(based on production and meat type) that performs one of the following six process combinations: first
processing, (2) further processing, (3) first and further processing, (4) first processing and rendering, (5)
further processing and rendering, or (6) first processing, further processing, and rendering.

        Matching Engineering and Economic Model Facility Classes

        EPA matched its economic model facilities to the engineering model facilities on the basis of two
characteristics: (1) the relationship between production process, meat type, and NAICS industry, and (2)
the relationship between facility production and revenues.

        The Census Bureau classifies the meat product industry into four groups. Census distinguishes
red meat facilities (either beef or pork) that perform animal slaughter (first processing), whether alone or
in combination with other processes (NAICS 311611) from red meat facilities that perform further
processing '(with or without rendering), but no slaughtering activities (NAICS 311612). Census classifies
 all facilities that perform poultry slaughter, poultry further processing, or both (with or without
 rendering), in the same NAICS code (311615). Finally, facilities that perform rendering, but no other
 processing activities, are classified in NAICS 311613 by Census.

         Thus, model economic facilities were matched to the model engineering facilities, based on
 production, as follows:
                Engineering facilities that process either beef or pork, and perform first processing
                (alone or in combination with further processing and/or rendering) were assigned an
                economic model facility from NAICS 311611.

                Engineering facilities that process either beef or pork, and do not perform any first
                processing, were assigned an economic model facility from NAICS 311612.

                Engineering facilities that perform any combination of processes on chicken or turkey,
                were assigned an economic model facility from NAICS 311615.
                                                3-25

-------
              Engineering facilities that perform rendering—whether red meat or poultry—but no
              other processes were assigned an economic model facility from NAICS 311613.

              Engineering facilities that process both red meat and poultry were assigned an economic
              model facility from NAICS 311612.10

All model engineering facilities were assigned an economic model from one NAICS code only.

       Because of data availability, economic model facilities were sized by employment class, while
engineering cost models were sized by production and flow. EPA classified engineering models into
small, medium, large, or very large based on examination of production and flow characteristics of
facilities. After determining the appropriate size for each engineering cost model facility, EPA calculated
the median production for all facilities in that class based on screener survey data. EPA then combined
median production data for the engineering model facilities with meat product indicator prices to
estimate revenues for each engineering model facility. These estimated revenues were then compared
with average revenues for each economic model facility. EPA then selected the appropriate employment
class for the economic model facility based on the closest revenue match within the NAICS code
assigned to each meat type and process combination.11  For more detail on matching economic model
facilities to engineering model facilities, see Appendix B.

        3.13  Financial Ratio Analysis

        EPA also examined the impact of the proposed effluent guidelines on the model establishment's
 balance sheet as well as its income statement. EPA performed two analyses of balance sheet impacts.
 First, EPA examined the effect of compliance costs on model establishment return on assets (ROA).
 Second, EPA examined if compliance costs cause corporate financial distress for a select group of firms.
 EPA selected the Altaian Z' score as its means of measuring financial distress.  For reasons stated below,
         10 No mixed meat model facilities estimated to incur costs were found to perform slaughter activities. ^
         11 EPA used the baseline prices from the market model as the indicator prices for the meat products (for
 more detail on the market model see Section 3.1.4.2).
                                               3-26

-------
EPA would prefer to use the Altman Z' score as the sole measure of financial distress for this industry's
effluent guidelines.  However, EPA cannot construct balance sheet statements for its economic model
facilities due to lack of appropriate data.  Therefore, a limited analysis of projected impacts on model
facility ROA is utilized. For many large, multi-establishment companies, including  many of those listed
in Section 2.5  of the industry profile, EPA was able to obtain sufficient financial data to perform the
Altman Z' analysis. For the final rule, EPA will use the Altman Z' score measure of financial health to
assess impacts to all affected firms based on Section 308 survey data. Section 3.1.3.1 describes the ROA
analysis, while section 3.1.3.2 presents Altman Z' score methodology.
        3.1.3.1  Return on Assets

        EPA selected return on assets (ROA) as perhaps the single best financial ratio to indicate facility
 profitability. ROA provides a reflection of the opportunity cost of investing in the meat product industry.
 Investors look for their best opportunity to receive a high rate of return on their capital. If the estimated
 compliance costs of the proposed effluent guidelines are projected to significantly lower the rate of
 return earned in the meat products industry, investors may exit that market in search of better
 opportunities; the meat products industry would then be likely to contract.

         EPA obtained data on ROA for SIC codes in the.meat products industry from Dun &
 Bradstreet's Industry Norms and Key Business Ratios, 1997-1998 (D&B, 1998). D&B provides the
  median, upper quartile, and lower quartile ratio for companies submitting financial data in each SIC
  code.12 Therefore, these data are not obtained from a representative sample and must be interpreted with
  care. D&B did not provide data for the rendering industry; EPA used the lowest median ROA ratio from
  among the other meat product industries as a conservative proxy  for the rendering ROA ratio.
          12 The relationship between NAICS and SIC codes is presented in Section 2.1.
                                                3-27

-------
       To project impacts on model facility ROA, EPA first used the median ROA for the appropriate
industry, combined with each model facility's net income, to estimate model facility total assets. ROA
equals net income divided by total assets. Therefore, EPA calculated:
                          , _  ...      ,          model facility net income
                     model facility total assets =  ——
                                                       median ROA
Given each model facility's total assets, net income, and compliance costs, EPA then calculated:
                                   (net income - posttax annualized compliance costs)
           post-regulatory ROA = ^—ป—    *                 —
           r      e                           (total assets  +  capital costs)
In addition to baseline and post-compliance ROA, EPA calculated the percent change in ROA as an
impact of the proposed rule.

        In past effluent guidelines, EPA has typically considered a firm impacted if its post-compliance
ROA falls below the lower quartile ROA for the industry. Because EPA has information only on the
distribution of income (not, that is, on the distribution of total assets), it did not try to estimate the
probability that the post-regulatory ROA will fall below the lower quartile value. EPA does, however,
present each industry's lower quartile ROA for informational purposes.
        3.1.3.2 Corporate Financial Distress Analysis

        The corporate financial distress analysis examines whether a company can afford the aggregate
 costs of upgrading all of its sites. EPA has chosen to use a weighted average of financial ratios to
 examine the impacts of increased pollution control on companies. Many banks use financial ratio
 analysis to assess the credit worthiness of a potential borrower. If regulatory costs cause a company's
                                               3-28

-------
 financial ratios to move into an unfavorable range, the company will find it more difficult to borrow
 money. EPA will consider a company in such a condition to be in financial distress.
 Financial ratios are calculated at the business entity or corporate parent level because:
                Accounting procedures maintain complete financial statements (balance sheet and
                income statement) at the business entity or corporate level, but not necessarily at the site
                level. The survey data indicate that many companies do not keep complete financial
                statements at the site level.

                Significant financial decisions, such as expansion of a site's capacity, are typically made
                or approved at the corporate level.

                The business entity (or corporate parent) is the legal entity responsible for repayment of
                a loan. The lending institution evaluates the credit worthiness of the-business entity, not
                the site.
        First, EPA describes the Altman Z' score, a weighted average of financial ratios used to assess
 financial distress. EPA then summarizes the data and methodology used for the analysis. Finally, the
 implications of a score below the cutoff are discussed.

        Altman Z' score

        EPA performed a literature search to review bankruptcy prediction literature from 1990 to 1998
 (Kaplan, 1999). Although new approaches have been developed (e.g., neural networks, logit models,
 and multiple discriminant analyses), there no clearly superior method and no consensus on what is the
best approach. EPA has determined that—given the goal of selecting a methodologically sound,
reproducible, and defensible approach—the Altman Z-score, a multiple discriminant analysis (e.g., a
weighted-average) of financial ratios, is appropriate.
                                              3-29

-------
        The Altaian Z-score is a well accepted standard technique of financial analysis with nearly two
decades of use (see Brealey and Meyers, 1996, and Brigham and Gapenski, 1997). The Z-score has

advantages over consideration of an individual ratio or a collection of individual financial ratios:
               It is a simultaneous consideration of liquidity, leverage, profitability, and asset
               management. It addresses the problem of how to interpret the data when some financial
               ratios look "good" while other ratios look "bad."

               There are defined threshold or cut-off values for classifying firms as in good,
               indeterminate, and poor financial health. "Rules of thumb" are available for some
               financial ratios, such as current ratio and times interest earned, but these frequently vary
               with the industry (U.S. EPA, 1995).
        Altaian (1993) developed several variations on the multidiscriminant function. EPA selected the
Z'score because it was developed to evaluate public and private manufacturing firms. The model is:



                       Z' = 0.717Xi + 0.847X2 + 3.107X3 + 0.420X4 + 0.998X5


where the pre-compliance components are:
        Z'

        X,

        X2

        X3

        X4

        X5
overall index

working capital/total assets

retained earnings/total assets

EBIT/total assets

book value of equity (or net worth)/total debt

sales/total assets
The detailed survey requested each piece of information for the analysis. (Working capital is equal to
current assets less current liabilities). Book value of equity is also called net worth (i.e., total assets

minus total debt). Total debt is the sum of current and non-current liabilities.
                                              3-30

-------
        EPA estimates financial distress based on changes in the Altman Z'score as a result of pollution
 control compliance costs. The estimates of post-compliance scores are calculated as follows:
        Z'
        X,.
        X2
        X3
        X4
        X5
overall index
working capital/(total assets + capital costs)
retained earnings/(total assets + capital costs)
(EBIT - pretax annualized compliance costs)/(total assets + capital costs)
book value of equity (or net worth)/(total debt + capital costs)
sales/(total assets + capital costs)
        Capital costs are those developed by the engineering staff for use in the cost annualization
 model. The annualized pollution control costs for each option were calculated from the engineering
 estimates of capital and operating and maintenance costs in the cost annualization model
 (see Appendix A).

        Taken individually, each of the ratios given above (X, through X5) is higher for firms in good
 financial condition and lower for firms in poor financial condition. Consequently, the greater a firm's
 distress potential, the lower its discriminant score. The thresholds for evaluating financial distress are:

        •       Altman Z'score below 1.23: financial distress is likely

        •       Altman Z'score above 2.9 indicates that financial distress is unlikely.

        •       Altman Z'scores between 1.23 and 2.9 are indeterminate.

As can be observed from the components of the post-compliance Z'score, incremental compliance costs
will lower a company's score. EPA examines a firm's pre- and post-compliance score to determine if it
crosses one of the thresholds.
                                              3-31

-------
        Data and Methodology

        EPA performed a preliminary Altman Z' analysis based on a partial database of detailed survey
responses to questions on the income statement and balance sheet at the corporate level, information
presented in the industry profile, and estimated facility level compliance costs.

        EPA first identified major meat product companies contained in both the industry profile listing
of largest companies (Section 2.5) and the preliminary detailed survey database. For this analysis, EPA
identified 20 major multi-facility meat product companies with sufficiently consistent data on which to
perform this preliminary Altman Z' analysis. EPA used data presented in the industry profile to
determine the number of facilities owned by each of the 20 companies listed above. In general, EPA did
not have sufficient information to further classify facilities by operation, size, or discharge type.

        For compliance costs, EPA used an average of estimated compliance costs by meat type (i.e., red
meat, poultry, rendering) weighted by the median production for each process combination, size and
discharge type. Thus each company in the analysis incurred an average cost that reflected in some
measure, costs for slaughter, further processing, and rendering operations, a range of sizes, and both
direct and indirect dischargers.
        Implications of a Z' score Below the Financial Distress Threshold

        What does it mean if a company's Altman Z'score falls below the threshold for "distress likely"?

        First, it should be noted that Altman used the phrase "bankruptcy likely" as well as "distress."
This does not mean, however, that a company will immediately declare bankruptcy once its score falls
into that danger zone. It is a warning flag. A company has the opportunity to change its behavior during
this warning period to avoid the projected bankruptcy. The Chrysler Corporation is an example; Altman
(1993) cites other examples.
                                              3-32

-------
       Second, taking Chapter 11 (bankruptcy) is not the same as taking Chapter 7 (liquidation). A
company that takes Chapter 11 is protected .from its creditors for a period of time while it reorganizes
itself. A company can continue to operate while it is in Chapter 11, and has the chance to emerge from
bankruptcy. In contrast, a firm is liquidated when there is no hope for rehabilitation. Altman notes,
"Economically, liquidation is justified when the value of the assets sold individually exceeds the
capitalized value of the assets in the marketplace" (Altman, 1993, p. 33).

       Third, other forms of response are possible. Shedding non-productive assets, merging with
another company, or being purchased by another company are all possible responses  to financial   .
distress.

        What this means for the economic analysis is that:
               A company that moves into the "distress likely" category as a result of added pollution
               control costs is considered to be distressed as a result of the regulation. It does not mean
               that EPA expects the company to liquidate immediately upon promulgation. The
               company, however, will have to change the way it operates to respond to the regulation
               and remain out of bankruptcy. In either case, EPA expects serious economic disruption
               for the firm.
               A company in the "distress likely" category before the rulemaking cannot be evaluated
               for a change in status. It does not mean that EPA expects the company to liquidate in the
               very near future.
         3.1.4   Market Model

         EPA developed a market model to examine the impacts of the meat products industry effluent
  guidelines on the price and output of various meat products. The distinguishing feature of EPA's market
  model is that it explicitly incorporates cross-market impacts among meat types into the analysis. The
  demand for meat products such as beef, pork, broilers, and turkey is closely related; an increase in the
  price of pork, for example, may cause a fall in the quantity of pork demanded and an increase in demand
  for beef.
                                               3-33

-------
       In the context of EPA's proposed effluent guidelines for the meat products industry, this
increases the complexity of the market analysis. Because EPA's proposed effluent guidelines may
simultaneously affect the price of beef, pork, chicken, and turkey, the market analysis for each product
depends not only on the compliance costs for that product but on the impact of compliance on the prices
of the other three meat products.

       For example, if the proposed effluent guidelines impose compliance costs on the producers of
beef products, then the  supply of beef products will tend to decrease (i.e., the supply curve for beef will
shift to the left; a smaller quantity of beef will be offered for sale at the current price). If all other things
remain constant, this would tend to increase the price of beef products while decreasing the quantity
sold. However, EPA's proposed effluent guidelines may also impose compliance costs on pork
producers, tending to increase the price of pork. All other things being constant, the  increase in the price
of pork would increase the demand for beef products; the demand curve for beef will shift to the right.
This would tend to increase the price of beef as well as increase the quantity of beef sold. The final
impact on the price and output of beef products will depend on the relative magnitude of supply and
demand shifts. Figure 3-3 illustrates the general rule behind this example.

       If all meat products incur relatively similar per-unit compliance costs, cross-market impacts
would tend to be roughly offsetting. However, if per-unit compliance costs are asymmetric (e.g., per-unit
compliance costs are significaptly larger for some subcategories than for others), then potentially
significant shifts could  occur between meat product markets. EPA's model was developed with the
flexibility to analyze the latter situation as well as the former.

       In order to incorporate both cross-market effects and international trade into the model, EPA
specified linear supply  and demand equations in  each market to make the model tractable. The slopes of
the equations were derived from estimated price  elasticities of supply and demand found in existing
research. These elasticities were then  converted to slopes at the baseline equilibrium price and quantity.
Because domestic supply,  domestic demand, import supply, and export demand are all specified as linear
functions, the model components are  additive, and simultaneous equilibrium can be solved for multiple
markets using linear algebra.
                                              3-34

-------
                                                  Decrease in supply of
                                                  Meat Product i caused by
                                                  ELG on Meat Product i
ppost
RP
re
                                                                              s2
                                                                                  s1
Increase in demand tor
Meat Product i caused by
ELG on Meat Product j
                                                               D2
                                                               D1
                                    Qpost   Qpre
                                                                            Q,
            D1, S1 = preregulatory market supply and demand conditions
            D2', S2 = postregulatory market supply and demand conditions
            ppre> Qpre = prerequlatory equilibrium price and quantity
               ^ Qpost = postrequlatory equilibrium price and quantity
                                          Figure 3-3

                   Impact of the Effluent Guideline on Market for Meat Product i
                                            3-35

-------

analyses of both the proposed and the fmalrules.







       3.1.4.1  Market Model Approach
 in Appendix C.
       cWes in the U

                                                                                 .,. production
   an excess
            demand function for each meat product.
                                                3-36

-------
convenient way to solve the equations simultaneously. Given pre-regulatory prices, quantities, and price
parameters, linear algebra is used to solve for the pre-regulatory intercept for all four excess demand
equations.

        Third, the supply curve shift for each meat product is calculated (imposing effluent guidelines
on the industry causes the supply curve for each meat product to shift.) The supply curve shift for a
meat product is estimated as a function of average per-unit compliance costs for that product. Once the
post-regulatory (i.e., post-shift) supply curve is estimated, the excess demand equation for each meat
product is  re-written.

      . Fourth, the post-regulatory excess demand equations for all four meat products—like the pre-
 regulatory equations—are expressed in matrix form. The post-regulatory intercept for each excess
 demand equation, however, is already known: it is a function of the pre-regulatory intercept, per-unit
 compliance costs, and the supply equation price parameter. By using linear algebra to invert the matrix
 containing the price parameters, then multiplying the post-regulatory intercept vector by that inverted
 matrix, EPA can evaluate the set of meat prices that results in simultaneous equilibrium for all four meat
 products.

         Finally,  the individual component equations for each meat product's domestic supply, domestic
 demand, import supply, and export demand are evaluated using the post-regulatory prices to solve for
 post-regulatory quantities. Changes in these four quantities for each meat product, as well as changes in
 the price of each meat product, measure the market-level impacts of a meat products  effluent guidelines.
                                                3-37

-------
3.1.4.2
            Data Sources for Market Model Analysis
     Baseline Market Quantities and Prices
      SiซปaซOT
  00,,00,-s
                                                           md boneless product, it is
Futten,oK, a,tbough Putn^ and Arouse present *ป. ป                    ^
dilecfly derived tern essenUaUy to same carcass Welght daซ as pre


reflect true retail sales.
  for use as the baseline prices.
         Compliance Costs
   .   ly d. sumof an^ed comp,iance







    than Outlook's data.
                                                3-38

-------
       EPA initially estimated compliance costs by process (first, further, and rendering) within general
meat type categories (e.g., red meat and poultry; see Section 3.1.2.3 for details). This meant that EPA had
to attribute (1) estimated compliance costs for red meat to beef and pork and (2) estimated compliance
costs for poultry to chicken and turkey. To do this, EPA first estimated total annualized compliance costs
for each subcategory and size class (e.g., red meat, further processors, medium size). Then, for each
subcategory size class, EPA calculated the quantity and percent of total meat production accounted for by
teach meat type (beef, pork, chicken, and turkey). Costs were attributed by the percent each meat type
made up of total meat production for that subcategory size class (e.g., if red meat/further processors,
medium sized facilities produced 70 percent beef, 70 percent of annualized compliance costs for that
 subcategory size class would be attributed to beef). Per-unit costs were estimated by dividing the
 attributed compliance costs for each meat type by the quantity of that meat type produced.

         To determine the average per-unit compliance costs for each meat type over all subcategories
 and size classes, EPA calculated a weighted average of the per-unit costs for each subcategory and size
 class by meat type. The weights were calculated as the meat type output within each subcategory and size
 class expressed as a percent of total output of that meat type over all subcategories and size classes. (Note
 that, to an estimation of market-level compliance costs per unit, the distinction between direct and
 indirect dischargers is irrelevant.) Finally, to estimate market-level impacts, EPA entered average per-unit
 compliance costs by meat type directly into the market model.

         Price Elasticities of Demand

         Domestic price elasticities of demand are widely available from a variety of sources, including
  USDA and academic research. The results of the literature search for demand elasticities is documented
  in the record. For use in its market model, EPA selected K. S. Huang's A Complete System of U.S.
  Demand for Food (1993).

         The advantage of Huang's estimates is that they were generated in a single, coherent, consistent
   framework that satisfies theoretical constraints of symmetry, homogeneity, and Engel aggregation. This
   should make using them better than selecting individual elasticities from among several sources with
                                                 3-39

-------
varying methodologies, degrees of aggregation, and time horizons..The internal consistency of Huang's
work is of particular importance because EPA is modeling cross-product impacts in the market model.
The own- and cross-price elasticities of demand are presented in Table 3-2.
        Price Elasticities of Supply

        EPA undertook a literature search for estimates of the price elasticities of meat supply for both
the feedlots and meat products effluent guidelines. This search resulted in a wide range of estimated
elasticities with little apparent consensus.

        Because of this lack of consensus, EPA chose to use the elasticities from the effluent guidelines
for concentrated animal feeding operations (CAFOs). These were selected with the concurrence of
EPA's expert consultants (U.S. EPA, 2001). It is reasonable to use these elasticities for the meat products
market model, because meat (hi the form of both live animals for slaughter and meat products) makes
up the majority of material costs in the meat products industry (79 percent in animal slaughtering, 63
percent in meat processing, and 76 percent in poultry (U.S. Census Bureau, 1999a through 1999d). In
 addition, the other major cost component of meat production is unskilled labor, and the price elasticity
 of primarily unskilled supply tends to be large. Thus, the CAFOs supply elasticities should represent a
 reasonable lower-bound estimate for the price elasticity of meat supply. The supply elasticities selected
 for use in the model are presented in Table 3-2.
         Import and Export Elasticities With Respect to Domestic Price

         EPA used an Armington-type specification to model the effects of international trade on U.S.
 meat products markets. If foreign-produced and domestically produced goods are perceived as perfect
 substitutes for each other—that is, if consumers do not differentiate between foreign and domestically
 produced goods—then one would expect a country to either import those goods or export them, but not
                                               3-40

-------


-------
to both import and export them simultaneously. However, if consumers perceive that foreign and
domestically produced goods in a particular class are close but not perfect substitutes, then their country
may import and export that class of products simultaneously. The U.S. both imports and exports meat
products; the Armington specification that EPA selected incorporates product differentiation in the meat
products industry market model.

        Econometrically, the Armington model measures the degree of substitutability between traded
products. This is expressed as the percentage change in market share of the imported product relative to
the domestically produced good caused by a change in the relative prices of the imported and domestic
goods. An elasticity of zero implies that consumers will not substitute imported meat products for
domestic meat products; the higher the elasticity, the more willing consumers are to make this
substitution. This means that if the elasticity of substitution is equal to one, then market shares remain
constant; if this elasticity is greater than one, then an increase in U.S. price means that U.S. market share
will decrease (Armington, 1969a).

        The Armington elasticity of substitution cannot be directly  used in EPA's market model.
However, Armington demonstrated that own price and cross price trade elasticities are a function of
 domestic demand elasticities, market shares of domestic and foreign products, and the value of the
 elasticity of substitution (Armington, 1969a,  1969b). EPA used Armington's results to derive formulae
 for the trade elasticities used in its market model.14

        The U.S. elasticity of demand for imports of each meat product with respect to the U.S. domestic
 price of that product is a function of its domestic elasticity of demand, the ratio of "rest of world"
 (ROW) and U.S. market shares (EPA assumed for simplicity that there are only two countries, the U.S.
 and the ROW), and the elasticity of substitution between U.S. and ROW meat products. The value of the
 import price elasticity is positive: that is, an increase in the U.S. domestic price of meat products is
 expected to increase U.S. demand for ROW meat products.
         14 Further details of this derivation may be found in Appendix C and the rulemaking record.
                                               3-42

-------
       EPA calculated the elasticity of ROW demand for U.S. meat products with respect to U.S. price
in a similar fashion. The value of this elasticity is negative: an increase in U.S. domestic meat price wdl
decrease U.S. exports of meat product, Due to a lack of data availability, EPA calculated a numerical
value for this elasticity assuming that: (1) the ROW elasticity of substitution for U.S. meat products U
identical to the U.S. elasticity of substitution for ROW meat products, and (2) the elasticity of ROW
demand for meat products with respect to ROW price equals the elasticity of U.S. demand for meat
products with respect to U.S. price.15

        Market shares of meat production were estimated at the carcass weight level of aggregation using
 quantity data from the United Nations Food and Agriculture Organization (UN FAO). Long-run
 Armington elasticities were obtained from Gallaway etal. (2000).- Table 3-3 presents a summary of the
 trade parameters and elasticities with respect to changes in domestic, price that were used in the model.
 Note that, in general, the elasticities of meat imports are relatively low; this is because meat imports mak.
  up a small share of the U.S. domestic market.
         3.1.5   National Direct and Indirect Impacts

         Impacts on the meat product industry are known as direct effects, impacts that continue to
  resonate through the economy are known as indirect effects (effects on input industries), and effects on
  consumer demand are known as induced effects. The U.S. Department of Commerce's Bureau of
  Economic Analysis (BEA) tracks these effects both nationally and regionally in massive "mput-output"
  tables, published as the Regional Input-Output Model (RIMS II) multipliers. For every dollar m a
   "spending" industry, these tables identify the  portion spent in contributing, or "vendor," mdustnes.
    substitution for each product described by a code.
                                                  3-43

-------
       For this analysis, EPA calculated direct and indirect impacts using the national-level final-
demand multipliers for BEA industries 14.0103 (meat packing plants, sausages, and other prepared
meats):

       •      Output: 4.9661 dollars of total output per dollar of meat products

       •      Employment: 46.9297 FTEs per $1 million in output in 1992 dollars

and these multipliers for BEA 14.0105, poultry slaughtering and processing:

       •      Output: 4.3518 dollars of total output per dollar of meat products
       •      Employment: 45.1800 FTEs per $1 million in output in 1992 dollars

Note that because employment multipliers are based on 1992 data, the value of lost output needs to be
deflated to 1992 dollars before estimating employment impacts. (U.S. DOC,  1996). EPA used Gross
                                        •
Domestic Product (GDP) data by industry for the years 1947 to 2000, compiled by the Bureau of
Economic Analysis (BEA), to calculate the implicit price deflator for the Food and Kindred Products
industry in the period 1992 to 1999 (U.S. DOC, 2001).
 3.2    METHODOLOGY FOR THE FINAL RULE

        Much of the methodology for the final rule follows the same principles as the methodology for
 the proposed rule, but uses site-specific data obtained from the detailed survey. Thus, the cost
 annualization model is essentially identical for both the proposed rule and the final rule. For the
 proposed rule, however, the model used general industry average data obtained from publicly available
 sources for certain key parameters. For the final rule, the model will use facility-specific values from
 survey data for those parameters. Similarly, the facility-level impact model compares facility income
 with estimated compliance costs. For the proposed rule, facility income was measured as an average
                                              3-44

-------
based on Census data, while for the final rule, facility income will be measured directly from detailed
survey data.                                                                    .

        Section 3.2.1 explains how the use of survey data will change the cost annualization model as,
used for the final rule. Section 3.2.2 presents the methodology for projecting facility closure impacts for
the final rule. The corporate financial distress analysis, market model, and the national and community
impact methodologies will essentially be unchanged from the proposed to the final rule except where
appropriate, data from the Section 308 detailed survey will be used.
        3;2.1   Cost Annualization Model

        The cost annualization model for the final rule has essentially the same structure as the model
  used for the proposed rule. However, certain inputs for the final rule's model-such as facility income .
  and the discount rate-will be based on facility-specific data from the Section 308 detailed survey
  instead of averages from publicly available information sources. Inputs to the cost annualization model
  will come from three sources: EPA's engineering staff, secondary data, and the 2001 Meat Products
  Industry Survey. The capital and O&M costs for incremental pollution control were developed by EPA's
  engineering staff. Differences in the methodologies for developing engineering costs for the proposed
  rule and for the final rule are discussed in the Development Document (U.S. EPA, 2002).

         As with the proposed rule, EPA will use the MACRS as the depreciation method in the cost
   annualization model. Secondary data will provide the average inflation rate from 1987 to 1999 as
   measured by the Gross Domestic Product (GDP) Price Deflator. EPA will use the average inflation rate
   to convert the nominal discount rate to the real discount rate. To determine tax rates, EPA will add the
   national average state tax rate and the federal tax rate.

          The 2001 EPA survey data provide discount rates or interest rates (the weighted average cost of
   capital or the interest rate supplied by the site) for survey sites. For any site that supplied neither a
                                                 3-45

-------
discount rate nor an interest rate, EPA will use the median discount rate of all sites. Figures for taxable
income—EBIT—will also come from the EPA survey. The value of EBIT for each site will determine
mat site's tax bracket. EPA will calculate average taxes paid from its survey data, using taxes for the
years 19??, 1998, and 1999. These numbers will be used to ensure that a site's tax shield cannot be
greater than the average taxes that site paid in 1997, 1998, and 1999. Tax shields will be estimated
according to corporate structure. In the model, a "C" corporation pays federal and state taxes at the
corporate rate, an "S" corporation or a limited liability corporation (LLC) pays taxes at the individual rate
(since EPA has no way of determining how many individuals receive earnings or those individuals' tax
rates, these rates are set to zero), and all other entities pay taxes at the individual rate.

        A sample cost annualization spreadsheet is located in Appendix A of this document. Section A.3
 of Appendix A details the calculations used to determine annualized costs.(before and after taxes) and
 present value of costs  (before and after taxes).

        The cost annualization model calculates the present value of the pre- and posttax cost streams.
 Then it calculates the  annualized cost based on the site-specific discount rate. Thus, as in proposal, the
 model will calculate four types of compliance costs for each site:  present value of expenditures (pre- and
 posttax) and annualized cost (pre- and posttax). The latest year for which financial data is available is
 1999, hence, the model will use 1999 dollars.
         3.2.2  Facility Closure Model

         EPA has developed a financial model based on facility specific data from the detailed
  questionnaire to estimate whether the additional costs of complying with the proposed regulation will
  make a site unprofitable. Sites designated as unprofitable are projected to close as a result of the
  regulation, leading to site-level impacts such as losses in employment and revenue. Hence, the site
  financial model is also called the closure model within this report.  In essence, this model will perform
  the same type of analysis for the final rule as the facility level model performs for the proposed rule.
                                                3-46

-------
The difference is that the facility level model for the proposed rule is based on averages of aggregate
industry data, while the closure model is based on facility specific data.

       In terms of perspective, the closure model focuses on individual sites. It attempts to answer the
question "does it make financial sense to upgrade this site?" using data and methodology available to
corporate financial analysts. The closure model interacts with the market model (see Section 3.1.4); the
industry proportion of costs that meat processors passes through to their customers via price, increases is
derived from the market model. EPA performs its primary analysis of facility level impacts assuming
that firms can pass zero percent of costs on to customers in the form of higher prices; impacts will be
more severe under this scenario. However, unless the demand for meat products is perfectly price elastic
 (or supply perfectly price inelastic), some percentage of increased production costs due to the proposed
 rule will tend to be passed through to customers. This is the point of interaction between the closure
 model and the market model.

         In contrast, the corporate financial distress model evaluates whether a company could afford to
 upgrade all of its facilities (see Section 3.2.3). In other words, each model provides a different
 perspective on the industry and the impacts potentially caused by the effluent limitations guidelines
 requirements.

         The closure model turns the question "does it make sense to upgrade this site?" into a
 comparison of future facility income with and without the regulation. The closure decision is modeled
         Post-regulatory status   =  Present value of future earnings
                                       (Present value of after-tax incremental pollution control costs
                                       * (1-percent cost pass-through)
                                                 3-47

-------
The site closure model calculates the long-term effects on earnings reduced by the added pollution
control costs. If the post-regulatory status is less than zero, it does not make economic sense for the site
owner to upgrade the site. Under these circumstances, the site is projected to close.17

        Although simple in concept, the model incorporates numerous choices, including:

        •      Whether or not to include salvage value.

        •      Net income or cash flow as the basis of projecting future earnings.

        •      Time frame for consideration.
        Section 3.2.2.1 reviews the choices EPA has made in these three areas for the site closure model.
 Section 3.2.2.2 describes the data preparation and forecasting methods used in this analysis. Section
 3.2.2.3 presents EPA's methodology for determining site closure when evaluating different approaches
 to estimating future earnings.
        3.2.2.1 Assumptions and Choices

        Salvage Value

        The closure decision equation can be modified to include consideration of the salvage value of
 the site. If salvage value is taken into account, that is, the post-regulatory status is zero if the present
 value of post-regulatory earnings exceeds the salvage value of the site. For the meat product industry,
 EPA will not include salvage value in the. site closure model. EPA made .this decision for several reasons.
         17 EPA assumes that, when a site is liquidated, it no longer operates and closure-related impacts will result.
 In contrast, facilities that are sold because a new owner presumably can generate a greater return are considered
 transfers. Transfers cause no closure-related impacts, even if prompted by increased regulatory costs. Transfers will
 not be estimated in this analysis.
                                                 3-48

-------
       First! the market for used capital equipment appears to be limited. Having few alternative uses,
capital assets tend to be specific to the industry and have limited mobility (Anderson etal., 1998). These
assets are not viable in their current locations (otherwise the site would not be shut down), but it would
be expensive to move them to a different location.

        Second, a significant percentage of salvage value may be composed of current assets.  It is not
appropriate to calculate salvage value based on significant current assets because the value of cash, cash-
equivalents, and inventory is so liquid that an owner would not base a long-term decision on it. That is,
an owner would not liquidate a site because it shows a relatively high cash position on the balance sheet,
and thus has a high salvage value relative to cash flow. The cash could be transferred to other corporate
operations without such a drastic step as closing down operations.

        Third, excluding salvage value brings the site closure model into greater consistency with
 economic modeling approaches. That is, if salvage value is left out, a site is assumed to remain in
 operation as long as its revenues meet or exceed its operating costs. Sunk (i.e., capital) costs are not
 considered.
         Fourth, firms often do not record the value of assets at individual facilities; this information
 tends to be tracked at the corporate level. Therefore, even with the availability of Section 308 data, EPA
 frequently cannot reliably determine the salvage value of individual facilities.
         Net Income Versus Cash Flow
         EPA examined two ways to estimate the present value of future plant operations:
                 Net income from all operations, calculated as revenues minus operating costs; selling,
                 general, and administrative expenses; depreciation; interest; and taxes (as these items are
                 recorded on the site's income statement).

                 Cash flow, which equals net income plus depreciation.
                                                 3-49

-------
EPA could not collect reliable data on depreciation from the detailed survey. Therefore, EPA will use net

income as the measure of facility income. Excluding depreciation from the evaluation of facility income

may be likened to setting aside an allowance for replacement of current capital equipment when it wears
out.
    18
        Time Frame for Consideration


        EPA will use a 16-year time period in forecasting future income. (This corresponds to the time

period used in the cost annualization model—see Appendix A.) Although it might be appropriate to use

the estimated actual lifetime of the equipment rather than the depreciation period, doing so would yield a

lower estimated annualized cost because of the greater number of years over which to spread the capital

investment. EPA prefers to use the more conservative (shorter) time frame. The first year's data will not

be discounted, again to keep the cost annualization and forecasting projections on a consistent basis.
        3.2.2.2 Present Value of Future Earnings


        Adjusting Earnings to an After-Tax Basis


        Depending on the corporate hierarchy of which a site is part, the earnings reported in the survey

 may have to be adjusted for taxes. A site may fall into one of several categories:
         18 The trend in corporate finance appears to prefer cash flow as the appropriate basis for evaluating
 investment decisions because depreciation reflects previous, rather than current, expenditures and does not actually
 absorb incoming revenues. For example, Brigham and Gapenski (1997) note that in capital budgeting it is critical to
 base decisions on cash flows or the actual dollars that flow into and out of a company during the evaluation period.
 The Financial Accounting Standards Board, in SFAS Nos. 105,107 and 119, recommends using the present value of
 future cash flows to identify market value (FASB, 1996). In addition, although depreciation may intuitively be
 thought of as a capital replenishment allowance, in general, the value of historical capital expenditures (as reflected in
 depreciation) is not a reliable indication of future capital requirements to maintain operations.
                                                3-50

-------
             It is part of a multi-site corporation. If so, its EBIT will be adjusted to an after-tax level
             according to the taxable income of the.corporation, using the appropriate corporate tax
             rate.

             It is part of a multi-site organization whose income is taxed at the rate for individuals
             (e.g., partnerships, sole proprietorships, etc.). If so, its EBIT will be adjusted to an after-
             tax level according to the taxable income of the business entity, using the appropriate
             individual tax rate.

             The site is, or is part of,  an S corporation or LLC. If so, no adjustment will be needed.

              The site is the business entity, so the complete income statement data are supplied for the
              site. If so, because net income is presented on an after-tax basis, no adjustments need to
              be made.
       Adjusting Earnings to After-Tax Net Income


    .   For the first two categories (multiple facilities under the same ownership), net income will be

calculated as:

                    net income = [(EBIT)  * (1  - (federal + state tax rates))]


where the federal and state tax rates are dependent on corporation type and income at the business entity
level. (See Section A.1 for more details.) That is, EPA will reduce operating earnings by estimated taxes.
EPA will not make a similar adjustment for interest, because interest is generally not held at the site level

and it may vary widely from company to company (while tax rates are consistent).


        S corporations and LLCs (the third category) distribute income to the partners and tax is paid by

the partners at each partner's personal tax level. (That is, the company does not pay  taxes, the partners
pay taxes.) Therefore, no adjustment is needed. For the fourth category—single-site businesses—net

income will be taken directly from the survey.
                                               3-51

-------
       Forecasting Methods for Future Net Income


       Site net income must be forecast over the 16-year project lifetime. All forecasting methods to be
examined for and used in the closure analysis incorporate the following assumptions and procedures:


       •       No growth in real terms.

               Constant 1999 dollars. Data from 1997 and 1998 are inflated using the change in the
               GDP price deflator.


EPA is making the "no growth" assumption to avoid assuming that a site can grow its way out of an
economic impact associated with additional pollution control costs; essentially, EPA will assume that
sites are running at or near capacity and that significant growth is unlikely without a major capacity

addition.


        EPA will examine several different forecasting methods to address  site-specific variations:
                Most recent year (1999 data) as best indicator of future net income.

                Three-year average (1997 to 1999 data after inflation to 1999 dollars).19

                Time-varying income option #1 (called "Cycle 1"), according to which net income
                follows this 3-year pattern:
                        1999 = 1999 net income
                        2000 = 1998 net income
                        2001 = 1997 net income
                        2002 = 1998 net income
                        2003 = 1999 net income (pattern begins again)
  sin;
        19 EPA requested 3 years of data in the survey to mitigate the uncertainty in the analysis resulting from a
aj.igle data point For new or newly acquired facilities, however, 1 year of data may be all that is available for
analysis. For facilities with a trend in income, the most recent year may be the more conservative estimate of future
net income. If only 2 years of data are available, the model will calculate the average of the two values. If only 1999
data are available, that year's data are used.
                                                 3-52

-------
                     2004 = 1998 net income, and so forth
             If the facility had a good/bad year in 1998, the result will be a good/bad year every 2
             years.
                                                                                         j
             Time-varying income option #2 (called "Cycle 2"), according to which net income
             follows this 3-year pattern:
                     1999 = 1999 net income
                     2000 = 1999 net income
                     2001 = 1998 net income
                     2002 = 1997 net income
                     2003 = 1997 net income
                     2004 = 1998 net income
                     2005 = 1999 net income (pattern begins again)
                     2006 = 1999 net income, and so forth
              If the facility had a good/bad year in 1998, the result will be a good/bad year every three
              years.

After detailed survey data become available, EPA will examine the implications of the four forecasting
methods. EPA will select three forecasting options that provide a spectrum ranging from relatively
optimistic to relatively pessimistic forecasts.


       Discount Rate


       The final step in estimating each site's pre-regulatory present value is to discount the stream of
net income back to the first year in the time series. This step does not adjust the stream for inflation,
because the projections are in constant dollars. Thus, the discount rate used for discounting must be a
real discount rate,  obtained by adjusting the nominal discount rate for the expected annual rate of
inflation (see Appendix A). The same site-specific real discount rate is used in both the cost
annualization and  closure models.
                                               3-53

-------
       3.2.2.3 Projecting Site Closures As a Result of the Rule

       With three forecasting methods, there are three ways to evaluate a site's status. If a site's post-
regulatory status is less than zero, the site will be assigned a score of "1" for that forecasting method. A
site, then, may have a score ranging from 0 to 3.

       Closure is the most severe impact that can occur at the site level and represents a final,
irreversible decision in the analysis. The decision to close a site is not made lightly; the business making
the decision is aware of and concerned with the turmoil introduced into its workers' lives, community
impacts, and how the action might be interpreted by stockholders. The business will likely investigate
several business forecasts and several methods of valuing their assets. In its decision to close a site, a
corporation would weigh not only all data, assumptions,  and projections of future market behavior, but
also the uncertainties associated with the projections. When a corporation examines the results of several
analyses, it is likely to find that the results are mixed. Some indicators may be negative while others
indicate that the site can weather the current difficult situation. A decision to close a site  is likely to be
made only when the weight of evidence indicates that closing the site is the appropriate path for the
company to take.

        EPA will emulate corporate decision-making patterns when determining if sites will close. A
score of 1 for a site may result from an unusual year of data. If the score is 2 or 3, in EPA's judgement,
the weight of the evidence indicates poor financial health. EPA believes that this scoring approach
represents a reasonable and conservative method for projecting closures.
        Pre-Regulatory Conditions

        The closure analysis will begin with an evaluation of the pre-regulatory status of each site.
 Several conditions may lead to a site having a score of 2 or 3 under pre-regulatory conditions:
                                                3-54

-------
              The company does not record sufficient information at the site level for the closure.
              analysis to be performed.

              The company does not assign costs and revenues that reflect the site's true financial
              health. Two important examples are cost centers and captive sites, which exist primarily
              to serve other facilities under the same ownership. Captive sites may show revenues, but
              the revenues are set approximately equal to the costs of the operation (cost centers have
              no revenues assigned to them).

              The site already appears to be in financial trouble.


       The first two conditions would exist if a site's earnings data are held at the company level, or if

the site has been established not to show a profit, but to serve the company of which it is part. In either

case, EPA would not have sufficient information to evaluate impacts at the site level as a result of the
rule. The impact analysis would default to the company level, because that is the level at which relevant

decisions are made.


        The third condition identifies a site with complete  site-level financial information and no

confounding factors (i.e., it is not a captive site, a start-up  site, or a cost center) to obscure the financial
condition of the site. If the site is unprofitable prior to the  regulation, the company involved may decide

to close the site. This is likely to occur before the rule is implemented: the company will likely seek to
 avoid additional investments in an unprofitable site. The projected closure of a site that is unprofitable
 prior to a regulatory action should not be attributed to the  regulation.
         Estimation of Site Closures As a Result of the Rule


         EPA will consider the rule to have an impact on any site that has a score of 1 or zero in the pre-
  regulatory condition and a score of 2 or 3 after incurring the costs of responding to the regulation. That
  is, any site that is profitable before the regulation, but not after.
                                                 3-55

-------
       Direct Impacts


       Again, closure represents a final, irreversible decision in the analysis. EPA will therefore
estimate direct impacts from site closures as the loss of all employment, production, exports, and
revenue associated with the closed sites. This is an upper bound analysis; that is, it will project the most
severe effects, because it will not account for other sites increasing production or hiring workers in
response to the closure of a site.20 The losses will be aggregated over all sites to estimate the national
direct effect of the regulation.
3.3    REFERENCES
Altman, Edward. 1993. Corporate Financial Distress and Bankruptcy. New York: John Wiley and
       Sons.

Annual Survey of Manufacturers. 2000. The NBER-CES Manufacturing Industry Database (1958-
       1996). Downloaded 9/22/00 from http://www.nber.org/nberces/nbprod96.htm

Anderson, Donald, W., Brian C. Murray, Jackqueline L. Teague, and Richard C. Lindrooth.  1998. Exit
       from the Meatpacking Industry: A Microdata Analysis. American Journal of Agricultural
       Economics. 80(February 1998):96-106.

Armington, Paul S. 1969a. A theory of demand for products distinguished by place of production.
       International Monetary Fund Staff Papers. 16(1): 159-177.

Armington, Paul S. 1969b. The geographic pattern of trade and the effects of price changes.
        International Monetary Fund Staff Papers. 16(2): 179-199.

Brealey, Richard A., and Stewart C. Myers. 1996. Principles of Corporate Finance (5th ed.). New York:
        The McGraw-Hill Companies, Inc.

 Brigham, Eugene R, and Louis C. Gapenski. 1997. Financial Management: Theory and Practice (8th
        ed.). Fort Worth: The Dryden Press, pp. 428-431.
        20 The market model, however, accounts for this effect.
                                              3-56

-------
Dun & Bradstreet. 1998, Industry Norms and Key Business Ratios, 1997-1998. Desk-Top Edition.


Financial Accounting Standards Board. 1996. Financial Accounting Standards: Explanation and
       Analysis. SFAS No. 105 (Disclosure of information about financial instruments with off-balance
       sheet risk and financial instruments with concentrations of credit risk), No. 107 (Disclosures
       about fair value of financial instruments), and No. 119 (Disclosure about derivative financial
       instruments and fair value of financial instruments). Bill D. Jarnagin, ed. 18th edition. Chicago:
       CCH Incorporated, pp. 564-586.

 Gallaway Michael P., Christine A. McDaniel,  and Sandra A. Rivera. 2000. Industry-Level Estimates of
        U.S. Armington Elasticities, Office of Economics Working Paper. Washington, D.C.: U.S.
        International Trade Commission. September.

 Harnett, Donald L. 1982. Statistical Methods (3rd ed.). Reading, MA: Addison-Wesley Publishing.


 Huang, K. S. 1993. A Complete System of U.S. Demand for Food. Technical Bulletin Number 1821.
       ' Washington, D.C.: U.S. Department of Agriculture, Economic Research Service.


 Kaplan, Maureen F.  1999. Review of recent bankruptcy prediction literature Memorandum to William
       'wheeler, U.S. EPA, dated 12 February 1999.

 Mendenhall, W., D. D. Wackerly, and R. L. Scheaffer. 1990. Mathematical Statistics with Applications
        (4th ed.)- Boston: PWS-Kent Publishing Co.

 Office of Management and Budget. 1996. Economic Analysis of Federal Regulations Under Executive
         Order 12866. Washington, D.C.: Executive Office of the President.


 Outlook. Various dates. Livestock, Dairy and Poultry Situation and Outlook. Washington, D.C.: U.S.
         Department of Agriculture, Economic Research  Service.

 Putnam, Judith J., and Jane E. Allshouse. 1999. Food Consumption, Prices, and Expenditures, 1970-97.
         Statistical Bulletin Number 965. Washington, D.C.: U.S. Department of Agriculture, Food and
         Rural Economics Division, Economic Research Service.


  Quash, 2001. Personal communication from Nishea Quash, U.S. Census Bureau, to Calvin Franz, ERG,
          September 10, 2001.

  U S  Department of Commerce, Bureau of Economic Analysis. 1996. Regional Input-Output Modeling
          System (RIMS II). Total multipliers by industry for output, earnings, and employment.
          Washington, D.C.
                                               3-57

-------
U.S. Department of Commerce, Bureau of Economic Analysis. 2000. Gross Domestic Product by .
       Industry for 1997-1999. Survey of Current Business. Washington, D.C.

U.S. Department of Commerce, Bureau of Economic Analysis.  2001.  Gross Domestic Product by
       Industry: 1947-2000.  Downloaded on January 14, 2001.

UN FAO data: downloaded 2/20/01 from http://apps.fao.org/page/collections?subset=agriculture
       Agricultural Production/Livestock Primary & Processed/World+, United States of America/.

U.S. Census Bureau. 1999a. Animal (Except Poultry) Slaughtering. EC97M-3116A. 1997 Economic
       Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
       November.

U.S. Census Bureau. 1999b. Meat Processed From Carcasses.  EC97M-3116B. 1997 Economic Census:
       Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.


U.S. Census Bureau. 1999c. Poultry Processing. EC97M-3116D. 1997 Economic Census:
       Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.


U.S. Census Bureau. 1999d. Rendering and Meat Byproduct Processing. EC97M-3116C. 1997
       Economic Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of
       Commerce. December.

U.S. Census Bureau. 2001. Special Tabulation of Census Data for NAICS 311611, 311612, 311613,
       311615. Washington, D.C.: U.S. Department of Commerce. May.

U.S. EPA. 1995. Interim Economic Guidance for Water Quality Standards: Workbook. EPA-823-B-95-
       002. Washington, D.C.: U.S. Environmental Protection Agency, Office of Water.

U.S. EPA. 2001. Economic Analysis of the Proposed Revisions to the National Pollutant Discharge
       Elimination System Regulation and the Effluent Guidelines for Concentrated Animal Feeding
       Operations. EPA-821-R-01-001. Washington, D.C.: U.S. Environmental Protection Agency,
       Office of Water. January.

U.S. EPA. 2002. Development Document for the Proposed Revisions to the Effluent Limitations
        Guidelines for the Meat Products Industry. EPA-821-B-01-007. Washington, D.C.: U.S.
       Environmental Protection Agency, Office of Water.
                                            3-58

-------
                                      CHAPTER 4

                        POLLUTION CONTROL OPTIONS




4.1    EFFLUENT LIMITATIONS GUIDELINES AND STANDARDS

       The Federal Water Pollution Control Act (commonly known as.the Clean Water Act [CWA, 33
U S C  ง1251 etseaJ) establishes a comprehensive program to "restore and maintain the chemical,

physical, and biological integrity of the Nation's waters" (ง101(a)). EPA is authorized under sections
301, 304, 306, and 307 of the CWA to establish effluent limitations guidelines and pretreatment standards

of performance for industrial dischargers. The standards EPA establishes include:   .

               K^t Practicable rnntrnl Technology Currently Available (BPT) . Required under
               section 304(b)(l), these rules apply to existing industrial direct dischargers. BPT
               HrmStionsle generally based on the average of the best existing performances by plants
               cHrious sizes ages, and unit processes within a point source category or subcategory.
               Best V^I-M- TW^nolnpv F.cnnomica"Y Arhitwahle (BAT). Required under section
               304(b)(2), these rules control the discharge of toxic and nonconventional pollutants and
               apply to existing industrial direct dischargers.

               p~~* r^,^^1 Polhitimt Control T^nnlogv (BCT). Required under section
               304(b)(4), these rules control the discharge of conventional pollutants from existing
               mdusrria direct dischargers.' BCT limitations must be established m light o a two-part
               cost-reasonableness test.  BCT replaces BAT for control of conventional pollutants.
                 rrtr-nr—           for Existing Sources (PSES). -Required under section 307.
                Analogous to BAT controls, these rules apply to existing indirect dischargers (whose
                discharges flow to publicly owned treatment works (POTWs).

                M*w Source Perfr™™™ St^dards (NSPS). Required under section 306(b), these rules
                control the discharge of toxic and nonconventional pollutants and apply to new source
                industrial direct dischargers.

                Pr.trP.atn,ent Star^r-Hs for New Sources (PSNSV Required under section 307.
                Analogous to NSPS controls, these rules apply to new source indirect dischargers
                (whose discharges flow to POTWs).
          i Conventional pollutants include biochemical oxygen demand (BOD), total suspended solids (TSS), fecal
   coliform, pH, and oil and grease.
                                                4-1

-------
EPA is proposing effluent limitations guidelines and pretreatment standards for the meat products
industry in this rulemaking effort.
4.2    TECHNOLOGY OPTIONS

       EPA does not mandate technologies when establishing effluent limitations guidelines and
pretreatment standards. However, EPA evaluates various technology options in order to base the
limitations on demonstrated technologies and to evaluate the economic impact of the cost of those
technologies on the regulated industry. This section briefly describes the pollution control options
evaluated for each subcategory within the meat products industry. The Development Document (U.S.
EPA, 2002) provides a detailed description of the meat products industry subcategories and pollution
control options for each subcategory.

       Table 4-1 summarizes the technology options considered for each meat products industry
subcategory. The first column indicates the option number that appears in the cost and impact tables in
Chapters 5 through 9. The second column identifies contains a brief description of the technology
option.
       In assessing costing technologies, EPA distinguished between direct and indirect discharging
facilities. All direct dischargers in the industry were costed for four sets of technology options
regardless of meat type (i.e., red meat or poultry) or processing stage (i.e., slaughter, further processing,
rendering), except for poultry processors, who were costed for a technology option incremental to
option 4 (BAT 5).  Similarly, all indirect dischargers were costed for four technology options regardless
of subcategory.  However, indirect dischargers  were costed for a different set of technologies than were
direct discharging  facilities.  In general, wastewater treatment technology options for direct dischargers
included lagoons and ultra-violet disinfection; indirect dischargers were costed instead for equalization
tanks. That is the primary distinction between technologies for direct and indirect dischargers.
       For both direct and indirect discharging facilities, the treatment train costed in the higher
numbered options builds upon the set of technologies costed for the first option. Thus, under BAT 1,
direct dischargers were costed for: preliminary  treatment, dissolved air flotation, lagoon, and ultra-violet
                                               4-2

-------
                                       Table 4-1
                Meat Products Industry Treatment Technology Options
     Option
                                             'Treatment Unit
               *~r ^ ' •;"  "?MI.T  DirectPiscfaargers     ' ,
                                                                                      ซ, p
     BAT1
                    Preliminary Treatment, Dissolved Air Flotation, Lagoon, Ultra-Violet
| (nonsmall facilities) Disinfection

       BAT1.       Preliminary Treatment, Dissolved Air Flotation, Lagoon, Ultra-Violet
   (small facilities)   Disinfection, Drying Beds

                    Preliminary Treatment, Dissolved Air Flotation, Lagoon, Nitrification -
                     Suspended Growth, Ultra-Violet Disinfection, Drying Beds
      BAT 2
      BAT 3
      BAT 4
                     Preliminary-Treatment, Dissolved Air Flotation, Lagoon, Biological Nitrogen
                     Removal, Ultra-Violet Disinfection, Drying Beds

                     Preliminary Treatment, Dissolved Air Flotation, Lagoon, Biological Nutrient
                     Removal - 3/5 Stage, Ultra-Violet Disinfection,Prying Beds
      BATS
   (poultry only)
                     Preliminary Treatment, Dissolved Air Flotation, Lagoon, Biological Nutrient
                     Removal - 3/5 Stage, Filtration, Ultra-Violet Disinfection, Drying Beds
      PSES 1
      PSES 2
       PSES 3
       PSES 4
                      Preliminary Treatment, Dissolved Air Flotation, Equalization, Nitrification -
                      Suspended Growth, Drying Beds    	^	

                      Preliminary Treatment, Dissolved Air Flotation, Equalization, Biological
                      Nitrogen Removal, Drying Beds

                      Preliminary Treatment, Dissolved Air Flotation, Equalization, Biological
                      Nutrient Removal - 3/5 Stage, Drying Beds
Changes between technology options indicated by italics.
                                             4-3

-------
disinfection.2-3  These components are also included in BAT options 2 through 5. BAT 2, 3, and 4 are
distinguished by a single component: BAT 2 utilizes nitrification (suspended growth technology), BAT 3
replaces nitrification with biological nitrogen removal technology, and BAT 4 replaces nitrogen removal
with biological nutrient removal (3/5 stage).  BAT 5, .which only applies to poultry processors, adds
filtration to nutrient removal.
       Similarly, under PSES 1, indirect dischargers were costed for: preliminary  treatment, dissolved
air flotation, and equalization. These components are also included in PSES options 2 through 4.  PSES
2 adds drying beds to the costed treatment train, which then become components of PSES 3 and 4.
PSES 2,3, and 4 are thus distinguished by a single component: PSES 2 utilizes nitrification (suspended
growth technology), PSES 3 replaces nitrification with biological nitrogen removal technology, and
PSES 4 replaces nitrogen removal with biological nutrient removal  (3/5 stage).
        Table 4-2 summarizes the technology options proposed for direct discharging facilities in each
moat products industry subcategory. Note that hi all subcategories, EPA is proposing different standards
for small facilities than for nonsmall facilities.  EPA defines small facilities as:
                Subcategory A through D: facilities that slaughter less than 50 million pounds (live
                weight kill) per year;
        •       Subcategory E through I: facilities that produce less than 50 million pounds of finished
                product per year. Because Subcategory E (small processors) is defined as facilities that
                produce less than 6,000 pounds of finished product per day, all facilities in Subcategory
                E are by definition small;
         •       Subcategory J: facilities that render less than 10 million pounds of raw material per year;
                Subcategory K: facilities that slaughter less than 10 million pounds per year;
                Subcategory L: facilities that produce less than 7,000 pounds of finished product per
                day.

 In general, EPA is  excluding small facilities hi subcategories A though J from the revised standards, and
 is setting less stringent standards for subcategories K and L. EPA  is not currently proposing  any changes
 to pretreatment standards for indirect dischargers in any subcategory.
         2 BAT 1 for small model facilities includes drying beds in the costed treatment train; drying beds also
 included in BAT 2 through 5 for nonsmall facilities.
         3 Note that EPA's survey results indicate that all potentially affected nonsmall direct dischargers have the
 BAT option 1 treatment technologies in place.
                                                4-4

-------
                                              Table 4-2
                 Technology Options for Meat Products Industry Subcategories
============
Selected Option
for
Final Rule
BAT1
(small facilities)
BAT2
BATS
BAT 4
BATS
Tnnr Siihr.atp.onrv A tl
Subcategory
A-I>

BPT, BCT1
BAT, NSPS1

loush D. EPA excl
====
Subcategory
E-I

BPT, BCT2
BAT, NSPS2

udes small facilities
=====
Subcategory
J

BPT, BCT,
BAT, NSPS3


5 (those that slaugh
Subcategory
K
BPT, BCT,
BAT, NSPS4

BPT, BCT,
BAT, NSPS4

— — — — e^=
ter less than 50 mil
Subcategory
L
BPT, BCT,
BAT, NSPS5

BPT, BCT,
BAT, NSPS5


ion pounds live
 f \jฑ vJ UUL/ClLwfcVJl J **• t.IJ.V/fc*^ai JLJ- j jt-i*. *. *. •—•.-—-- — —	—                     —
"/eight kill per year) from the proposed revisions to the effluent guidelines.     ''.„.„.       -A  f
2 For Subcategory E though I, EPA excludes small facilities (those that produce less than,50 million pounds of
finished product per year) from the proposed revisions to the effluent guidelines.  Note that all facilmes m
Subcategory E (those that produce less than 6,000 pounds of finished product per day) are therefore excluded by
definition from the revised effluent guidelines.
3 For Subcategory J, EPA excludes small facilities (those that render less than 10 million pounds of raw product per
year) from the proposed revisions to the effluent guidelines.

-------
4.3    REFERENCES


U S EPA  2002  Development Document for the Proposed Effluent Limitations Guidelines and
Standards for the Meat Products Industry. EPA-821-B-01-007. Washington, DC: U.S. Environmental
Protection Agency, Office of Water.
                                             4-6

-------
                                         CHAPTERS

                                  ECONOMIC IMPACTS

       This chapter presents the projected economic impacts resulting from the costs of complying with
the proposed effluent limitations and guidelines (ELG) on the meat products industry.  The impacts are
estimated using the methodology outlined in Chapter 3. Impacts are estimated from the smallest scale to
industry-wide impacts, i.e., in the following order - facility level, corporate level, market level, and    ,
national level.  Impacts presented in this chapter are for medium, large, and very large model facilities
combined. Because small model facilities are almost without exception small business owned facilities,
impacts for small model facilities are presented in Chapter 6.1

        For each of the four facility level analyses, impacts are presented at a two-tier level, by:

                40 CFR -."52 subcategory (hereafter, subcategory), and
                meat type and process class (hereafter, class; see Section 2.4 for more detail).

 In addition, EPA presents a range of impacts. EPA first estimated the incremental compliance costs of
 purchasing new equipment to  match the technology train used as a basis for analyzing an option; costs are
 incremental in that facilities are costed only for additions to current treatment in place necessary to match
 the technology train.

         However, EPA determined that it may be possible for some establishments to upgrade (or retrofit)
 current treatment in place to meet the specified technology train at lower cost than if they purchase new
 equipment. For example, a facility that currently owns a nitrification system (specified for option 2) can be
 retrofitted to become a nitrification and denitrification system, which will meet the requirements of option 3
  (see Development Document, Section 4.6.4 for details). EPA only estimated retrofit costs for options 3  and
  4.  For the remainder of Chapter 5, EPA will present, where applicable, the costs (and associated impacts)

          1 No small facility impacts are included in the analyses presented in Chapter 5. As documented in
  Chapter 6, EPA estimates that a total of four small facilities in Subcategory L are potentially affected oy the
  proposed rule  EPA projects that these four facilities will incur posttax annualized compliance costs of $2,600
  ($700 per facility) and that none of these facilities should close as a result of the proposed rule.
                                                  5-1

-------
of purchasing new equipment as an upper-bound estimate, and the upgrade or retrofit costs that will meet
the same requirements as a lower-bound estimate.

        The facility level analysis is discussed in Sections 5.1 through 5.4. Section 5.1 presents total and
average facility compliance costs for the industry. Section 5.2 discusses projected facility level incremental
closure and employment impacts. Section 5.3 reports facility nonclosure impacts and Section 5.4
completes the facility level impact analysis with a financial ratio analysis. Section 5.5 discusses financial
distress at the corporate or business entity level. Market level and international trade impacts are presented
in Section 5.6.  EPA examines secondary and indirect employment and output impacts in Section 5.7. EPA
estimates new sources in the meat products industry in Section 5.8. Finally, EPA summarizes impacts
under the proposed options in Section 5.9.

        The economic analysis is based on a wide variety of sources including the screener survey and
publicly available data. However, the facility counts in each class and subcategory are based on estimates
derived from the stratified random sampling procedure used to determine survey recipients. Sixty-five
facilities were specifically selected to receive surveys ("certainty facilities"). Information on these 65
certainty facilities was not available in time to complete subcategorization and analysis of these facilities
because information on these facilities was collected in the detailed survey and it could not be processed as
quickly as the screener survey.  Therefore, to project potential impacts to these 65 certainty facilities, EPA
totaled impacts by subcategory (or class) and discharge type, then inflated these impacts by 8 percent.
EPA is thus implicitly assuming that the 65 certainty facilities are similar to the model facilities used in the
remainder of the analysis, and impacts are therefore proportionate to impacts projected for other facilities.
However, EPA could not identify the subcategories or classes in which these impacts may occur in time to
include precise estimates for all aspects of the analysis.
 5.1     TOTAL AND AVERAGE COMPLIANCE COSTS

        In order to estimate impacts, EPA calculated total and average facility compliance costs in 1999
 dollars by subcategory, meat type and process class, discharge type, and technology option.  The
 compliance costs include estimated capital costs, annual operating and maintenance costs, pretax
                                                5-2

-------
annualized, and posttax annualized compliance costs.  The annualized costs are analogous to a mortgage
payment that spreads the one-time investment of a home over a series of constant monthly payments. They
are calculated as the equal annual payments of an annuity that has the same present value as the stream of
cash outflow over the project life and includes the opportunity cost of money or interest (see Section 3.1.1
for more detail).

        In general, estimated annualized compliance costs for direct dischargers consistently increase with
the technology option.  Also, all direct discharging facilities have sufficient treatment in place to meet the
requirements of BAT 1,  and therefore costs for BAT 1 are zero for all classes.  For indirect dischargers,
PSES 2 has the highest cost .per facility in several classes, and PSES 3 is estimated to have lower costs
than either PSES 2 or PSES 4.  Within each subcategory, generally, indirect dischargers  incur higher
compliance costs than direct dischargers on a per facility basis for equivalent technology options.
        5.1.1   Total and Average Compliance Costs by Subcategory
        5.1.1.1  Upper-Bound Costs

        Table 5-1 presents total and annual compliance costs by subcategory, discharge type, and
 technology option.  As the table shows, for the direct dischargers, total posttax annualized compliance costs
 range from a low of $0.2 million under BAT 2 for.Subcategory E through I, to a high of $72 million under
 BAT 4 for Subcategory A through D. Estimated average posttax annualized costs range from $11,000 per
 facility under BAT 2 in Subcategory L, to $1.1 million under BAT 4 for Subcategory A through D. Under
 the proposed option, BAT 2 for Subc,ategory J and BAT 3 for all other subcategories, average posttax
 annualized costs per facility are as follows:
                Subcategory A through D:
                Subcategory E through I:
                Subcategory J:
                Subcategory K:
                Subcategory L:
$550,000
 ,$22,000
 $14,500
$335,000
$120,000
                                                5-3

-------
1


11
1 w
0
1
•<




11
11
1



„
8










11
"I
1|
-5


s
0
U
i
O
1
^
c
a
u
I "2
S j
SS *^"
' S
c
u
5
*S

O
0
[
-e
'5
c
U

c
i

i*
1 i





























Q
a
•*
^


'

3

ฃ



o



%




o

fe<

c




o





H
PC


VC

3
CO
CO
&e
CO
fee



CO
VC
•f.

CN
ir
o
fe=



in
oo
oo
1
V
fee
c
VD
C
0
in
co
CO
oo


vo
CN
OO
V
CN
OO
ff



&
PC




o
N
O
in
m
fee
o
in
CO
oo
&e



m
m
CO
ir;
•f.

3
vc
fee


u-
r-
VC
CO
fee
r-
oo
VO
c
w
\r
oc
CO
O
vo
CN

1
VO
CO
VO
j*
ฃ\
fee

H





CN
VO
8
D
\
fft


oo
SO
VC
fee

1
in
o
CO



oo
in
CO
CO
1
oc
VO
CO
c
os
C
OS
0
00
00
CN


*e
c
vr
c
CN
VO
u
&e

pi
PO








































n

S
00
op
ft



N
in


vc
ir
ir



ง
CN
CO
C
OO
CN
VC
fee
o
o
CO
co
•F.

oc
ฃ
CN
co
V


V.
on


c
vc

s;
2
ป
•o
CO
ft


R
oc
CN

VO
oc
o"

^

c<
VO
S
0
oo
oc
o
0
o
CO
^J-
r-

O
oo
VD
CO
in
CN
VI
fee







CO
n
CN
oo
Is
o
CO
in
00
T-H*
ye


in
in
oo
;*
VO
fee

X
Os
VO
VO,
r-^



\r
in
VO
OS
00
in
V
VO
OS
CN
a
oo
o
c
o


o
CN
CN
OC
VC
^
fee

ft
on




OS
ON
CN
00
CO
CN
VO
00
ye


I
Os
ie

CN
VO
S
o

fef

vc
Os
C
VO
CO
S
C--
Vฃ
VD
C
OS
r-


^;
CO
t~-
CN
o
V
fey

O"
V-
P-

































;






ฃ

ft



o



%




o

o

o




o





H
PC


OS

VO
CN
VO
ฃ
Os



oo
oo
fee

in
CO
OS
ฃ



1
S
VC
r
C1"
VO
OS
oo
in
CO
w


p-
VC
u-




H
m




X)
N
CO
S
co
•o



CN
OO
Os
OS
fee

1
OS
CN



OO
fe=
os
Os
O
VC
if
VO
oc
CO
***


\r
oo
VD
VI
CN




pa




CN
oo
N
3S
Tป



n
OS
S
w-

g
CO
oo
VC
fee


as
CO
OO
CN
OC
o
c:
vt
oc
cf
c
CO


Os
in
VI
o
c<"
fee


EH
PC








































in
^
vo
CO



oo
5?
fee-

ง
CO
VO
CN
fee



Os
as
S
oo
o
O
O
r-
C<"
oo
oo
oo
O


CO
CO
CN
co
V



[X
ฃ/•


O.

3
CN
VD
N
3s
vq,
CO



•o
oo
*3

ง
VC
VO_
fee-


VC
in
OS
vc
vc
c<
OS
CN
V
a
U"
c
VO
V!
CO


c
m
oo
r-
OS
CO
oo


rN
V.
O-




n
S
o,
cT
CO
o



1
00
VD
fee

00
oo
01
V
fee


oo
S
in
r~
OC
c
r-
oo
o
i
c
OS
(T
Os


C
vc
V
o


m
[/;
V
PL




d
Os
OO
ft
$435,725



3s
X>
S
fee

Os
O



c-
o
a
fee
r-
00
VO
CO
C
2
5
OS
C*"
c*"
VO


c
CO
in
r-
OS
CN


^
W
PL







































V
fee

o



%



&




ฃ

.-6*

C




O





H
PC


T—
CN

in
fee
co
'ft



y\
S


%




vc
S
co
fee
V
ff
\r
v
r-
V



c:





E-
pc




CN
OS
ป
-&e
VO
XD
n
>e-



o
?>
CO
CO
fee

in
X
V
C/D


1
I/-
feซ
VO
Os
CO
CO
r-
co
\r
fee
vo
c
co
oc
CN


C
in
co
c
CN



C-j
PC




oo
CO
oo
•&e
2
xf
N
fee



9
^
fee

cr
cs
C



ง
CN
oc
C*"
fee
t-
CN
ex
V
fee
co
c
OS
c
CN


O
OO
oo
co
r-



E-j
PC





-------
 o
  B 8
  I'i
3ซ 8P

ill
ซj i — i ^
^ I-1 en





ง
1
•**











t—
H











S *>
en M
||
•G

{f
O
1
'O
!


CS
U
I"8
81
1
^

X 'C
C^ <
ฃ a
1
1
1
O
O
•
o
;

C
c
&
•fi .*•
i r
s
o
VO
S




vฃ
vo
•EA




vo
c

O
o
c
ei"
€/=


T—
V
p.
P

•I/"

en
vo
oo
oo
oo
•-i
1O
cs
{A

o
^

>n
X3
^"^
ro"


CS
IT
I
rซ-j

VO
VO
in
en
ir
o
CS
m
C*"
O
OO
o
EA
Os
en
oo
60
O
e>


CN
V
t
p.



en
oo
A
so
•n

Os
3S
f.

m
Os
VO
fA


VO
VO
en
r-
r~
rf\

m
o
\r
CN
ii
C
C""

CS
m
VC
o
cs

V)
I
p.



3
1
5*
o
n

oc
-

VO
VC
IT
fi.
JA


9
a
EA

OO
c
Ss-
CN
ci"
EA
CS
c
o
en

EA
o
P.
P-

































•
o

E


S


s




E



o

0


o




r—
E-
ec



CNl
OS
A
CS
O
A

>C
Os

VC
SA



f-
en
cn
VC
o
•fA


g
v<
E/
O
r-i
OS
e*"

i
o
EA


ฃ
PC



rs
en
en
Sn
oo
n

CN1
er

\r
in
ฃ


in
CS
oo
O
ir
O
ff

w
en
oo
oo
t—
VO
oo
OS
o
cs

s
r-
rs
CN
CN
ff
H
PC



n
VO
ง
OS
VO
rt


(N
or
cs
?
f.

^
en
y


m
r-
r-
"T.
r~
m
•ฃ/^

cs
oo
cs
en
i — i
VO
m
oo
vo
m
oo
vo
t-;
\r
cs
•EA
VO
OO
OS

ff
?
<
PC



3O
CS
cs
cs
en
VO
•EA

r-
S
en
•6A

VO
CS
oc
VC
r-
e<-


en
oc
VD
o"
$f

en
0
oo
C
&
vo
cs
en
O
en
VC
vo
rs

'tf-
C
C
r-
c
•6ฐ
E^
p:


































0
*
oo
T)
N

en
J;
3

g
en

en
•ee-
o
Os
O

^
CS

0
•*




OS
oo
Os


oo
cs
in
in
o
in
cs
in
r-
oc

oo
oo
O
en
OS
O
in


[
p.
oc


in
en
A
3
OS
VO

^
OO

V
oo
*


ง
V
en
VO
C/j

VO
in
m
cs
V
V
0
r-
in

c
^
in
r-


CN
V
p.
C/"
PL



en
en
OS
A
Os
?s
cs
en
•n

S
(f.

oo
en
oc
m
47


r-
VO
r-
en
oo
in_
i^
C/J

CS
ฃ
o
s
OS
VC
p-
Vฃ
e*"

CN
\r
en
en
r-
OS


en
V.
p.
p.



VO
OS
N
OS

m
en

Os
oo
IT
VC
•— i
CS
4A


OO
OO
m
t/j

oo
oo
r-
oc
en
en
vo
O
o

in
VO
en
r—
c
^


c
p.




-------
11
      o
              ON
              vo
              01
              fee
              fee-
                  >
                                                                                           s
                                                                                           •^3  rt
                                                                                           ?  I

-------
       Among the indkect dischargers, PSES 1 under Subcategory J has the lowest total posttax
annualized compliance cost at $0.8 million, while PSES 2 under Subcategory K has the highest cost at
$114 million. The range for average posttax annualized cost is from $10,000 for PSES  1 under
Subcategory J to $1.5 million under PSES 2 for Subcategory A through D.  EPA has chosen not to propose
any options for indkect dischargers.                                                   .
     ,   5.1.1.2 Upgrade Costs

        Table 5-2 presents total and annual upgrade compliance costs by Subcategory, discharge type, and
 technology option. For the dkect dischargers, average posttax annualized costs for upgrading range from
 $16,000 per facility under BAT 3 in Subcategory E through I, to $600,000 under BAT 4 for Subcategory
 A through D.  The lower end of the retrofit cost range is not much different than the lower end of the upper-
 bound cost range.  However, at the top of the range, retrofit costs are about 45 percent'lower than the
 comparable upper-bound costs. For the proposed dkect discharger options, average posttax annualized
 upgrade costs per facility are as follows:
                Subcategory A through D:

                Subcategory E through I:

                Subcategory J:

                Subcategory K:

                Subcategory L:
                     $374,000
 68 percent of upper-bound costs
                      $16,000
 73 percent of upper-bound costs
                      $14,500
100 percent of upper-bound costs
                     $229,000
 68 percent of upper-bound costs
                      $85,000
 71 percent of upper-bound costs
  In general, except for Subcategory J for which retrofit costs were not estimated under option 2, retrofit
  costs are 27 to 32 percent lower than the upper-bound costs presented in Section 5.1.1.1.
                                                5-7

-------
r-g
II
ll
18
r





CrJ
O
1
-<3?
^^










•






>~
ฃ•<










C3 " -U
t/j ฃ5
S I
P

5

i
O
^r
s
og
0
••'S
u
^5
^Js
a
a
H
11
C
0

||
2 *~
Hi a
s
•<
|
c

ง

O
O
u
"3

B
0
c
B
O2
tj .2
1 1






































j;
|

O
Q
t
Co
2


g



|





2





2




2



2




2




ฃ
PH


VO
vo

g


g



|





g





2




2



2




2




S
PH




VO
CN
CO
,-jT
ro
fee
o
CN
ON
m
fee


m
in
CO
in
ON
CO
fee-


co
CN
in
CN
CO
f^
fee-

vo
ir

ro
fee
ON
vo
oo
o
o
ON
CO
fee
oc
CO

vn
O
oo
m
vo
oo
m
CO
o
fee
P
PH




P
ro
sO
O
CO
in
T— H
CO


co
oo
VO*1
^^
ie


•
2


g



g





2





2




2



2




2




p
pa


CN

2


2



|





2





2




2



2




2




S
(T




ON
VO
ON
T-H
fee
CO
00
VO
00
00
1 — 1
fee


ON
CO
co
fee


oo
CO
co
ON
T-H
in
C/J


vo
cr
v

•fee-

co
c-
VO
ON
CO
fee
VO
ON
cr
oc
CN

|
E
ON
O
ff.

CO
p:




T 	 1
T-H
fee
1
ON
i — i
CN
&e


i — i
o
Vj1
5e


VO
in
in
C^-
CO


in
ON
VO
ON



vc
O
vo
fee
i
ON
ON
CN

O
co
m
in
fee

ฃ
pa





-------

-------













a
O
a!
H
>
S "3
S ซ
2 N
O *CB
r\^ 'T*
WH jj*
C
ntf*



gs
(11 iM
•*1
•3


S


1
O


•S3

o
O
*
•s
A
C3
0


•i 1
i|
a
<


I'S
ll
pug
C
n
t-^.
co



CO
^j*
T-H
CO
VO
1 — 1
(ft



PJ:
in
n
oo
n
CM
(ft

P3




m
no

ON
m
ro
(ft




•vo
S
CM
r-
(ft



in
0

co
(ft



CM
CO
0

VO
,T-H
(ft



VO
T-H
m
CO
CM"
T-H
OO
(ft


vo-
r-
•*
oo
CM
ON
CM
T— (
<<^


CO
s
•*f
CO
c-~
ON
00
fft


$375,404,565

1
m





P"^
2








|





T-H
CO
^;(-
co
CO





1
1



s?
in
[-C
>n
CM



oo
00
CM
T-H
00
CO
CM
(ft



in
CM
oo
co"
>n
CM
(ft


o
ro
ON
S


ro
CM
oo
oo
ON
T— (
(ft

in
r-
ro
06"
CO
00
(ft
CO


'


in

ON
ON
CO
(ft




ON
CO
I



p;
T— (
^-fT
a



^j.
00
oo
O
S
CO
ซซ•



^
CO
CM
co
o
co
(ft


CO
O
•<*•
r-~
m
VO
(ft


S
o"
CO
co
ON
T-H
(ft

CO
co
CO
CM"
CM
T-H
CM"
co
CO
(ft.
I





0
•s

•3

E^
'B
a
u
to
_w
m
H
<
n
ฃ

1
o
ฃ


en
3
.2
<8
8-
1.^
-c:
tJ
5

d>
X
ง
*o
s
w
B
_O
1
&
S3
.S •
v •&
a ง
fcป D-
is o
udes pou
ther BAT
a ฐ
.S >H
J-S
e>^
P ^
ay w
If
c/i w
. w
S ซ
o .ซ
nj •--'
und in Poultry oper
the number of faci
•a 8
^ s
H
ซ 1
^?
< 0
CQ 
-------
       Among indirect dischargers, the average facility posttak annualized costs for upgrading range from

$185,000 for PSES 3 under Subcategory J to $1.1 fnillion undel PSES 4 for Subcategory A through D.
        5.1.2   Total and Average Compliance Costs by Class
        5.1.2.1 Upper-Bound Costs


        Table 5-3 presents total and average compliance costs by meat type and process class, discharge

type, and technology option. For the 12 direct discharging classes:


        •      BAT 4 is the highest cost option (posttax annualized costs) in seven classes:
               —     red meat first processing;
               —     red meat further processing;
                       red meat first processing and rendering;
               	     red meat further processing and rendering;
               	     red meat first processing, further processing, and rendering;
               —     mixed further processing;
                —     rendering.

                BAT 5 is the highest cost option (posttax annualized costs) in five classes (there is no BAT
                5 option  for the red meat classes):
                —     poultry first processing;
                —     poultry further processing;
                	     poultry first and further processing;
                —     poultry first processing and rendering;
                	     poultry first processing, further processing, and rendering.


         For the 13 indirect discharging meat type and process classes:


                 PSES 2  is the highest cost option (posttax annualized costs) in nine classes:
                 	      red meat first and further processing;
                 	      red meat first processing and rendering;
                 	    ,  red meat further processing and rendering;
                 —      poultry first processing;
                 —      poultry further  processing;
                 	      poultry first and further processing;
                 	      poultry first processing and rendering;
                 	      poultry further  processing and rendering;
                                                 5-11

-------











ฃ3
O
1
^5*







.













i-3
•
Q
i
e
ซ-
i**
•^
"^
as
69






O
69





O





O







o
69




69





^
ฃf^




O






i — i
m


vo


s






o
69





O
69





O
•fA







O
69




8





r- J
Q*^




O
69






M





VO
in

vn
69



Tf
P-
CO
^
fA



OO
ON
ro
1 — 1
T—4
fA


O
fA







co
in
O
69


m
oo
vn
69-



ON
OO
co
00
VO
69

69






i
m





00
ON
CO
T-H
rN
i — i
69

ON
OO
in
So

69


VO
8


T— C
^9-

ro
00


00
69


00
CO
oo"
CN
r-
69

co
f^
T-H
1 — 1


in
in

o
o
vo
69

ON
T-H
o
8
oo
ฃ




M














































^
i
ฃ••
o
1
to
.1
$

J*
OH
k.
*-
ja
ซi!
"S
as
o






o





o
69





O







O




O





0





o






T 	 1
S3


1 — 1


ON
ON
•^j-
^•^



CO
oo
fA




o
f —
c/^



f5
00
ro





VO
oo
CO
ON
>n


CM
in
o
o
T-H
69


ON
p-
oo
ON


CO
oo
VO
in
69




S





CM
i — (
P-
m
fA



CN
fA




CN
OO
O
p-
•FA



OO
VO
O

^yCj.



O
in
CO
VO
69


1
O
1 — 1
1 — 1
69


ro
oo

co


CN
ฃ?
69




m





>n
VO
o
00

T— 4
69

i — i
O
CO
VO
CN
CN
69


in
VO
in
•^>
T— H
T-H
69

vo
00
p-
m
0
T-H
69

O
00
r-
VO
fA

T— 4
VO
in

p-
CN
fA


ro
oo
P-
p~
co
69

CM
P-
co
ON
vo
CM"
i — i
69



1






























































co
p-
00
in

(fy


oo
CM
P-
fA



O
CO
VO
p-

fA


^
P-
in
ro
rN
69


O
t — •
VO
fA

VO
OO
vn"
p-

CN
69

ON
P-
00
1 — 1
0
o
00
69

VO
CO
ON
ON
ON
CO
69



1


VO


ON
00

in
00
T 	 1
69

"ง
in
oo
rN



T-H
in
in

69

in
VO
T-H
r-_)
ro
CN
i — i
69

T-H
T-H
CN
vo
CO
69
ON
CN
in
CO
o
oo
69

in
ro
p-
3
T-H
vo"
CN
69

00
VO
in
ro
oo
vo
O
CN
69



CN
GO
W
CO
OH





T-H
r-
CN
oo

T 	 1
69

0
oo
p-
p
(M
69


VO
VO
-Tf
•if
1 — 1
69

VO
rN
vo
CN
rN
rN
T-H
69

CN
in
ON
ON
69
ON
O
in

ON
^
69

O
in

0
CO
69

1
O
in
O
CN
69



CO
CO
W
CO
OH





OO
ON
CO
ON
rN
CN
69

co
1 — 1
in
oo
ro
69


OO
VO
VO
T 	 1
69

VD
O
CO
0
CN
P-

69

co
oo
oo
CO
m
oo
ro
69
ง
10
in
oo
m
69

ro

CM
in
o
00
CN
69

in
VO
co
^
O
ON"
00
69



1









































Q~
'
"^
r*>


S
-B
i;
00
O
?
a,
k.

69

VD
co
P-
ON
VO
ON
O
T-H
69



CN
CO
1X1
CO
OH





OO
in
oo
m
ON

69

VO
VO
vo
T-H
in
p-
69


T 	 1
ON
T—4
in
ro
69

CO
T-H
CO
ro
oo
p-
ro
69

S
O
s
oo
CO
69
vq
VO

o
T—4
CM
69

T-H
p-
VO
CO
in
oC
69

OO
VO
P-

in
O
69



CO
co
CO
OH





S
in
ro
0
in
69

in
•s
r-
69


S
T-H
in

ro
69

in
VO
T—4
V)
ro
ON
ro
69

ON
vo
ON
ON
O
i — I
69
CO
VO
vn
o
CO
T-H
CN
69

CN
C^
ON

-------
i
                           n
                       VO

                   CN  -CN
                   fee  fee
                                   OS
                               fee-
                                oo
                                en  vo
                                OS  CN
                                cs
                                    CN
                                    •*
ID
                         OS

o
fee




o
fee

o
fee


5
T-H
T 	 1
in
fee

en
in
Os
3
V5-

r-
T— 1
3


en -3-
in O
oo en
vo vo"
CN in
fee vo

Os
,Os
fee

en
oo
OS
in
vo
fee


fee
en
r-
o
en
fee
T-H
o
•n
en
. en
fee
                                    fee
                             fee
                                 vo





O
fee



o
fee

I
1
>^
i — i
CN
.fee-


o
OS
r-
CJ
cs
T-H
fee

m
OS
"T.
•*
o
CNl
fee-

in
VO
fee


VO
Os
VO
s
&e


r-<
r— 1
"i;
ฃ
T-H
fee

r~
•n
en
en
fee
OS
o
in
en
r-
y
fee
in
T-H
CN
in
tN
VO
CN
fee

                      o
                              OS
                              fee
                              CN
                      oo
                                         to
                                         fZ
                                         •S
                   --  ซ
                    o  fa
                   ..•—-.

o
fee-


g
:
i
'
>
>
is
* ซ
A
3
1}
S .,*
0
d_
fee-
en
oo
m
Os
r-H
fee

5
OO
VD
OO
fee

g
PH

fee
o
••*
VD
OS
fe^

0
en
OS
S
s

s
pp
=
VO
CN
fee
g
1
i— (
^e-
CN
VD
T-H
oo
CN
^.
en
fee
1
pa

                                                              os
                                                                  VO
                                                                  fee
                                                                ^

                                                              of
                                                          vo
                                                          OS
                                                                                     ซซ•
                                                               Os
                                                                          Os
                                                                                                           v,
                                                                                                               fee
                                                                                  fee
                                                                                         tN
                                                                                         eN
                                                                                         fee
                                                                                                        CN
                                                                                                        OS
                                                                                         OO  Os
                                                                                         en  en
                                                                                                         vo
                                                                                                         -&e
                                                                                                 Os
                                                                                                 vo
                                                                                                 CN"
                                                                                                 m
                                                                                                 VD
                                                                                          Os

                                                                                                     fee
                                                                        CN
                                                                           eN
                                                                                          Os
                                                                —
                                                                OO
                                                                w
                                                                00
                                                                PH
                                                                                           fee
                                                                                              cs
                                                                                                         CN
                                                                                                         OS
                                                                                                                     en
                                                                                                                     T—<

                                                                                                                     m
                                                                                                             Os
                                                                                                             CN
                                                                                              os
                                                                                                      S
                                                                                                      fee
                                                                               K
                                                                                                      r-

                                                                                                          CN
                                                                                                          CN

-------

-------

-------

-------
_

-------
               —     poultry &st processing, further processing, and rendering.

       •       PSES 4 is the highest cost option (posttax annualized costs) in four classes:
               •—     red meat further processing;
               —     red, meat first processing, further processing, and rendering;
               —     mixed further processing;
               —     rendering.

       For each subcategory in Section 5.1.1.1, average facility costs actually consist of a weighted

average of class level impacts. Hence, under the proposed BAT options (BAT 3 for Subcategory A

through D, E through I, K, and L, and BAT 2 for Subcategory J), the range of average facility costs by

class within each subcategory are as follows:
               Subcategory A through D
               — red meat first processing
               — red meat first processing and rendering

               Subcategory E through 1:
               — red meat further processing
               — mixed further processing

               Subcategory J:
               — rendering2

               Subcategory K:
               — poultry first processing
               — poultry first processing, further processing, and rendering

               Subcategory L:
               — mixed further processing
               — poultry further processing
$550,000
  $7,000
$970,000

 $22,000
  $6,000
 $92,000

 $14,500
$335,000
$265,000
$853,000

$120,000
  $92,000
$124,000
 In sum, average posttax annualized costs per facility for the proposed options range from a low of $6,000

 for the red meat further processing class to a high of $970,000 for the red meat first processing and

 rendering class.
        2 In Subcategory J, the class (rendering) is identical to the subcategory.

                                                5-18

-------
       5.1.2.2  Upgrade Costs


       Table 5-4 presents total and average upgrading compliance costs, by meat type and process class,

discharge type, and technology option.  The rank order of costs among classes is unchanged: BAT 4 is the

highest cost option (posttax annualized costs) for red meat, mixed, and rendering classes. EPA did not

estimate upgrade costs for BAT 5, which thus remains the highest cost option for poultry processors.

Because upgrade costs do not apply to option PSES 2, it remains the high cost option for most indirect

discharging classes; PSES 4 is the highest upgrading cost option for the remaining classes.


        The range of average facility costs for the proposed options and a percentage comparison to uppei-

bound costs under each subcategory are:
                Subcategory A through D
                —     red meat first processing
                —:     red meat first processing and rendering

                Subcategory E through I:
                —     red meat further processing
                —     mixed further processing

                Subcategory J:
                — rendering

                Subcategory K:
                —     poultry first processing
                	     poultry first processing, further processing, and rendering

                Subcategory L:
                —    mixed further processing
                —    poultry further processing
$374,000
  $7,000
$658,000

 $16,000
  $5,000
 $64,000

 $14,500
 $229,000
 $181,000
 $578,000

  $85,000
  $64,000
  $89,000
  Average upgrade posttax annualized costs for the proposed dkect discharger options range from a low of

  $5,000 under the red meat further processing class to a high of $658,000 under the red meat first

  processing and rendering class (about 33 percent lower than the upper-bound costs for this class).
                                                5-19

-------
H
       .a
       •s
11
ฃ2
        a
        a
        _

         c
  '&
                 VO


                 vo"
                                                               oo
                                                               ON
                                                               CO
                                                                                  >n

                                                                                  J*
                                                                                                 VO
                                                                                                 VO
                                                            O
                                                            oo
                                                            r-
                          ON
                          •fee-
                          S
                                                                          oo
                                                                          co
                                                                          ON
                                                                          00
                                                                          ON
                                                                          VO
                                                                                  CN
                                                                                  CN
                                                                                  *f!r
                                                                                                     ON
                                                                                                C-^
                                                                                                i—i
                                                                                                ซft
                                         (N
                                             VO
                                             •ae-
                                             CM
                                                            CN
                                                                               ON
                                                                VO
                                                                VO
                                                                                            CM
                                                                                            CN
                                                                                            CN
                                                                                                 VO
                                                                                                vq
                                                                                                in
                                                                                                vo
                                                                VO
                                                       CN
                                                       CN
                                                       CN
                                                                               CN
                                                                               CN
                                                                               ro
                                                                               •6ft
                                                                                  CN
                                                                                                 ON
                                                                                                 ON
                                                                                                 CN
                                                                                                CN
                                                                                                     VO
                          ON
                                         CN
                                                                               VO
                       V,
                       o"
                                                                                  V,
                                                                                                 CO
                                                                                                 CN
                                                                                                     VO
                                                                CO
                                                                in
                                                                                                m
                                                                                                CN
                                                                                                m
                                                            ฃ
                                                                                                  CN)
OO
00
                           ON
                           in
 ON
• oo
 u-

 vo"
 OS
                                             o>
                                        oo
                                        co
55
                                                    CN
                                                    o

                                                    i/-
                                                    ON
                                                                                                     1
                                                                                                     O
                                                                                                     vo"
                  CO
                  oo
                  VO
                  ซซ•
                                                                                   ON
                                                                                   ON
                                                                                       &
                                                                                       o
                                          ee-
                                                            c<-

                                                            c<-
                              j;
                          s
                          h.
                          1
                          •g
                                                           o,
                                                           oc
                                                                VO
                                          a

                                          I
                                         -Cj
                                          s
                                                                o< -
                                                               •I
                                                                                             in
                                                                                             oo
                                                                                                  Sf*
                                                                O
                                                                ON"
                                                                oo
                                                                o
                                                                v$
                                                    ON
                                                    VO
                                                    oc

                                                    in
                                                    oo
                                                    ON
                                                                                                     •vc
                                                                                                     >n
                                                                                                 in
                                                                                                 r-
                                          •fc
                                          a
                                         m
                                                     CO
                                                     PH
                                                oo
                                                vo
                              co
                              H
                              CO
                                   m
                                   oo
                                                                               CO
                                                                               CO
                              P-l
                              OO
                                                             V
                                                             t

                                                            (fc
                                                                                              H
                                                                                      J
                                                                VO
                                                                CO

-------

-------
I'








w
0
-
63
%








a ซ y
a i
5-4 (cont.)
rage Retrofit C
id Process Clas
•iS*
C3 n
ex
oc

Cf
c
VJ
ir
OS


C
CN

r--
CO

•5/3


ฃ
PC





Q
in
CO
N
A

o





r;
44



oo
00
00
X
CO



c*"

ir
o
r-~
c<"
,_,
5?

e
CO
in

,_H
co

fe<3


?
pc





2




*





*




55






55

55

55



1








E-
fC


















































S;




55





55




55






*

55

^



j








p.
ty
p.



CN
O1

i;




55





2




55






^

<

%



*








CN
q.
V
p.






ซ2
•*
^o

00
•^1-
vd
i — i





wH
CO
*4



^
^
VC
C/3



r-
r~
T
(T

OC
ff
C
oc
O
CO
*e

ir-
on


^N;
c
ff


%.
f.





j\
o\
XT
•oy-

vq,
CO
n





oc
T:
A


OO
VO
OS
^J-





CO
CN
OS
CO
Os

c
CO
V
0
oo
oc
^H
co

CO
r-
o
r-

o
rN
V
en
fW


IL
V
P-
































— .
ฃ
^








Q,

fr


*

^




55




^
55



•rf*
55






55

55

^



j









PC





ฃb




55




^
*



^
55






*

55

^



^3








ฃ
PC






oo
00
•6/5'


•o
o




^.
VC
•ee-



CN
t •
ij




t—
co
ir

oc
c^
CN
oc
oo
t—
OS
CO
T— 1
Vi

oc
CO
"5

o>
••^
{/=



?
PC





^
vo
1
C/5

CO
oo




Q^
VC
^=>


T— H
J
•
cr
vS
1— 1

oo
co




f*^
c-
tfl-


r-
0
i/-^
r-;
f&



Q
O
s"
VD
OS

i — i
00
t— 1
O
$
S.
s
CO
ซซ

OS
VO
co
CN

OS
T-H

6^


tf
P-





oc
VO
CO
fee-

VC
Os
OS
•^
V
in
fe<*
os
O
o
w
VO
fee

?
o

00
*~~

6=


C/"
g-
p-






























Q
JU

c?





C

•^
Cr


ฃt


-
Q.
^




55




^^
*



^^
J5






55

S

55



55








H
ca





53




^




.^
55



^
55






55

55

55



55








E-
PQ





IT)

-------
CO


-------

-------
Wl




U1

-------
       5.1.3   Comparison of Upper-Bound and Retrofit Compliance Costs by Class

       Table 5-5 compares upper-bound (new equipment) and upgrade (retrofit) capital costs by meat
type and process class. Estimating upgrade costs reduces capital investment for options 3 and 4 because
facilities now pay to modify equipment already purchased rather than having to pay the entire cost of a new
piece of equipment. O&M costs, however, are unchanged for options 3 and 4.

        Retrofit has a much larger impact  on costs for direct dischargers than for indkect dischargers.
Overall, upgrade costs are 55 percent lower than new equipment costs under BAT 3, and 63 percent lower
under BAT 4. For indkect dischargers, upgrading capital costs for PSES 3 and PSES 4 are 10 and 9
percent lower than new equipment costs respectively. Within classes, the difference between upper-bound
costs and upgrade costs may vary substantially.
 5.2    FACILITY CLOSURE ANALYSIS

        Facility level closure impacts are estimated using the closure model described in Chapter 3. The
 closure model addresses the impact of compliance costs on the financial health of the individual facility. In
 effect, the closure analysis models the financial evaluation a facility owner might make when deciding
 whether to upgrade pollution controls, or to close the facility because, with pollution controls in place, the
 facility is no longer economically viable.

         In general, because the methodology is based on a cumulative probability distribution (see Section
  3.1.2.1), the relative size of impacts is dkectly related to:

                 the average estimated compliance costs per facility as a percent of cash flow in a
                 subcategory or meat type and process class, and
                 the number of facilities in the subcategory  or meat type and process class.

  As per facility costs as a percent of cash flow increase, so will the incremental probability of closure.  As
  the number of facilities in a subcategory or meat type and process class increase, so will the number of
                                                 5-26

-------
                    Table 5-5
Comparison of Upper-Bound and Retrofit Capital Costs
                                                Differencem
                                                Capital Costs
\Kfd Meat Firrt Process"" (Suhrateqorv A - D) 	 _^
6 I
I
]
3
Red Meat Fi
12 ]
]



168



JAT1
?AT2
3AT3
3AT4
iriher Pro
BAT1
BAT2-
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
$0
$0
$0
$4,805,019
$0
$0
$0
$800,836
NA
NA
$0
$0
NA
NA
$0
$0

NA
NA
$0
$4,805,019
cessing (Subcategory E - I) i 	 , 	 — r
$0
$45,683
$247,412
$12 693 792
$0
$3,807
$20,618
.$1,057,816
NA
NA
$111,335
$160,818
NA
NA
$9,278
$13,402

$39.599,365! $235,711
$206,835,648J
$205,401,202
$289,011,365
$1,231,165
$1,222,626
$1,720,306
NA
NA
$205,401,202
$289,011,365
NA
NA
$1,222,626
$1,720,306
NA
NA
$136,077
$12,532,974

NA
NA
$0
$0
Red Meat First an^ Further Processing (Subcategory A - D) 	 r__ 	 ,
28



PSES1
PSES2
PSES3
PSES4
Red Meat First Proce
36




15



Red Meat
4



BAT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
Further P
BAT1
BAT2
IBAT3
BAT4
$7,674,552
$109,691,736
$105,932,768
$110,184,632
$274,091
$3,917,562
$3,783,313
$3,935,165
NA
NA
$91,985,869
$99,994,413
NA
NA
$3,285,210
$3,571,229
NA
NA
$13,946,899
$10,190,219
ssing and Rendering (Subcategory A - D) 	 , 	
$0
$6 252 839
$269,463,940
$312 997 176
$0
$173,690
$7,485,109
$8,694,366
• • NA
NA
$121,258,772
$175,151,561

$9 946,909
$311,479,620
$210,194,072
$211,683,958
$663,12'
$20,765,308
$14,012,938
$14,112,26'
NA
NA
$210,194,072
$211,683,958
NA
NA
$3,368,299
$4,865,321

NA
NA
$14,012,938
$14,112,26'
rnr.essin.% and Rendering (Subcategory E- 1) 	 	
$(
$86,86'
$263,93(
$13,428,16:
) $C
1 $21,71'
) $65,98:
I $3,357,041
i NA
7 NA
.NA
L NA
I $118,7691 $29,69:
} $171,55^
5 $42,88<
NA
NA
$148,205,168
$137,845,615

NA
NA
$0
$C

L NA
L NA
I $145,161
) $13,256,60'

NA
NA
0.00%
100.00%

NA
NA
55.00%
98.73%

NA
NA
0.00%
0.00%

NA
NA
13.17%
9.25%

. NA
NA
55.00%
44.04%

NA
NA
0.00%
0.00%

NA
L NA
L 55.00%
1 98.72%||
                         5-27

-------
                 Table 5-5 (cont.)
Comparison of Upper-Bound and Retrofit Capital Costs
Number
of
Facilities
7




Option
PSES1
PSES2
PSES3
PSES4
UPPER-BOUND
Total
Capital Costs
$3,588,406
$37,076,732
$30,127,418
$34,521,628
Average
Capital
Costs
$512,629
$5,296,676
$4,303,917
$4,931,661
RETROFIT
Total
Capital Costs
NA
NA
$26,398,992
$31,268,069
Average
Capital
Costs
NA
NA
$3,771,285
$4,466,867
Difference in
Capital Costs
NA
NA
$3,728,426
$3,253,559
Percent
Difference
NA
NA
12.38%
9.42%
Red Meat First Processing, Further Processing, and Rendering (Subcategory A - D)
24



BAT1
BAT2
BATS
BAT4
$0
$1,993,987
$5,172,769
$249,497,464
$0
$83,083
$215,532
$10,395,728
NA
NA
$2,327,746
$3,362,300
NA
' NA
$96,989
$140,096
NA
NA
$2,845,023
$246,135,164

17



PSES1
PSES2
PSES3
PSES4
$14,504,126
$203,365,424
$144,061,380
$280,904,584
$853,184
$11,962,672
$8,474,199
$16,523,799
NA
NA
$72,030,690
$161,805,662
NA
NA
$4,237,099
$9,517,980
NA
NA
$72,030,690
$119,098,922
Poultry First Processing (Subcategory K)
49




BAT1
BAT2
BATS
BAT4
BATS
$0
$0
$97,162,006
$130,989,236
$146,285,848
$0
$0
$1,982,898
$2,673,250
$2,985,425
NA
NA
$43,722,902
$63,155,304
NA
NA
NA
$892,304
$1,288,884
NA
' NA
NA
$53,439,104
$67,833,932
NA

92



PSES1
PSES2
PSES3
PSES4
$33,447,312
$406,506,200
$351,742,064
$376,110,848
$363,558
$4,418,546
$3,823,283
$4,088,161
NA
NA
$332,158,512
$362,017,073
NA
NA
$3,610,419
$3,934,968
NA
NA
$19,583,552
$14,093,775
Poultry Further Processing (Subcategory L)
13




BAT1
BAT2
BATS
BAT4
BATS
$0
$142,827
$10,898,624
$15,381,507
$17,719,557
$0
$10,987
$838,356
$1,183,193
$1,363,043
NA
NA
$4,904,381
$7,084,105
NA
NA
NA
$377,260
$544,931
NA
NA
NA
$5,994,243
$8,297,402
NA

155



PSES1
PSES2
PSES3
PSES4
$36,434,378
$236,758,364
$201,922,369
$271,880,434
$235,061
$1,527,473
$1,302,725
$1,754,067
NA
NA
$201,922,369
$271,880,434
NA
NA
$1,302,725
$1,754,067
• NA
NA
$0
$0
NA
NA
55.00%
98.65%

NA
NA
50.00%
42.40%

NA
NA
55.00%
51.79%
NA

NA
NA
5.57%
3.75%

NA
NA
55.00%
53.94%
NA

NA
NA
0.00%
0.00%
                        5-28

-------
                 Table 5-5 (cont.)
Comparison of Upper-Bound and Retrofit Capital Costs
Number
of
Facilities <
^^ssss^=s=
Dntion
UPPER-BOUND
Total
Capital Costs
Average
Capital
- '•'':" - CoStS
RETROFIT
,-'•} Toiai
Capital Costs
Average
Capital
-''• r Costs •
Difference in
Capital Costs
Percent
Difference
Poultry Fir-it and Further Processing (Subcategory K) 	 	 	 , 	
16





29



3AT1
BAT2
BATS
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
$0
$1,018,875
$37,748,307
$60,619,846
$67,733,811
$0
$63,680
$2,359,269
$3,788,740
$4,233,363
NA
NA
$16,986,738
$24,536,399
NA
NA
NA
$1,061,671
$1,533,525
NA

$0
$96,159,047
$116,164,392
$122,980,483
$0
$3,315,829
$4,005,669
$4,240,706
NA
NA
$82,391,629
$94,558,645
NA
NA
$2,841,091
$3,260,643
NA
NA
$20,761,569
$36,083,447
NA

NA
NA
$33,772,763
$28,421,838
NA
NA
55.00%
59.52%
NA

NA
NA
29.07%
23.11%
[Poultry First Processing and Rendering (Subcategory K) •_ 	 	
17





r



BAT1
BAT2
BATS
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
$0
$466,032
$47,375,431
$61,101,413
$68,021,691
$0
$27,414
$2,786,790
$3,594,201
$4,001,276
NA
NA
$21,318,944
$30,794,030
NA
NA
NA
$1,254,056
$1,811,414
NA

$0
$46,412,547
$29,064,025
$30,098,859
$0
$9,282,509
$5,812,805
$6,019,772
NA
NA
$29,064,025
$30,098,859
. NA
NA
$5,812,805
$6,019,772
NA
NA
$26,056,487
$30,307,383
NA

NA
NA
$0
$0
Poultry Further Processing and Rendering (Subcategory L)
15



PSES1
PSES2
PSES3
PSES4
Poultry First Proces
6




BAT1
BAT2
BATS
BAT4
BATS
$2,640,352
$45,671,602
$38,125,831
$40,626,708
$176,023
$3,044,773
$2,541,722
$2,708,44'7
NA
NA
$35,359,327
$38,711,023
NA
NA
$2,357,288
$2,580,735
sing Further Processing, and Rendering (Subcategory K)
$c
$(
$38,990,37(
$40,129,511
$45,039,29^
) $C
) • $(
) $6,498,39f
L $6,688,25:
t $7,506,54<
i "^f^
) NA
i $17,545,66-y
> $25,343,74:
) N/
NA
NA
' .$2,924,27!
$4,223,95'

NA
NA
$2,766,504
$1,915,685

NA
L N^
5 $21,444,702
1 $14,785,77(
L NP
NA
NA
55.00%
49.60%
NA

NA
NA
0.00%
0.00%

NA
NA
7.26%
4.72%

NA
NA
55.00%
) 36.85%

                         5-29

-------
                 Table 5-5 (cont.)
Comparison of Upper-Bound and Retrofit Capital Costs
Number
of
Facilities <
12 I
L_J^_

	 i
	 i
Mixed Furtl




11
fi 97


1
Rendering (
11 21
i 	
8
1

1 75


1
\Total Cost.
209



101 '

715

I
11
=P
Jption
'SES1
3SES2
'SES3
?SES4
lerProces
BAT1
BAT2
BATS .
BAT4

PSES1
PSES2
PSES3
PSES4
'Subcategi
BAT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
UPPER-BOUND
Total
Capital Costs
$8,960,599
$222 320,423
• $140,102,742
$141,530,779
Average
Capital
Costs
$746,717
$18,526,702
$11,675,228
$11,794,232
, * , ... . ,- ^, * • , ' ••___' . . -
RETROFIT
' ••, Total
Capital Costs
NA
NA
$132,094,302
$138,953,449
Average
Capital
Costs
NA
NA
$11,007,858
$11,579,454
miP (fil nercent Subcatesorv E - 1, 39 percent Subcategory L
$0
$30,519
$3,205,753
$9,742,008
$0
$6,104
$641,151
$1,948,402
NA
NA
$1,442,589
$2,083,739
NA
NA
$288,518
$416,748

$30,400,918
$237,813,392
$204,321,312
$337 282 624
$313,412
$2,451,684
$2,106,405
$3,477,140
NA
NA
$204,321,312
$337,282,624
' NA
NA
$2,106,405
$3,477,140
yrvJ) 	
$0
$0
$24 235 794
$27 388 270
$0
$0
$1,154,085
$1,304,203
NA
NA
$10,906,107
$15,753,267
NA
NA
$519,338
$750,156

$3,497,420
$82 708,839
$121,046,542
$130,924,926
$46,632
$1,102,785
'$1,613,954
$1,745,666
NA
NA
$78,857,861
$92,106,957
NA
NA
$1,051,438
$1,228,093
Difference in
Capital Costs
NA
NA
$8,008,440
$2,577,330

NA
NA
$1,763,164
$7,658,269

NA
NA
$0
$0

NA
NA
$13,329,687
$11,635,003

NA
NA
$42,188,681
$38,817,969
Percent
Difference
NA
NA
5.72%
1.82%

NA
NA
55.00%
78.61%

NA
NA
0.00%
0.00%

NA
NA
55.00%
42.48%

NA
NA
34.85%
29.65%
y Erch"Hnp fi5 dertaintv Facilities 	 __, 	 — , 	 —
BAT1
BAT2
BATS
BAT4
BAf5

PSES1
PSES2
PSES3
PSES4
$0
$10,037,629
$534,764,336
$938,773,404
$344,800,201
$C
$48,02^
$2,558,681
[ $4,491,73$
$3,413,862
NA
NA
$240,643,95(
> $347,596,815
J NA
NA
NA
$1,151,406
$1,663,142
NA

$190,694,33'
$2 242 799,57^
$1,898,206,11'
$2,377,741,82!
1 $266,70f
1. $3,136,78:
J $2,654,83'
5' $3,325,51:
j N^
5 N/
[ $1,702,180,16:
3 $2,159,372,53
. NA
L NA
I $2,380,67:
$3,020,10
NA
NA
$294,120,386
$591,176,58f
NA

L NA
L N/
I $196,025,95'
. $218,369,29'
NA
NA
i 55.00%
62.97%
L NA

L NA
L NA
3 10.33%
7 9.18%fl
                        5-30

-------
                                         Table 5-5 (cont.)
                       Comparison of Upper-Bound and Retrofit Capital Costs
Slumber
of
Facilities
Optioji
UCTER-BO^^
Total
Capital Costs
Average
Capital
Costs
RETROFIT
Total
Capital Costs
Average
; Capital
•• •.'•'••' ' ' -.Costs
Tntnl C.nxtx Including 65 Certaintv Facilities
226




BAT1
BAT2
BATS
BAT4 .
BAT5
$0
$10,840,639
$577,545,483
$1,013,875,276
$372,384,217
$0
$51,869
$2,763,376
$4,851,078
$3,686,972
NA
NA
$259,895,466
$375,404,565
NA
NA
NA
$1,243,519
$1,796,194
NA
Difference in
Capital Costs

NA
.NA
$317,650,017
$638,470,712
NA

772



PSES1
PSES2
PSES3
PSES4
$205,949,884
$2,422,223,540
$2,050,062,606
$2,567,961,174
$288,042
$3,387,725
$2,867,220
$3,591,554
$0
$0
$1,838,354,575
$2,332,122,333
$0
$0
$2,571,125
$3,261,710
$205,949,884
$2,422,223,540
$211,708,031
$235,838,841
•% • ,' " " :
Percent
Difference

NA
NA
55.00%
62.97%
NA

100.00%
100.00%
10.33%
9.18%
1 Option BAT 5 is only found in Poultry operations.
                                                5-31

-------
1 incremental closures for a given probability of closure.  Because the number of projected closures is so
 directly related to the number of establishments in a category, this presentation will focus on the ratio of
 compliance costs to net income and the probability that posttax compliance costs exceed cash flow, rather
 than the absolute number of closures. These measures can be directly compared between subcategories and
 classes to get a sense of the relative magnitude of impacts.

         Section 5.2.1 below outlines impacts by subcategory and Section 5.2.2 does the same by meat type
 and process class.  Results presented include pretax and posttax annualized compliance costs per facility,
 the ratio of compliance costs to model facility net income and cash flow, the probability that cash How is
 less than compliance costs, and finally, projected incremental facility closure and employment impacts.3
         5.2.1   Projected Closure Impacts by Subcategory
         5.2.1.1  Upper-Bound Cost Closure Impacts

         Table 5-6 presents a summary of facility closure and employment impact results by subcategory
  groupings, discharge type, and technology option.  For direct dischargers, facilities in Subcategory J have
  the highest probability of closure under BAT 4: 1.6 percent. Given that there are 21 facilities in this
  subcategory, 0.3 facilities are projected to close under this option. Although facilities in Subcategory K
  have a lower probability of closure under BAT 5 (about 1 percent), with 88 facilities in the subcategory, 1
  closure is projected, the largest impact among the direct dischargers.  For the proposed direct discharging
  options, BAT 3 for all subcategories except J for which the proposed option is BAT 2, the ratio of
  compliance costs to net income and the incremental probability of closure in each subcategory is as follows:
                  Subcategory "A through D:
                  Subcategory E through I:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
1.90 percent
0.34 percent
0.40 percent
0.06 percent
          3 Closure impacts under alternative assumptions about the cumulative distribution function can be found
   in Appendix E.
                                                  5-32

-------
                Table 5-6
Economic Closure Impacts: Upper-Bound Costs
         40 CFR 432 Subcategories
— p
1 Option
Subcategt
[BATI
BAT2
BAT3
BAT4

PSES1
PSES2
PSES3
PSfeS4
=^==r=
Number '
of
Facilities
Annualized
Compliance Costs
per Facility *
Pretax
Posttax
================p
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Less Than
Compliance
Costs3
Projected
Facility Impacts 4
Closures
try A through D 	 , 	 , 	 -r
66




60



$0
$139,344
$835,010
$1,655,105

$108,802
$2 337 820
$1,485,337
$1,861,723
$0
$83,256
$550,223
$1,095,962

$71,591
$1,521,794
$982,758
$1,238,299
0.00%
0.28%
1.90%
4.11%

0.57%
10.35%
7.21%
8.14%
0.00%
0.25%
1.66%
3.58%

0.44%
8.09%
5.59%
6.39%
0.00%
0.05%
0.34%
0.74%

0.09%
1.73%
1.19%
1.36%
0.0
0.0
0.2
0.5

0.0
1.1
0.6
0.7
Employment

0
0
318
794

0
1,230
609
768
Subcategorv E throueh I - 	 '• 	 • 	
BATI
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
Subcatei
BATI
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
19




234



s>oryj
21




7^



$0
$19,641
$33,648
$340,790

$74,306
$403,679
$330 879
$435,725
$0
$11,626
$21,782
$224,821

$47,519
$262,073
$217,257
$289,705
0.00%
0.14%
0.40%
2.91%

0.80%
' 4.53%
3.72%
5.06%
0.00%
0.12%
0.33%
2.44%

0.67%
3.77%
1 3.09%
4.21%
0.00%
0.02%
0.06%
0.46%

0.13%
0.72%
0.59%
0.81%

. $0
$24,340
$255,876
$278 194

$16,406
$287 088
$344,581
$360,74")
$0
$14,458
$168,926
$184,386

. $10,425
$186,712
$228,36f
$239,901
0.00%
0.68%
8.03%
8.78%

ป 0.50%
\ 8.78%
> 10.79%
[ 11.36%
0.00%
0.56%
6.55%
7.16%

0.41%
7.13%
8.78%
> 9.25%
0.00%
0.12%
1.45%
1.59%

0.09%
1.58%
> 1.95%
, 2.06%
0.0
0.0
0.0
0.0

0.3
1.8
1.3
1:9

0.0
0.0
0.3
0.3

0.0
1.2
, 1.5
p l.C
0
0
0
0

91
495
346
492

0
0
14
14

0
66|
81
> : 89]]
                    5-33

-------
BBATI
|BAT2_
pAri"
[JAT4^
BEATS

 IgSESl
 |PSES2_
 IPSESS
Number
      of
Facjlities
 >K__
      88
      138
        15
  BBAT2
  |JBAT3^
  JBAT4
  |BAT5~

  PSESI
  |PSES2
                                 Table 5-6 (cont.)
                    Economic Closure Impacts: Upper-Bound Costs
                             40 CFR 432 Subcategones
              Annualized
            Compliance Costs
              per Facility1
       208
  _    $0
   $50.7621
  $508.9591
  $644.469
  $695.432J

_J$72.738
"tr.267.800l
   $892,461
  ^$916.1361

       loi
    $18.6781
   ^$182.548
   $267,851
   JS274.471

     $67.967
    $469.256
   _ $332,1991
    $419,2711
                         Compliance Cost
                          as a Percentage
                         ซf Model Faculty 21
      $0
_$29.922J.
1335,2371
'$426,6571
'$462.287

  147,101
  $824,567
 l590,677|
  $608,171|

       "$ol
   $11,2031
  •$119,997
   $177,456
   $182,451

   '$43.876
   $304.357
   $219,332J
   $279.7691
 0.00% I
 0.34%
j.98%
J5.14%1
1.61%

 ^0.55%
 1.71%
  6.53%
  Q.00%
 1.39%
  4.23%
 1.04%
  6.71%]

   1.50%)
   9.63%|
   7.00%
   8 96%|
 0.00%l
_0.27%
1.20%
 4.13%
 4.50%

 0.43% I
 6.95%
Probability
 Cash Flow
 Less Than
Compliance
     Costs3
        •••—

     0.00% I
     0.06%1
     0.72%^
      0.93% I
  Q.00%
  0.32%l
  3.54%
  5.04%
   1.26% .
   8.06%
   5.87% I
      1.59%]

      1.23%]

      0.00%]
      0.07%[
       0.77%[
                                                    Projected
                                                 facility Impacts4:
       0.27%l
        1.75%[
        1.27%|
                                                                             0.6
                                                                             Tel
                                                                                        1.20811

-------
                                                  Table 5-6 (cont.)
                                  Economic Closure Impacts: Upper-Bound Costs
                                             40 CFR 432 Subcategorifis
=====
"otal Inc
RAT1
BAT2
RAT3
RAT4
RAT5

PSES1
PSES2
PSFS3
PSES4
All impac
weighted
1 Total an
2 Ratio of
3 Probabil
=====
Number
of
^Facilities

226





772



========f
Annualized
Compliance Costs
per Facility l
: 1
Pretax Posttax
'
Compliance Cost
as a Percentage
of Model FacUity2
Net Income
Cash Flow
=========
Probability
Cash Flow
Less Than
Compliance
Costs3
ertainty Facilities 	 	
NA
• NA
NA
NA
NA

NA
NA
NA
•NTA
*•••••-
:s presented in this table are t
>y the number of facilities in
nualized compliance costs for
posttax annualized complianc
ity net income or cash flow le
NA
NA
NA
NA
NA

NA
NA
NA
NA
======
le average of res
each combinatio
subcategory anc
e costs to net in<
ss than posttax ฃ
NA
NA
NA
NA
NA

NA
• NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
• NA
NA
NA
NA

NA
NA
NA
' NA
Projected
Facility Impacts 4
Closures

0.0
0.0
1.2
1.7
1.1
1.1
10.6
.8.1
9.Q
Employment

0
0
662
1,470
656

240
4,531
2,736
3,391
ults for each subcategory, discharge type and model facility, size combination,
n.
1 discharge class divided by number of facilities in that class.
:ome and cash flow. • '
mnualized compliance costs minus probability net income or cash flow less than
zero.
4
 Closures- probability cash flow less than annualized compliance costs multiplied by the number of facilities in the subcategory.
ป Opt^BAT 5 is onry found in Poultry operations. Subcategory L includes poultry further operations and mixed further operations. The
couS for BAT 5 is for poultry further operations only and hence, the number of facilities is smaller than for other BAT options.
                                                           5-35

-------
               Subcategory J:

               Subcategory K:

               Subcategory L:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
0.68 percent
0.12 percent
3.98 percent
0.72 percent
4.23 percent
0.77 percent
Projected closure impacts total about 2 facilities under the proposed option with associated employment
losses of about 600 workers. The largest impacts measured in terms of the highest ratio of compliance cost
to net income and the highest incremental probability of closure occur in Subcategory L. The largest
closure impacts occur in Subcategory A through D because there are four times more establishments in that
Subcategory than Subcategory L.               .

        For indirect dischargers, Subcategory L incurs the largest impacts with 3.6 projected incremental
closures uivier PSES 2.  However, Subcategory J actually has a higher cost to net income ratio and a
higher incremental probability of closure under PSES 4. Larger impacts are projected for Subcategory L
because there are a total of 208 facilities in Subcategory L as opposed to 75 in Subcategory J.  In general,
impacts to indirect dischargers are larger than impacts to direct dischargers for each option.  This is
because: (1) indirect dischargers tend to incur higher compliance costs per facility resulting in a higher
incremental probability of closure, and (2) there are usually more indirect dischargers than direct
dischargers in each subcategory.
        5.2.1.2  Upgrade Cost Closure Impacts

        Since costs for upgrading are lower than new equipment costs under options 3 and 4, generally
 closure impacts for the upgrade scenario are lower than under the upper-bound cost estimates presented
 above.  There will generally be lower cost to net income ratios, lower incremental probabilities of closure,
 and lower projected closure impacts.

        A summary of facility closure and employment impact results using upgrade costs by subcategory
 groupings, discharge type, and technology option is presented in Table 5-7. For direct dischargers,
                                               5-36

-------
              Table 5-7
Economic Closure Impacts: Retrofit Costs
       40 CFR 432 Subcategories
Option

Number •
of
Facilities
Annualized
Compliance Costs
per Facility *
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
i^ess man
Compliance
Costs 3
Projected
Facility Impacts 4
Closures
Employment
Suhnategorv A through D 	 , 	 	 	 , 	 • 	 , 	
BAT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4 '
66




60



NA
NA
$592,740
$1,031,530
NA
NA
$374,326
$643,172
NA
NA
1.30%
2.38%
NA
NA
1.13%
2.07%
NA
NA
0.23%
' 0.43%

NA
NA
$1,333,647
$1,633,619
NA
NA
. $872,626
$1,072.687
NA
NA
6.53%
7.36%
NA
NA
5.05%
5.75%
NA
NA
1.07%
1.22%
.NA
NA
0.1
0.1

- NA
NA
0.6
0.7
NA
NA
159
159

NA
NA
609
768
Subcategorv E through 1 	 	
BAT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
19




234



NA
NA
$26,108
$171,523
NA
NA
$16,269
$101,755
NA
NA
0.29%
1.36%
NA
NA
0.24%
1.14%
NA
NA
0.05%
0.22%

NA
NA
$329,193
$434,254
NA
NA
$216,033
$288,637
NA
.NA
3.71%
5.05%
NA
NA
3.09%
4.20%
NA
NA
0.59%
0.81%
NA
NA
0.0
0.0

NA
NA
1.3
1.9
NA
NA
0
0

NA
NA
346
492
Suhfntp.onrv .7 	 _ _ 	 . 	
BAT1
BAT2
BATS
BAT4

PSES1
PSES2
Hฃ_ 	
PSES3
IJPSES4
21




• 75



NA
NA
$188,683
$219,544
NA
NA
$119,699
$141,417
NA
NA
5.70%
6.74%
NA
NA
4.65%
5.49%
NA
NA
1.02%
1.21%
NA
NA
0.3
0.3

NA
NA
$285,034
$305,958
NA
NA
$184,74C
$199,761
NA
NA
> 8.74%
9.47%
NA
NA
7.11%
7.71%
NA
NA
1.58%
1.71%
NA
NA
1.2
1.2
NA
NA
14
14

NA
NA
66
66
                  5-37

-------
                                      Table 5-7 (continued)
                              Economic Closure Impacts: Retrofit Costs
                                    40 CFR 432 Subcategones
BOfition
IjubcategO:
IBATI
|BAT2
IJAT3
|BAT4
 |BAT5
Number
      of
Facilities
   K_
      88
              138
  IJSES3
  |gsis4
   Sttbcateo
  IBATI
  JBAT2.
  Co
  |BAT4
  BBATJ
   r_
   igSESl
   |gSES2
   HJSES3
   JPSES4
        15
       13s
        208
     NA
     NA
^$362.560
 $465,220
     NA
                 209
             $845,389
             $881.546
	   NA
J135,235
 $187,951
      NA
     —  '-

      NA
      NA.
  $330,790
  $418,296
    _NA_
     NA
	       •*
J228,901
 $296,460
      NA
                 _NA
                  NA
             _$85,410
              $J19,025
                   NA
                  •  -

                   NA.
                   NA
               $218,309
               $279,061
   NA
   NA
_2.73%
 3.56%
   NA

   "NA'
    NA
   —   —
  6.16%
  6.52%
                  NA
                  NA
                3.01%
                4.12%
                   NA

                  _NA
                __NA
                 6.99%
                 8.95%
                                    2.86%
                                     _NA

                                     _NA
                                      NA
                                     ...
                                    Jr.89%
                                     5.17%
                NA
               _NA
             "2.52%
              3.44%
                NA
                 NA
              _5.86%
               7.50%
J).49%
 0.64%
   NA

   JNA
    NA
   -
 J).98%
  1.18%
    NA
  Q.55%_
  0.75%
     NA

     NA
                                                                _
                                                               1.27%
                                                               1.62%

-------
                                                    Table 5-7 (cont.) ;
                                       Economic Closure Impacts: Retrofit Costs
                                               40 CFR 432 Siibcategdries
Option
Number
of
Facilities
Annualized
Compliance Costs
per Facility 1
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
;Net Income
Cash Flow
Probability
Cash Flow
Less Than
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
Tntnl Jnrludine the 65 Ceftaintv Facilities
BAT1
BAT2
BATS
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
226





772



NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.8
1.1
0.0

NA
NA
NA
NA
.NA
NA
NA
NA
NA
• NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
. NA
NA
NA
7.7
9.3
NA
NA
262
490
0

NA
NA
2,678
3,180
weighted by the number of facilities in each combination.
1 Total annualized compliance costs for subcategory and discharge class divided by number of facilities in that class.
2 Ratio of posttax annualized compliance costs to net income and cash flow.
3 Probability net income or cash flow less than posttax annualized compliance costs minus probability net income or cash flow less than zero.
4 Closures: probability cash flow less than annualized compliance costs multiplied by the number of facilities in the subcategory.
5 Option BAT 5 is only found in Poultry operations. Subcategory L includes poultry further operations and mixed further operations.
                                                             5-39

-------
Subcategory K incurs the largest impacts under BAT 4 with an incremental probability of closure of 0.6
percent and 0.5 projected closures: This is, however, 44 percent lower than the largest facility closure
impacts assuming upper-bound costs.  For the proposed direct discharging options, the ratio of compliance
costs to net income and the incremental probability of closure are also lower than in Section 5.2.1.1. They
are as follows:
               Subcategory A through D:

               Subcategory E through I:

               Subcategory J:

               Subcategory K:

               Subcategory L:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
costs / net income:
probability of closure:
1.30 percent
0.23 percent
0.29 percent
0.05 percent
0.68 percent
0.12 percent
2.73 percent
0.49 percent
3.01 percent
0.55 percent
Projected closure impacts are 0.5 facilities under the upgrade cost scenario, with employment losses of
about 230 workers under the proposed options. Note that impacts to Subcategory J are unchanged from
the upper-bound cost estimates because retrofit costs were not estimated for BAT 2.
        5.2.2   Projected Closure Impacts by Meat Type and Process Class

        5.2.2.1 Upper-Bound Cost Closure Impacts

        Table 5-8 summarizes projected facility closure and employment impacts by meat type and process
class, discharge type, as well as technology option. The class level data allows more insight into the range
of impacts projected to occur under the proposed option than does the Subcategory data.  The impacts listed
for each subcategory in Section 5.2.1.1 above actually consist of a weighted average of class level impacts.
Thus, for each subcategory, the overall ratio of compliance costs to net income and the range of those
impacts in component classes is as follows:
                                               5-40

-------
                Table 5-8
Economic Closure impacts: Upper-Bound Costs
       Meat Type and Process Classes
^==s=
Option
RedMea
BAT1
BAT2
BATS
BAT4
1
RedMea
BAT1
BAT2
BATS
BAT4

isESl
SES2
SES3
PSES4
Red Me
PSES1
PSES2
PSES3
PSES4
Red Me
BAT1
BAT2
BATS
BAT4

PSES1
IIPSES2
IiPSESS
HPSES^
=^====
^Jiimlipr
of
Facilities
t First Proc
6


	 1 	 : 	 : 	 • 	 • 	 II
1
Annualized
Compliance Costs
per Facility 1
Pretax
Posttax
========7=
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
=====5=
Probability
Cash Flow
Less Than
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
'fSftin-P (Suhrateqorv A - D) . • 	 -, 	 : 	
$0
$0
$11,374
$184 589
$0
$0
$6,756
$121,398
0.00%
0.00%
0.25%
4.50%
0.00%
0.00%
0.21%
3.74%
0.00%
0.00%
0.04%
0.77%
0.0
0.0
0.0
0.0
0
0
.0

t Further Pr<"-<> ซ•'""<* (Suhratef>orv E - I) 	 __, 	 _, 	 1|
12




168



at First' one
28



at First Pn
36




If



$0
$8,376
$9249
$226,301

$72481
$285 920
$273,780
$348,513
$0
$4,949
$5,712
, $148,065

$45,873
$185,489
$178 271
$229 398
0.00%
0.07%
0.08%
2.19%

0.71%
2.89%
2.78%
3.58%
0.00%
0.06%
0.07%
1.83%

0.60%
2.42%
2.32%
2.99%
0.00%
0.01%
0.01%
0.35%

0.12%
0.47%
0.45%
0.58%
0.0
0.0
0.0
0.0

0.2
0.8
0.7
1.0
0
0
0
0

71
282
247
353
1 further Processing ( Subcatesorv A -D) 	 r 	 , ||
$64,589
$1,035,076
$751,666
$760,945
$41,841
. $663,388
$495,858
$503,560
0.84%
13.31%
9 95%
10.11%
0.60%
9.58%
7.16%
7.27%
0.13%
2.08%
1.55%
1.57%
, 0.0
0.6
0.4
0.4
tce-xxinv and Rendering (Subcategory A - D) 	 	 r 	 	 —
$0
$139 773
$1,475,132
$1,684,423

$134,63')
$4,131,62S
$2,656,20;
$2,610,9(r
$0
$84 239
$974 022
$1,114,43C

r $88 595
) $2 725 09'
S $1,761,91$
1 $1,736,95*
0.00%
0.29%
3.32%
1 3.80%

} 0.30%
> 9 29%
) 6.019?
) 5.929!
0.00%
0.25%
2.91%
3.33%

> 0.269?
> 8.149?
, 5.269!
•> 5.199
0.00%
0.05%
0.60%
0.699?

, 0.059?
-, 1.719!
•> 1.099
9 1.089
0.0
0.0
0.2
, 0.3

•> O.C
3 0.2
9 0.1
0 -0.1
0
436
291
291



318
476

0
476J
1591
1591
                     5-41

-------
                             Table 5-8 (cont.)
                Economic Closure Impacts: Upper-Bound Costs
                       Meat Type and Process Classes
                                                     Probability
                                                     Cash Flow
                                                     Less Than
                                                     Compliance
                                                         Costs3
           i^^^^
           Compliance Cost
           , as a Percentage
           of Model Facili
   Annualized
Compliance Costs
   er Facili
                                                                              projected
                                                                          facility Impacts
                                            Closures! Emplo
                                                          0.00%
                                                          0.03%
                                                          __
                                                          0.02%
                                                          0.48%
 $30,197
 $16,674
$433,757
                                                           0.08%
                                                           —    —
                                                           0.98%
                                                           0.62%
                                                           0.69%
                                                         A-JD
                                                           0.00%
                                                           0.06%
                                                           0.03%
                                                           0.81%
                          0.44%
                          5.27%
                          3.34%
                          3.71%
                         Subcaiego
                           0.00%
                           0.31%
                           0.15%
                           3.92%
                       and Renderin
Red Meat First Prป<*™™. Further Process*
                           Q.00%1
                           0.35%
                           0.17%
                           —.
                           4.47%
          $1,078.756 $1,311,902
                                                0.32%
                                                5.60%
                                                —   ——
                                                3.28%
                                                6.00%
                0.36%
                6.39%
                3.74%
                6.85%
   . Q00.742  $1.873,902
    fifiO.618 $1,097,217
First Processing (Subcatego
                                     0.00%
                                     —   •-
                                     0.24%
                                     __
                                     3.33%
                                     4.31%
                                      4.72%
        $0
   $19,048
   $264,617
   $341,425
   $372,064
         $0
    $32,617
    $402,059
    $515,806
    $560.232
              $84.3681    $54.737
             $952.857   $622,905
             $739.031
             $769,8591  $511,067
                                        5-42

-------
             Table 5-8 ^cont.) •
Economic Closure Impacts: Upper-Bound Costs
       Meat Type arid Process Classes
Option
Number
of
Facilities
Annualized
Compliance Costs
per Facility 1
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Less Than
Compliance
Costs3
Projected
Facility Impacts 4
Closures) Employment
Poultry Further Processing (SubcategoryL)
BAT1
BAT2
BAT3
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
13





155



$0
$18,084
$189,147
$251,412
$274,471

$68,468
$401,506
$289,937
$358,060
$0
$10,853
$124,240
$166,155
$182,451

$44,034
$260,392
$190,988
$238,006
0.00%
0.40%
4.56%
6.11%
6.71%

1.72%
10.20%
7.45%
9.33%
0.00%
0.33%
3.81%
5.10%
5.61%

1.45%
8.59%
6.28%
7.86%
0.00%
0.07%
0.84%
1.13%
1.24%

0.32%
1.91%
1.39%
1.75%
0.0
0.0
0.1
0.1
0.1

0.5
2.9
2.1
2.7
Poultry First and Further Processing (Subcategory K)
BAT1
BAT2
BATS
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
16





29



$0
$50,359
$487,028
$726,500
$786,050

$9,939
$953,462
$800,429
$823,911
$0
$30,367
$319,898
$481,243
$522,705

$5,805
$606,678
$527,679
$544,926
0.00%
0.30%
3.38%
5.12%
5.62%

0.07%
5.92%
5.42%
5.70%
0.00%
0.24%
2.68%
4.06%
4.45%

0.05%
4.73%
4.31%
4.52%
0.00%
0.05%
0.60%
0.92%
1.01%

0.01%
1.07%
0.97%
1.02%
0.0
0.0
0.1
0.2
. 0.2

0.0
0.3
0.2
0.3
Poultry First Processing and Rendering (Subcategory K)
BAT1
BAT2
BAT3
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
17





5



$0
$61,494
$560,984
$703,209
$745,836

$19,013
$2,279,835
$1,142,017
$1,149,785
$0
$36,447
$370,537
$466,017
$497,125

$11,150
$1,474,420
$756,188
$763,895
0.00%
0.49%
5.24%
6.68%
. 7.24%

0.17%
18.27%
10.02%
10.30%
0.00%
0.42%
4.43%
5.65%
6.12%

0.14%
15.45%
8.48%
8.72%
0.00%
0.09%
0.98%
1.25%
1.36%

0.03%
3.50%
1.89%
1.94%
0.0
0.0
0.1
0.2
0.3

0.0
0.1
0.1
0.1

0
0
16
16
16

80
488
360
456

0
0
38
174
174

0
211
174
211

0
0
16
152
168

0
16
16
16
                    5-43

-------
             Table 5-8 (cont.)
Economic Closure Impacts: Upper-Bound Costs
       Meat Type and Process Classes
	 L— —
Onfinn
\Poultrv I

|p** Prnr.essine. and Rendering (Subcategory K) 	 . 	 n
6





12



"wrth&r Pro
5




91



ing (Subca
21



$0
$169,617
$1,293,051
$1,310,040
$1,415,110

$157,724
$4,020,330
$2,187,182
$2,163,118
$0
$99,056
$852,850
$865,627
$939,292

$103,338
$2,626,437
$1,452,861
$1,440,597
0.00%
0.83%
7.38%
7.61%
8.29%

0.82%
19.07%
10.96%
10.97%
0.00%
0.67%
5.96%.
6.14%
6.68%

0.67%
15.69%
8.98%
8.989o
0.009o
0.15%
1.34%
1.38%
1.51%

0.15%
3.58%
2.02%
2.03%
0.0
0.0
0.0
0.0
0.0

0.0
0.5
0.2
0.2
cessing (61 percent in Sปh™te<,arv E-1.39 percent in Subcategory L) ^ 	
$0
$22,640
$138,552
$377,450

$74,822
$622,276
$431,45C
$623,29C
tegory J)
( $(
$24,34(
$255,87<
$278,19'
$0
$13,538
$91,709
$252,797

$49,043
$405,605
$287,192
$421,25S
0.00%
0.30%
2.03%
5.60%

1.09%
8.99%
6.37%
9.34%
0.00%)
0.25%
1.68%
4.64%

0.90%
7.44%
5.27%
7.73%
0.00%
0.05%
0.32%.
0.88%

0.17%
1.42%
1.00%
1.47%

) $(
) $14,45!
j $168,92(
I $184,38
) 0.00%
3 0.689
5 8.039
5 8.789
9 0.00%
9 0.569
9 6.559
'0 7.169
, 0.009
, 0.129
o 1.459
0 1.599
0.0
0.0
0.0
0.0

0.2
.1.4
1.0
1.4

O.C
D 0.(
b o.:
b o.:
Employment

0
174
38
38
	
0
0
0
0
0

0
582
174
1741

0
0
0
0


228
163
228

) 0
) 0
3 14
3 14
                     5-44

-------
                                        Table 5-8 (cont.)
                         Economic Closure Impacts: Upper-Bound Costs
                                 Meat Type and Process Classes
— .—
fer[
PSES2
PSES3
PSES4
Total Ex
BAT1
BAT2
BATS
BAT4
BATS

FS1
PSES2

PSES4
Total In
BAT1
BAT2
BATS
BAT4
BATS

PSES1
PSES2
PSES3
PSES4
=====
All impa
========
dumber
of
Facilities
75



eluding 65
209



101 5

715



eluding 65
226





772


=========
cts presente
=================
Annualized
Compliance Costs
per Facility 1
Pretax
$16,406
$287 088
$344,581
$360,747
Posttax
$10,429
$186,713
$228,365
$239,901
1
Compliance Cost
as a Percentage
of Model Facility2
Net Income
0.50%
8.78%
10.79%
11.36%
Cashflow
0.41%
7.13%
8.78%
. 9.25%
============
Probability
Cash Flow
Less Than
Compliance
Costs3
0.09%
1.58%
1.95%
2.06%
Projected
Facility Impacts 4
Closures
0.0
1.2
1.5
1.6
Employment
0
66
81
89
Certainty Facilities J 	 , 	 ,- 	 — r— 	 II
NA
NA
NA
NA
NA

NA
NA
NA
NA
NA
NA
NA
NA
NA

NA
NA
NA
NA
NA
	 NA
NA
NA
NA

	 NA
NA
NA
NA
NA
NA
NA
' • NA
NA

NA
NA
• NA
- . NA
Cfrfninly Facilities 	 : 	
NA
NA
NA
NA
NA

NA
NA
• NA
NA
i in this table a
NA
NA
NA
. NA
NA

NA
NA
NA
NA
=====
re the average
NA
NA
NA
NA
NA

, NA
k. 	 NA
L NA
*. NA
of results for ej
NA
NA
NA
NA
NA

NA
NA
NA
NA
ich meat type ai
NA
NA
' NA
NA
NA

NA
NA
NA
NA

NA
NA
NA
NA
NA
—
NA
NA
NA
•NA
=====
id process class
[nation.
0.0
0.0
1.1
1.6
1.0

1.0
9.8
7.5
9.2

0.0
0.0
1.2
1.7
1.1

1.1
10.C
8.1
9.S
, discharge tj
0
0
613
1,361
607

222
4,195
2,533
3,140

0
0
1 662



240
4,531
2,736
3,391|
/pe and model
                          less dm annuafed complimce coso nmltiplMb, Ita numb™ of faciMes m the
subcategory.
5 Option BAT 5 is only found in Poultry operations.
                                                5-45

-------
               Subcategory A through D:
               — red meat first processing, further
                       processing, and rendering
               — red meat first processing and
                       rendering

               Subcategory E through I
               — red meat further processing
               — mixed further processing

               Subcategory, J:
               — rendering

               Subcategory K:
               — poultry first processing
               — poultry first processing, further
                       processing, and rendering

               Subcategory L:
               —     mixed further processing
               —     poultry further processing
costs / net income:
costs / net income:
costs / net income:
costs / net income:
costs / net income:
1.90, percent
0.17 percent

3.32 percent
0.40 percent
0.08 percent
2.03 percent

0.68 percent
3.98 percent
3.33 percent
7.38 percent
4.23 percent
2.03 percent
4.56 percent
The largest ratio of compliance costs to net income under the proposed options is projected in the poultry
first processing, further processing, and rendering class (7.38 percent — Subcategory K), followed by

poultry first processing and rendering (5.24 percent — Subcategory K), and poultry further processing
(4.56 percent — Subcategory K).
        5.2.2.2 Upgrade Cost Closure Impacts


        Table 5-9 summarizes projected facility closure and employment impacts based on upgrade costs

by meat type and process class, discharge type, and technology option.


        Under the proposed options, (BAT 3 for all classes except rendering and BAT 2 for rendering),

there are a total of 0.4 facility closures projected with employment losses totaling 229 for all classes

combined. Comparing the range of disaggregated class level cost to net income ratio for the proposed
option with the Subcategory level ratio:
                                              5-46

-------
              Table 5-9
Economic Closure Impacts: Retrofit Costs
     Meat Type and Process Classes
1 Option
RedMea
BAT1
BAT2
BATS
BAT4
RedMea
BAT1
BAT2
BATS
5BAT4
"•"•

|j 	
PSES1
PSES2
PSES3
PSES4
Red Me
PSES1
pQ-pCO
PSES3
PSES4
LRe^Mg
BAT1
BAT2
BAT3
BAT4

PSES1
PSES2
PSES3
PSES^
=^^=^=p
Number ~
of
Facilities
t First Proi
6



	 ~
Annualized
Compliance Costs
per Facility 1
Pretax
Posttax
1
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
'Liess Than
Compliance
Costs 3
Projected
Facility Impacts 4
Closures
wisine •(Suhcaiegorv A - D) 	 , 	 r 	 r
NA
NA
$11,374
$99 815
NA
NA
$6,756
$59,290
NA
NA
0.25%
2.20%
NA
NA
0.21%
1.83%
NA
NA
0.04%
0.38%
NA
NA
0.0
0.0
t Further Prnr.p.fmine (Subcatesorv E- 1) 	 , 	 r
12




168



at First am
28



at First Pn
36




If



NA
NA
$8,049
$115,742

NA
NA
$273 780
$348,513
NA
NA
$4,840
$67,795

NA
NA
$178,271
$229 398
NA
NA
0.07%
0.99%

NA
NA
2.78%
3.58%
NA
NA
0.06%
0.83%

NA
NA
2.32%
2.99%
NA
NA
0.01%
0.16%

NA
NA
0.45%
0.58%
NA
NA
0.0
0.0
Employment

NA
NA
0
0

NA
NA
0
0

• NA
NA
0.7
1.0
NA
NA
247
353
1 Further Processing (Subcategory A - D) 	 	 i 	 , 	 	
NA
NA
$698 938
$722 420
NA
NA
$457,576
$475,589
NA
NA
9.18%
9.54%
NA
NA
6.61%
6.87%
NA
NA
1.43%
1.48%
NA
NA
0.4
0.4
NA
NA
291
291
irr.m'w nnA Rendering (Subcatesorv A- D) 	 . 	 , — _ 	 	
NA
NA
$1,039,337
$1,279,089

NA
NA
$2,656,202
$2,610,90'
NA
NA
' $657,618
$820,142

NA
NA
\ $1,761,91<
i $1,736,95?
NA
NA
224%
2.80%

NA
L • NA
) 6.019?
) 5.929?
NA
NA
1.96%
2.45%

NA
NA
j 5.2695
3 5.1991
NA
NA
0.40%
0.51%

L NA
L • NA
5 1.099?
, 1.089?
NA
NA
0.1
0.1

NA
NA
> 0.1
0.1
NA
NA
159
159

. NA
NA
159
159||
                   5-47

-------
                                       Table 5-9 (cont.)
                            Economic Closure Impacts: Retrofit Costs
                                 Meat Type and Process Classes
                                         Compliance Cost
                                          as a Percentage
                                         of Model FacilV"2
                    Annualized
                 Compliance Costs
                       Facili  1
Probability
 Cash Flow
 Less Than
Compliance
    Costs3
   Projected
Facility Impacts
                                       TSfet Tncomel Cash Flow
Rod Meat Further Processins and Renderin
,.m i   ill"    "   I "    *~             _ _
                                           nd Rendering (Subcate
•R^M^at First Processing, Furt er Process*
  Poultry First Processing (Subcat
                                                 5-48

-------
           Table 5-9 (cont.)
Economic Closure Impacts: Retrofit Costs
     Meat Type and Process Classes
Option

Number "
of
Facilities
=======Tf=
Annualized
Compliance Costs
per Facility J
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Less 1 nan
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
Poultry Further Processing (Subcategory L) 	 _, 	 __, 	
BAT1
BAT2
BATS
BAT4
BAT5

PSES1
PSES2
PSES3
PSES4
13





155



NA
NA
$140,337
$183,847
NA

NA
NA
$289.937
$358,060
NA
NA
$88,567
$116,777
NA

NA
. • NA
$190,988
$238,006
NA
NA
3.25%
4.29%
NA

'NA
NA
7.45%
9.33%
.NA
NA
2.72%
3.59%
NA

NA
NA
6.28%
7.86%
NA
NA
0.60%
0.79%
NA

NA
NA
1.39%
1.75%
NA
NA
0.1
0.1
NA

NA
NA
. 2.1
2.7
NA
NA
16
16
NA

•NA
NA
360
456
Poultry Firxt and Further Processing (Subcategory K) 	 __, 	 ,__ 	 . —
BAT1
BAT2
BATS
BAT4
[BATS

PSESl
PSES2
PSES3
PSES4
Poultry
BAT1
BAT2
BATS
BAT4
BAT5

PSESl
PSES2
PSES3
PSES4
	 16





29



First Proc
17





<



NA
NA
$349,667
$487,768
NA
NA
NA
$220,169
$307,915
NA
NA
NA
2.34%
3.30%
NA
NA
NA
1.85%
2.62%
NA
NA
NA
0.42%
0.59%
NA

. NA
NA
$677,150
$720,163
NA
NA
$438,173
$469,602
NA
NA
4.50%
4.88%
NA
NA
3.58%
3.88%
NA
NA
0.81%
0.88%
essing and Rendering (Subcategory K) 	
NA
NA
$398 732
$514,487
NA
NA
NA
$252,506
$328,714
NA
NA
NA
3.59%
4.70%
NA
NA
NA
3.04%
3.98%
NA
NA
NA
0.67%
0.88%
, , NA

NA
NA
$1,142,01'
$1,149,78:
NA
NA
' $756,18*
> $763,89f
NA
NA
> 10.029?
> 10.309?
NA
L- NA
3 8.489?
3 8.729!
NA
L NA
3 1.899?
3 1.949?
NA
NA
0.0
0.1
NA

NA
NA
0.2
• 0.2

NA
NA
0.1
0.1
NA

NA
NA
3 0.1
3 0.1
NA
NA
0
38
NA

NA
NA
174
174

NA
NA
16
16
NA

NA
NA
16
16
                   5-49

-------
           Table 5-9 (cont.)
Economic Closure Impacts: Retrofit Costs
     Meat Type and Process Classes
n
Option
\Poultrv 1
PSES1
PSES2
PSES3
PSES4
Pew/fry j
BAT1
IBAT2
BAT3
HBAT4
JBAT5

I] 	 .
PSES1
HPSES2
UPSES3
|PSES4 , 	 , 	 	
m':,, .JvTi1 ~*^^^^**^^p"
Number '
of
Facilities
_ i
Annualized
Compliance Costs
per Facility 1
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Less i nan
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
further Processing and Rendering (Subcategory L) 	 __i 	 — , 	
15



NA
NA
$499,078
$523,737
NA
NA
$326,905
$344,666
NA
NA
3.81%
4.11%
NA
NA
3.00%
3.23%
NA
NA
0.68%
0.73%
NA
NA
0.1
0.1
NA
• NA
38
38
First Procfssin" Further Processing, and Rendering (Subcategory K)
6





12



NA
NA
$914,703
$1,049,175
NA
NA
NA
$578,155
$676,229
NA
NA
NA
5.02%
5.89%
NA
NA
NA
4.05%
4.76%
NA
NA
NA
0.91%
1.07%
NA

NA
NA
$2,116,535
$2,140,383
NA
NA
$1,401,569
$1,424,090
NA
NA
10.55%
10.81%
NA
NA
8.65%
8.85%
NA
NA
1.95%
1.99%
NA
NA
0.0
0.0
NA

NA
NA
0.2
0.2
NA
NAJ
0
0
NA

NAI
	 : 	 11
NA
174
174
mixed Further Processing (61 verceht in Subcateeory E - 1, 39 percent in Subcategory L)
BAT1
BAT2
BAT3
BAT4

PSES1
PSES2
PSES3
HPSES4
Render
BAT1
BAT2
EBAT3
IBAT^
5




97



NA
NA
$101,224
$215,312
NA
NA
$64,361
$134,011
NA
NA
1 .43%
2.97%
NA
NA
1.18%
2.46%
NA
NA
0.22%
0.46%

NA
. NA
$431,450
$623,290
NA
NA
$287,192
$421,259
NA
NA
6.37%
9.34%
NA
NA
5.27%
7.73%
NA
NA
1.00%
1.47%
NA
NA
0.0
0.0

NA
NA
1.0
1.4
NA
NA
0
0

NA
	 ™I
163
228
ing (Snbc(it*-'">rv J) 	 :, 	 r- 	
21



NA
NA
$188,682
$219,54^
NA
NA
! $119,69S
1. $141,41'
NA
L NA
) 5.709?
J 6.749?
NA
NA
> 4.6595
? 5.499?
NA
NA
> 1.0295
j 1.2195
NA
NA
, 0.2
, 0.:
NA
NA
i 14
\ 14
                   5-50

-------
                                      Table 5-9 (cont.)
                          Economic Closure Impacts: Retrofit Costs
                               Meat Type and Process Classes
                                        Compliance Cost
                                         as a Percentage
                                        of Model Facili
Probability
 Cash Flow
 Less Than
Compliance
    Coste3
        NA
        NA
      1.58%
      1.71%
   Annualized
Compliance Costs
            -1
   .Projected
Facility Impacts4
                              Posttax
                                  NA
                                  NA
                             $184,740
                             $199,761
  Pretax
     NA
     NA
 $285,034
 $305,958
    Excluding 65 Certainty Facilities
                    793.448  $7.748.641
Total Including ^ Certainty Facilities
                     ซ17.309  $2.406,923
                   $6101.244  $3,877,251
                  ^.7^.9231 -$8.368,532
                                        ;          NA       NA         ^^-\       "•-'
                                        A results for meat type and process class, ^charge ^ aud UMxki laulu,
  5UOptionฐBAT 5 is only found in Poultry operations.
                                                 5-51

-------
              Subcategory A through D:
              — red meat first processing, further
                      processing, and rendering
              — red meat first processing and
                      rendering

              Subcategory E through I:
              — red meat further processing
              — mixed further processing

              Subcategory J:
              — rendering

               Subcategory K:
              — poultry first processing
               — poultry first processing,- further
                       processing, and rendering

               Subcategory L:   •
               — mixed further processing
               — poultry further processing
costs / net income:
costs / net income:
costs / net income:
 costs / net income:
 costs / net income:
1.30 percent
0.14 percent

2.24 percent
0.29 percent
0.07 percent
1.43 percent

•0.68 percent
 2.73 percent
 2.28 percent
 5.02 percent
 3.01 percent
 1.43 percent
 3.23 percent
The largest ratio of compliance costs to net income under the proposed option is projected in the poultry

first processing, further processing, and rendering class (5.02 percent — Subcategory K), followed by

poultry first processing and rendering (3.59 percent — Subcategory K), and poultry further processing

(3.25 percent — Subcategory L).
5.3     FACILITY NONCLOSURE IMPACTS


        EPA calculated nonclosure impacts for facilities impacted by the proposed effluent guideline.

These impacts include:4                                           •


        •       ratio of pretax annualized compliance costs to model facility revenues,
        4 As discussed in Chapter 3, nonclosure impacts are estimated assuming that the distribution for each of
 the four income measures is normal. Appendix E presents a sensitivity analysis based on the assumption that
 revenues have a lognormal (i.e., positively skewed) distribution. Also note that in'the above analysis, EPA nets out
 the probability that facilities earn negative baseline income under each of the four income measures.  .
                                                5-52

-------
             ratio of pretax annualized compliance costs to model facility EBIT,

             ratio of posttax annualized compliance costs to model facility net income,

             ratio of posttax annualized compliance costs to model facility cash flow,

             number of facilities expected to incur pretax annualized compliance costs exceeding 1,3,
             5 , and 10 percent of revenues, and
              number of facilities expected to incur posttax annualized compliance costs exceeding 3, 5,
              and 10 percent of cash flow.
Because there are gene* no definitive <*-ซ*>. ซ- -* - ฐ* ซ- ซ"" te™ ™ **
cause a faciUty ,o dose if exceeded (other than if the ratio o, comp,iance costs ซo cash flow exceeds 100
percent), EPA calls these ratio measures "nonclosure impacts."

        As discussed in the dosure analysis, the relative size of impac* is direc,ly reiated to the estimated
 cornice costs per faciuty as a percent of facility income and to number of faciMes in ti,e subcategory
 or meattype and process dass. Hence, in generate .argerte (1) ratio of pretax annuahzed cos* ซo
 revenues or BWI. ซ ซUo of posซax annnaiized cos. ซo net income or cash floซ, and (3, me number o
 faciUnes in ** subcategory, *e gr=attr wiU be the number of faciMes projected ซo incur comphance cosB
 exceeding any given impact threshold (e.g., greater than 3 percent of revenues).

         NoK*a,foranygivenoption,mesizeofsonKrafcsre1a,ivetoeacho,hercanbeunan*iguously

  ranted. Tne ratio of pretax compUance cos* ,o revenues vvi,, always be s-er than theratio of preซ
             costs to EBIT; bod, ratios have the same numerator (pretax compHance costs,, bu, because the
   m
  the resuiting ratio is aiways tager. SinuMy, the ratio of posttax compliance costs to ne, income ซD
  aiways be smaller man the ratio of posttax compnance costs to cash flow; bo* ratios have the same
  Aerator (posttaxcomphancecos^butbeeausemedenon^atorne, income is a,wa,s smaHer than
  denominator cash flow (since cash flow e,ua,s ne, income plus depreciation) a iarger ratio wrU ป*  I.
   genera,, *e cash flow and EBIT ratios cannot be unambiguously ranked. The denominator cash flow
   shouM be smaHer ta, the denominator EBTT. However, the numerator posKax compliance cos. . aiso
   smauer than the numerator pre^x comphance costs, therefore the re,ative size of me two ratios w,U depend
                                                5-53

-------
on taxes and depreciation, which may vary.  For the meat products industry analysis, the cash flow ratio is,
with the exception of some options in the rendering subcategory, larger than the EBIT ratio.


       5.3.1   Nonclosure Impacts by Subcategory
        5.3.1.1 Upper-Bound Cost Nonclosure Impacts


        Table 5-10 presents a summary of impacts by subcategory, discharge type, and technology option
(the ratio of compliance costs to net income may be found on closure impact tables 5-6 through 5-9).
Among the dkect dischargers, the largest impacts are seen under BAT 5 for Subcategory K. Of the 88
facilities in that subcategory, 19 are projected to incur compliance costs greater than 1 percent of revenues
(22 percent of all facilities in Subcategory  K), and 4 will face compliance cost greater than 3 percent of
revenues (5 percent). Twenty-one facilities are projected to incur costs greater than 5 percent of cash flow

(24 percent).


        Results for the proposed direct discharging options, BAT 3 (Subcategories A through D, E through
I, K, and L) and BAT 2 (Subcategory J), are presented below. The ratio of compliance costs to average
facility revenues, and the number of facilities projected to incur compliance costs greater than 1 percent of

revenues or 3 percent of revenues are:
                Subcategory A through D:
                Subcategory E through I:
                Subcategory J:
                 Subcategory K:
costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:
0.12 percent
2.1 facilities
0.6 facilities

0.05 percent
0.2 facilities
0.1 facilities

0.17 percent
0.9 facilities
0.3 facilities

0.43 percent
12.2 facilities
 2.8 facilities
                                                 5-54

-------

w o
V "S
y w
J-y
g V
g 61
!^1 _H
Kfi S3
i 0
g- a.
"u tซ
A V
to v

w &
S^^ fli
o w
11
3 ^J
u a
•S3
•Is h
s .si
(y ^
fa (^



S
O> -.
.M W
o S
il
1*
3

1

ง
^ป
^1




,•••*-ป
1
ฃ
;o

8
jy

|
1
o
*"'
1
ฃ
VI
- -a
i*

f>

.•g
ง
g

CH
i-l
i
.43
ฐ
M
W

1
OJ

55 1

•3
ซ
fa


S
•.s
Q.
0

































Q
&C
g
t-i
^

j^1


3
w
4
cq

5^
O
60
^5
o
•ง
O
o



o
d



o
0
o
d

o
d


o
d




o
0



i
o
o

o
o
o

o
o
o

ON
l-H






i 	 [
<^
P3
O
o



o
d



^H
0
o
d

o
d


o
d




T— 1
O



CN
< — i
O

r-H
O

*
CN
O
O








rN
ง
i — i
O



CN
d



m
0
o
d

i — i
d


i — i
d




CN
0



en
en
O

en
O

in
O
O








en
m
ON
o



oo



CN
en
d.

en
d


in
d




r-
,— ^



S
CN

ON
^

en
en
O








^.
i












































o
en



i — i
vd



en
O
in
d

ON
d


\o
T— t




r— (
in



1
o

i — i
m
O

ON
o
o

en
rN






i — i
fTl
00
oo
r-



CS
vd
en


o
i — i
VO
p
en

vd


(N
r-H




00
o



f-
t—
en

1
en

CN
in
o








CN
OO
OH
VO
r— 4



O
d
en


O
i — i
in
en
CN

ON


in
06




—4
O
en-


1
en

oo
en
CN

T-H
o








en
oo
OH

O
CN



en
i — i
^ซ


en
oo
VO
CN
en

oo
vd


OS
r— t




•3-
en
**


CN
<^-

CN
CN
m

in
in
O








Tf
00
OH





































kปj
.^
^
O

1
O
o
o



o
d



o
o
o
d

o
d


0
d




o
o



8
0

1
o

o
o
o

CN






,_,
M
en
O



in
d



OS
o
o
d

CN
d


en
d




ON
o



VO
m
o

1
o

r-
l-H
O








rN
ffl

en



CN



in
i — i
p
i— i

oo
T— t


en
en




p-
o
'^


in
in
VO

oo
en
VO

in
oo
i — i








m
OH
r-
en



oo
r~



en
CN
l-H
O
1 — 1

1—4
CN


en




TJ-
i__i
*"*


VO
f-

1
VO

CN
O
CN








^t
s<
OH

-------
 Is
o1? a
g ซf
t> S u
  a3
•sfg
C-i ป-5 (E<
 o U

f"*
3 *
c/i O

r-


^


CN
V
t

NO



ON
CO
CO



oo
oo








^
ON


^
'~~l



o\
^

o
oc


1:
oo


fป
o

t

-



"NO"
CO



CN
O



1^:




ON
ON


^
1—1



T
a

ฃ
ON


Nl
HN
OO


fe-
es


t
C/"










































^
•
•3
o



o
o



o
o



^




o
0


C5




O
o

i
o


6-
0


c
o
o
oc
oo
T—
E-

o



^,
— 1



oo
•""



(5




o
ฐ


0
c>




NO
o

ฃ
o


o
!O








ON
CO


p




NO
CT

OC
ซ-,


c
•*


r~-
NO
O

(
CM
ON



NO
ON
CO



CN
NO








*-;
"*


oo




co
r-
co

0
10


o
"*


1
O

c/
p-
p.










































-J


o



o
o



o
0








c
^


o
C5




o
o

0
o


0
o


c:
o

PC

o



CN
o



CO
o



o




p
'-'


o
o




>— 1
o

e>
CO
o


f
O


ฃ
C
O

g
PC





NO
CN



NO
"*








CN
—


C)




f>
CM

S
CO


V
CO


fe-
oo
o

<
PC





ON
CO



co
NO


CN





"^
—


OO
o




o


c
<0


0
**


c
NO
o

?
ec





oo
CO



CM
VO


CN





•*
—


OO
o




o
^

Nฃ
>0


0
"*


ฃ>
r-
o
CO
PC













































IO



ON
^



NO
CN


NO





"*



H




OO
OO

NO
CN
1 '


S
r— *


fe-
oo
o
oo
V
i
c

*



CO
CO
00



CM
s



>0




r*-
r-<


CM
r-




<— '
O

Nl
OO


C
c~"


Vj


CN
C/
V.
i

ON



v~t
NO



^
oo
ON


-n-





1 ;



5




ON
r-

r--
oo
10


o
CO
>o


c-
oo
o

CO
V
p.
c.

co



O
oo



"~J
NO



^*




"^
^


CO
o




"*
r-
ON

V
r-


o
NO


C


o
p.
c

-------
  u
  T3

   ง J
  .O b
oซ a
•s  sir
O  fti S
u  S u
^•S-ja
o i-1 a
ปT  5S ซ>
m * JN

•S  S.9
   = r^
H ซ fe
   g P
   s S
   s ^

  1
   o

S. &
pliance Cos
of Cash Fk
E ซ•>
curringCo
liPercehtag
>& !8
Facilities
Greater Tli

ll
;a>\S
Compliant
ntagebf Re
a -•ซ
.52
3 :C

l;|
Percent
• to •••!S
05 t
SI
'S3 S
1*
1
O;,
Number
\...-.

••s
••ฃ
a>
-. PH
.. .O
.-l-t

/I
V)

3Percei

g
V
•'-• ".(1.
- o.

••:, "I
•w
i
- .-**
;;l
••..-.3
.;*
: " S
Revenues
of
Facilities
'-•'" • j
















Co
o
S
1
VO
cx:
^
O
O

o
o
0
o
o
o
o
o
o
o
o
0
2
2
2
g
CN
<
pa
Os
0

in
CN
CN

0
o
CN
o
en
O
OS

2
2
2

P
pa
Os
en

t--
Os
CN
oo
OS
00
i— i
oo
en
CN
r-
r-
r-
CN
*
*
*

en
pa
VO
o

t^
en
en
r-
t-.
CN
en

Os
OS
00
00
en
*
*
*

ง
oo
1 — 1

r-

oo
Os
en
CN

VO
CN
O

CN
"
*
'*
*
c
h
pa





















i — i

0
en
CN
00
en
in

00

o

in
r-
*
*
*
ฃ
PL
V
cu
CN
OS

0
1 — 1
CN
^H
en
OS
*— i
in
en
en
vo
CN
c^
0
*
*
*

CN
00
w
w
cu

VO

OS
OS
ri-
CN
OO
CN
in
CN
*— H
r-
CN
VO
oo
•<*.
\D
OO
OO
1 — 1
2
*
*

e<-
ff.
0-c
O
Tj-
f— H
1 — 1
cn
i — i
CN
CN
en
O
i — i
i — i
CN
en
r— <
OS
00
CN
en
CN
25
^
55

Cu
















3
I
s
1
vo
.s
o
1
tS
o
o.

o
0
o
0
0
o
0
o
o
o
o
o
z
*
*
VO
CN
CN
<
O
T 	 1

r^
CN
in

o
o
CN
O
en
0
_ <
CN
^
*
2

CN
pa
o
>n

^
CN
en
oo
en
in
OS

i — i

oo
r-
Os
Os
CN
*
^
*

I
CN
CN

CN

O

Os
CN
r-

r-
o
i — i
OS
1 — 1
z
2
2

pa
r-
CN

r~

0

en
i — i
00
CN
Tf

,— i
in
CN
^
*
*

E-
pa





















CN-

OO

o
en
vo
i — i
p
en
Tj-

Os
oo
2
z
2
f-
V.
Oi
Os
t-~
i — i
in
vo
O
o
in
cn
T 	 1
VO
CN
VO
cn
VO

00
o
25
2
2

CN
VO
en

en
S
T— H
o
in
O
en
in
en
oo
00
CN
in
CN
m
VO
en
2
2
2

en
OH
en
CN

>— '
O
en
CN
i — i
0
in
en
CN
vo
i — i
r-7
en
oo
en
vo
•^1-
m
CN
2
2
2

oo
o:
DH


| J
B J
bo s
> -g
ombination, i
ge type.
: number of f
S a ฃ
.a-g >-'
w ซ 4D
&^ T3
S "ฐ ป'
rt &1 "3
ID O ' C
•s SP r
e type and m
zes by subcat
cash flow.
equal to zero
%•% P .83
tegory, dischs
er all facility
)liance costs
ame measure
s S s y
for each sub
ts is the sum
jnnualized co
probability ii
bO C3 <> en
d as the avera
ing those imp
atio of postta,
e measure (le
ts are presente
'acilities incurr
es and EBIT; i
ntageofincorn

>ercent of facili
,ach combinatic
zed complianc
e costs exceed
a .s ง i
ฃy CO C ^H
fill
11 fl
0 ง-ฃ*
-------
      Subcategory L:
                                     costs / revenues:
                                     exceeding 1 percent:
                                     exceeding 3 percent:
0.48 percent
2.5 facilities
0.4 facilities
             „      psES 2 for Suta,egc,ry L has the largest nonclosure impacts. There are
For indirect dischargers, PSES 2 for Sub   g ry                  ^ ^.^ ^ .^

 ป—  -s,Ln^-^-ซ---
 5.3.1.2 Upgrade Cost Nonclosure Impacts
Therauo
          Subcategory A through D:
          Subcategory E through I:
    .      Subcategory J:
    .      Subcategory K:
            Subcategory L:
         oofco^ceco.to averaged
                                        costs / revenues:
                                        exceeding 1 percent:
                                        exceeding 3 percent:

                                         costs / revenues:
                                         exceeding 1 percent:
                                         exceeding 3 percent:

                                         costs / revenues:
                                         exceeding 1 percent:
                                         exceeding 3 percent:

                                          costs / revenues:
                                          exceeding 1 percent:
                                          exceeding 3 percent:

                                          costs / revenues:
                                          exceeding 1 percent:
                                          exceeding 3 percent:

   0.09 percent
    1.4 facilities
    0.3 facilities

    0.04 percent
    0.2 facilities
    0.1 facilities

    0.17 percent
    0.9 facilities
     0.3 facilities

     0.30 percent
     7.6 facilities
      1.7 facilities

      0.36 percent
      1.5 facilities
      0.3 facilities
                                         5-58

-------
 Results for all options and discharge types at the subcategory level are presented for upgrade costs in Table
 5-11.
         5.3.2   Nonclosure Impacts by Meat Type and Process Class


         5.3.2.1 Upper-Bound Cost Nonclosure Impacts


         Table 5-12 shows nonclosiire impacts by meat type and process class, discharge type, and
 technology option. From this table, EPA presents the upper and lower nonclosure impacts by class within
 each overall subcategory average for the proposed direct discharging options (BAT 3: Subcategories A
 through D, E through I, K, and L, and BAT 2: Subcategory J) below. The range for the ratio of estimated
 compliance costs to average facility revenues in each subcategory is:
                Subcategory A through D:
                — red meat first processing, further
                        processing, and rendering
                — red meat first processing and rendering

                Subcategory E through I:
                — red meat further processing
                — mixed further processing5

                Subcategory J
                — rendering

                Subcategory K
                — poultry first processing
                — poultry first processing, further
                        processu^and rendering
costs / revenues:
costs / revenues:
costs / revenues:
costs / revenues:
0.12 percent
0.01 percent

0.22 percent

0.05 percent
0.01 percent
0.27 percent

0.17 percent
0.43 percent
0.32 percent
0.84 percent
         The number of mixed further processing facilities for which compliance costs are greater than any given
income threshold is allocated to Subcategory E through I and Subcategory L in the following way: 0.61 percent of
them are placed in Subcategory E through I and 0.39 percent are placed in Subcategory L. For example, the
number of facilities with costs greater than 1 percent of revenues in the mixed further processing class is 0.4.  This
number is scaled by 0.61 to estimate the number of impacted mixed meat facilities in Subcategory E through I, and
by 0.39 to estimate those impacted facilities in Subcategory L. This results in 0.2 impacted facilities (rounding to
the nearest tenth of a facility) allocated to each subcategory (see Section 2.2.2.1 for more detail).

                                                5-59

-------

-------
   42
    '
a  2  |
 a  iu  -g
 8*  2
   r

2 |
_Q r^
^> •*
^ es
ฃ3 CM
a. o
= 
•a fc
M
ซ ฃ


1
Si
tn 0
0 fa
11
a *
0
U
-w
1
I*
pi
S


*•ป
fl
' 
-------

-------
3
   VJ
 11
 M
 ?"
 S 2
<2 a S

3^1
o> ••


ip
 S ft
 "t^
 a H
 = r,
 S^


 I

-2 ^
% 5
o f*<
^
^5


k.
j
5
TS.
ft*
3
o







o
o


o
o





^
o


0
o

o
o

0
0

C
o


C
o




^
C
o



ฃ3



pc

o







o
o


o
o





o
o


o
o

o
0

o
o

1
o


0
o




tp
C
o




!

H
pa

o







o
o


ป—H
o





0
o


o
o

o
o

o
o

s
o


V
0




t^
C
o






H
03

o







oo
O


>n
— '





,-H
ฐ


o

o

SO
o

en
oo
•-1


C
r— 1




b&
o






.ฃL
pc



















































_;







en


SO
SO





en
o


C5

o

oo
CM

0
o


C
o




t^
r~
o



oo
SO


n'
t/a
PH

oo







00
SO


OS
oo





! 	 1
"'


n
•^j-




J.Q
Os
O






OO
ffl
PH





























__^
4
o
5i
u
c<

*l^

^
K

S
s

C
^
|
1
<

^

o







o
o


0
o





o
o


p

o
o

o
o

0
C
o


C
o




fป
o
o



SO
en


H
m

O







o


SO
o





o
o


o
o

o
p

>— 1
o

s
o


I/I
CM
o




tA
CM
O
o






H
pa

CM







oo


-— i
oo





t— I
o


en
O

SO
0

<— <
CM

OS
CM


en
.so
CM




v.33
CM
O
O






H
m

CM







3


CM
OS





—*
O


en
O

r-
d

>o
CM

en
en
en


ง
en




ซJ3
o






&
P3
                                                    en
                                                    so

                                                    in

-------

1
11



Wo,
Wo —
13 ,ง0
g™ 8
*ฃ! CJ O
l&i
ฃ-ฃ
-21
•A 8,2
Ji E &
glfr
** a "ra
ง>ซ
•3*
e
S

1





iliance Costs J
jf Cash Flow 1
P 
Pซ jS
1!
6-
P
.2 Is
r
Number


10 Percent
0.0
•W T-
1 =
ฃ
VJ
3 Percent
m
10 Percent
nn
5 Percent
A A
*j e
1 =
ฃ
to
*> c
k! <
1
*•<
i!
rv*.
•f
3
i
n
m
1
Revenues
ot
Facilities
c
1

cs oo

>n
ป— (
<-< CM
S 8

I-;
ป— t
VO
CO
i n
•\ i —
5 C5
T CO
3 0
T m
3 0
SN o\
-H 1-H
S ?ฃ
0 0\
N T-I
n in
* *
to in
p- vo
•* •*
* *
o\ o>
CO IT
0 C

CO • •
0
o
o
d
o
o
o
o
0
o
o
o
0
0
*
o
0
o
tf
s
o
•^t
.
.
.
4
I
\ r-
is

o
o
o
o
o
o
0
o
o
c>
o
o
o
o
*
oo
o
*
OO
T-H
0
^
S
c

IS

o
d
o
o
o
o
o
0
o
<5
'
o
o
o
o
^
o
o
*
o
i-H
o
*?
S
0

15

d
T^-
o
r-
C5
o
d
*-H
d
T— (
0
^J-
o
^
00
m
0)
i
CO
CN
*
00
CO
O

iS

















^
*

O
1 — i
O
CN
d
o
d
o
d
o
o
i — i
o
tฃ
•*
TT
o
2
^
o
*
S
c
t-
r~
z
%

o
in
T— i
in
oi
1—4
d
CO
d
ซn
o
o\

*
55
m
ง
S
•*
^
;O
00
O

< 0
i g
J Z

o
OS
o
V3
T— 1
1— 1
d
cs
d
CO
o
T— I
V— 1
^
•*
CO
co
*
S
co
#
o>
"*
c

1 f
II

o
o

oo
i — i
i—i
d
 -n
i ty
1 a
LS








cj
1
^c
&•
So
3
1
52.
g
•8
ง
ฃ
1
g
1
u
1
X.
ซ)
I
&
E
1
n!
!
ฃ
;l
1 -5
J <2

O
o
o
o
d
o
d
c
o
o
o
o
o
^
o
0
o
^
o
p
d
ER
o
o
o
S
k
>
1 —
is

o
co
o
>n
d
o
d
u
o
o
o
1-H
o
*
1— (
CO
o
*R
r— (
cn
>_J
^
ฃ
c

iE

O IN
CN •*
O •*
co co
d t>
O CN
d d
0 0
o in
o o
O OS

^ ^
m CM
i— ( OS
O CO
* #
S 5?

^ ^
— 1 OS
o cs
d d

' P •?*•
< H H
3 < <
3 m ซ

d oi *-n
CN ^t; \o
d "* 
-------
in
VD

-------

)liance Costs
of Cash Flow2
B 
3 rt
ฃ 1
Facilities]
Greater Th

r*
&'%
- i
u e
Compliance
ntageofReve
60 0)
.a K
H
^j
s^
ซ t-
U ^
^1

c
U
Q> "^
eu !>•
ince Cost as !
dtodelFacilit
a u
g,e
a
S3'
A
&


•*•!
a
ง
ฃ
o
TH
"*•*
5 Percen
•W
CJ
fci
ฃ
to
-*j
ง
ฃ
c
1
If)
3 Percent
1 Percent
<
(s<
1
ฃ
ฃ|
(S
en
i
J
M V
O j
fr
CZ

o
o
o
c>


p

o
o
o
c>
o
o
SS
in
o
o
fi
c
o
&.
c
o
o>
cs
I

in
CO
CO

ซrf

fT

00
T
O
r-
^
CO
l-ซ
•E
• o
ON
f
5
O

CN
V
f

co
"r^
VO
CO

fr

f-;
o
VO
m
S

&~
ON
C*1
CO
&~
ฃ
C

c*"
o*
I

CN
CO
(— v

OO

fT

OO
c*
ON
V
fe-
es

ฃ>
CO
If
CO
&~
ON
V
o

c/
U.
i












^
o
B<
ซ:
a
<:
^
cu
.=
•?
c:

•*
T— i
VO
vd
*
CN
VC3
*
VO
in
in
^
o\
oc
c>

II"
E-
<
(T


















O
O
O
o

H

o
o
o
o
o
^
o
&~
in
o
&~
cs
c
o
if
^
p.
CL

CM
O
CO
VO
CO


oo
vq
CO
in
V
&-
w
^
VO
c^
c-

CN
%•


T— 1
0)
oo
cs
^

fT
VO
o
oo
c-
^
oo
oo
&-
r~-
r-
i
o

c<~

O\
C-
^
?
oo
fe-
vo
ON
t~-
E
C

o
I
c











^
L
Q
s<
i
i
i
c/
^
Q^
I
k.
ft.

o
cs
o
co
0



o
o
o
^
oo
O
&-
c-
C"
o
&>
c
o
in
i—
y
c
O-

i-4
CO
co
[•-
in


fr
VO
o
o
CO
^
f-

&-
CO
C<"
6-
vo
Vl
o

r*
I
a


•*
CS
CN
•*
^

o
C)
"*
o
t—
E^
CO
C*"
6-
ON
CO
CN
&-
O

f
w
C
C


VO
0)
VO
-*


c<
**
o
oc
E^
CO
CO
&-
CO
V
r*
&-
r<
O

o
P-
C









ฃ?
? (Subcateeon
J
^
c
•^
c
fa.
c
it
a.

o
o
o
o
o



o
o
o
o
^
o
fc-
c
o
&-
o
o
o
V
ฃ
ff

o
N
o
CO
o
Q


o
o
o
^
VO
o
&-
co
VC
o
fe>
o

o
E-
ff

o
o
CN|
"— '
CO
^


^
cs
o
^
VO
ON
m
&-
o
ฃR
D
O

f
cc

o
o
cs
CO
co
—;


•ซ*•
O
o
^
VO
6-
c
V
&<
c
C

?
<
PC

S
— 1
"S
in
CO
(_;

O)
m
C3
in
o
^
oo
VO
VO
ฃฃ
CO
in
^
CO
ON
O

• in
i ^
i oa
VO

^
in

-------
   3

   U  ฃ
   '"e  $
^- ง  —

 I?  %
.5 oj  o>
      •s
   1
                                                                                                                             IO

-------

-------
             Subcategory L:
             	mixed further processing
             — poultry further processing
                                                          costs / revenues:
                        0.48 percent
                        0.27 percent
                        0.52 percent
       5.3.2.2 Upgrade Cost Nondosure Impacts

       TaHe 543 contains the results of the nonclosure hnpac, analysis by meat We and process class,
             , and techno.ogy option for retrofit or upgrade costs, From this table, EPA presents the
direct discharging options (BAT 3: Subcategories A through D, E tough I, K, and L, and BAT 2.
slaซegOTy ป L.. Using -upgrade costs instead of new equipment costs in ซhe analvs,s, ซ. range for

the ratio of estimated compliance cos* to-average facfflty revenues in each snbcategory ,s:
                Subcategory A through D:
                — red meat first processing, further
                       processing, and rendering
                — red meat first processing and rendering

                Subcategory E through I:
                — red meat further processing
                	mixed further processing

                 Subcategory J
                 — rendering

                 Subcategory K
                 	poultry first processing
                 — poultry first processing, further
                         processing and rendering

                  Subcategory L:
                  	mixed further processing
                  	poultry further processing
                                                             costs / revenues:
costs / revenues:
costs / revenues:
 costs / revenues:
 costs / revenues:
0.09 percent
0.01 percent

0.15 percent

0.04 percent
0.01 percent
0.19 percent

 0.17 percent
 0.30 percent
 0.23 percent
 0.60 percent
  0.36 percent
  0.19 percent
  0.38 percent
                                                  5-69

-------
o
f-

-------

-------
   i
       2
fO
<:~J
                                                                                                                                     VO
                                                                                                             8ซ
                                                                                                                                     ^
                   •2    -a
                                                         ^
                                                         oo
                                                         nj

-------
^1
.5 tS

J,|(


sl'
ซ M
is
  s,
•ri


.gi pfi
•g.a
•JS " ฃj
wig

jjฃ
>s|
•Is ซ-

u c;
fao

~
N
CA CL
O 5

"H, ฐ
• C fl
o e
.s a
'E ^
3 c
5,?

•3 ฃ
1 ซ
ฃ&< r*"1

•ซ•ป
ง
w
ฃ ~
as •*J
ซ 1
(ง'-
ซ•=
e-ซ-
i-
ft
g
1

1
. s_
u
: PH
O
i-l
1

PH
UJ
1
g
tฃ
CO

1
Id
P*
O
i— (
1
ฃ

V)
1
a>
rn

s
t,
&
M


^*
	 r-
M
W


I
^
a

	 u ซM — 5
^j
'!
ft;
•^




n^

•
fcc.




Q.
.
I
Z




g



g




g



z



z



g


z



<ฃ
^



"g


VC




?
<
p:
Z




g



g




Z



Z



z



z


g




-------

-------
I-;-.
ซo

-------
                                                                                           as
5.4    FINANCIAL RATIO ANALYSIS

       EPA also examined the impact of the proposed ELG on the model establishment's balance sheet
well as its income statement, using the methodology outlined in Section 3.1.3.  As explained in that section,
return on assets (ROA) was used as the financial ratio to indicate firm profitability. ROA provides a
reflection of the opportunity cost of investing in the meat product industry. Investors look for their best
opportunity to receive a high rate of return on their capital.  If the proposed ELG significantly lowers the
rate of return earned in the meat products industry, investors may exit that market in search of better
 opportunities; the meat products industry would therefore tend to contract.
       5.4.1   Financial Ratio Analysis by Subcategory

       5.4. J 1 Upper-Bound Cost Financial Ratio Analysis

       Table 5-14 displays median.ROA, model facility net income, estimated model facility total assets,
the post-compliance ROA, and the percent change in ROA as an impact of the proposed rule by
subcategory and technology option. EPA presents impacts in terms of the percent change from baseline
ROA to post-compliance ROA. The greatest change in ROA is witnessed under BAT 4 in Subcategory J:
the baseline ROA is 2 percent and the post-compliance ROA is 1.8 percent, resulting in a 10 percent drop
in ROA due to compliance costs. For the proposed options (BAT 2 for Subcategory J and BAT 3 for all
others), the subcategories have the following percentage change in ROA:
               Subcategory A through D:
               Subcategory E through I:
               Subcategory J:
               Subcategory K:
               Subcategory L:
                                                                                      -2.6 percent
                                                                                      -0.5 percent
                                                                                      -0,7 percent
                                                                                      -4.5 percent
                                                                                       -4:8 percent
                                               5-76

-------
                  Table 5-14
Unpacts to Return on Assets Ratio: Upper-Bbund Costs
            40CFR432Subcategones
                            Return on Assets
                                                                                  Percent
                                                                                  Change
                                                                                    ROA<
                                                    Post-
                                               Compliance
                                                   3ROA
                   Model Facffi""1
      NUmbSf~Netlncome| 'total:As^
                                                                                  0.00%
                                                                                 -0.31%
                                                                                 -2160%
                                                                                 -5.68%
                                                    5.30%
                                                    5.28%
                                                    516%
                                                    5.00
SubcategoryAjhrou
                                                     5.26%
                                                     4.60%
                                                     	
                                                     4.78%
                                                     —   •—
                                                     4.71%
  'SES1
  SES2
  'SES3
  'SES4
                                                      5.50%
                                                      5.49%
                                                      5.44%
                                                           ~
                                                      5.17%
                                                      ..   ——
                                                      5.22%
                                                      .
                                                      5.11%
                                                                   O.OQ%
                                                                  -0.68%
                                                                  -9.03%
                                                                  -9.90%
                                                                           2.00%
                                                                           1.99%
                                                                           1.82%
                                                                            1.80%
iSubcategory J
                                                                   -0.54%
                                                                        —
                                                                   -9.70%
                                                                  -12.12%
                                                                  -.
                                                                  -12.79%
                                                                            1.99%
                                                                            1.81%
                                                                            1.76%
                                                                            1.74%
  SES1
  SES2
  'SES3
  'SES4
                           5-77

-------
                                          Table 5-14 (cont.)
                        Impacts to Return on Assets Ratio: Upper-Bound Costs
                                      40 CFR 432 Subcategories
          Number
                of
          Facilities
                         Model Facility
     Net Income
       Total Assets
          (x $1,000)
                               Baseline Return on Assets
                                    . Median
•  Lower
Quartile
      Post-
Compliance
     ROA3
BAT1
BAT2
BAT3
BAT4
BAT5
15



13s
$4,655



$4,676
$21



$23
                                                         2.5%
                                                         2.0%
                                                      -0.3%
                                                      -0.5%
                                                        2.46%
                                                        2.45%
                                                        2.35%
                                                                                    2.28%
                                                                                    1.85%
 PSES1
208
$4,493
                                       $198,535
                                         2.6%
                                                                      -0.2%
                                                                    2.59%
                                                                                    2.34%
                                                                                    2.42%
                                                                                    2.34%
Percent
Change
 ROA4
tlSIMUUl
ISubccttsso
IBATI
IBAT2
(BATS
BBAT4
llRAT1!
1
	
HPSESI
jpcpco
HPSES3
HDOCCM
ryK
88





138



-
$12,016





$12,305




$600,816




2.0%




-0.5%





$615,266



2.0%



-0.5%




2.00%
1.99%
1.91%
1.88%
1.87%

1.99%
1.81%
1.85%
1.84%

0.00%
-0.34%
-4.54%
-5.88%
-6.43%
II
-0.62%
-9.72%
-7.43%
-7.77%
                                                                                                 0.00%
                                                                                                 -0.39%
                                                                                                 -4.84%|
                                                                                 -7.02%
                                                                                 -7.63%
                                                                                                 -1.71%
                                                                                -10.93%
                                                                                 -8.16%
                                                                                -10.58%
Aggregating impacts to account for the 65 certainty facilities is not applicable for these impacts
                  in this table are the average of results for each subcategory, discharge type and model facility size
                      he number of facilities in each combination.    ,                                  •nr\*\
                                             a; model facUity total assets calculated as 
-------
I                ; psES2inSubcategoryAthroughD,thepercentagedropinROAisl3
  ^^"zz*-**"-™™'-'''"**
   compliance ROA is 4.6 percent.
        .41.2 Upgrade Cost Financial Ratio Analysis
The p^e change in ROA for the
others) are as follows:
          Subcategory A through D:
          Subcategory E through I:
          Subcategory J:
          Subcategory K:
          Subcategory L:
                               options (BAT 2 for
            E trough I ,o 50 pe.en. sn-aHer in
                                        K.
                                                         -1.6 percent
                                                         -0.4 percent
                                                         -0.7 percent
                                                         -3.0 percent
                                                          -3.3 percent
           .*W Vpper-Bound Cost Financ
           by
                   das,
                                    5-79

-------
                 Table 5-15
Impacts to Return on Assets Ratio: Retrofit Costs
          40 CFR 432 Subcategories
.
Number
of
Facilities <
r '
\SubcQtegot
I 66
I
1 	
r — f

r 60

—
—
i — '
flSlibcCltBBO
} 19

	
	
	
234

	
	
\\Subcatest
HV-
r~~
—
0

Model Facility '
Option
Net Income
(x $1,000)
Total Assets
(x $1,000)
Baseline Return on Assets 2
Median
Lower
Quarttie
,- • . • '
Post-
Compliance
ROA3
11
Percent
Change
ROA4
y A through D . 	 1 	 	 	 1 — 	 	
3AT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
$26,901



$507,564



5.3%



2.2%




$17,963



$338,932



5.3%



2.2%



• NA
NA
5.21%
5.15%
NA
NA
-1.62%
. -2.84%

NA
NA
4.84%
4.77%
ry E through 1' ' 	 > 	 ; 	 ^
BAT1
BAT2
BATS
BAT4

PSES1
PSES2
PSES3
PSES4
jjyj
BAT1
BAT2
BATS
BAT4
$8,558



$155.592



5.5%



1.3%




$6,370




$2,080



$115,819



5.5%



1.3%




$104,00;



2.0%



-0.5%



NA
NA
5.48%
5.42%

NA
NA
5.22%
5.11%

NA
N.A
1.88ซ
1.859!
NA
NA
-8.73%
-9.92%

NA
NA
-0.36%
	 	 H
-1.45%
II
NA
NA
-5.06%
-7.04%

NA
NA
, -6.16%
? -7:40%|
75

	
bm
PSESl
PSES2
PSES3
PSES4
$2,076



$103,801



2.0%



-0.5%



NA
NA
1.81%
1.79%
NA|
NA!
	
-9.63%
-10.50%|
                      5-80

-------
                                             Table 5-15 (cont.)
                             Impacts to Return on Assets Ratio: Retrofit Costs
                                         40 CFR 432 Subcategories
Number
of
Facilities

88





138



•. . •" -'•• •>•'-.' ~:-'.. ' . ' "-• . .- ;•., - .. '
Option
ryK •
BAT1
BAT2
BAT3
BAT4
BATS

PSES1 .
PSES2
PSES3
PSES4
Netincome
(i $1,000)
TotalrAssets
' (x $1^000)
Baseline Return bn;Assets 2
Median
. • ;••:•-. Lower
Quartile
Compliance
-•-'.^•'-"•'•'XOA1*

$12,016




$600,816




2.0%




-0.5%




NA
NA
1.94%
1.92%
NA

$12,305



$615,266



2.0%



-0.5%



NA
NA
1.86%
1.85%
Percent
Change
ROA4

NA
NA
-2.98%
-3.93%
NA

NA
NA
-6.99%
-7.42%
Subcategory L 	 	 	
15

1

13s

208
1"


BAT1
BAT2
BAT3
BAT4
BATS
$4,655



$4,676
$214,016



$233,818
2.5%



2.0%
-0.3%



-0.5%

PSES1
PSES2
PSES3
PSES4
$4,493



$198,535



2.6%



-0.2%



' NA
NA
2.38%
2.35%
NA

NA
NA
2.42%
2.34%
.NA
NA
-3.29%
-4.51%
NA

NA
NA
-8.14%
-10.57%
Aggregating impacts to account lor me oj ueii
-------
                   Table 5-16
Impacts to Return on Assets Ratio: Upper-Bound Costs
          •Meat Type and Process Classes
33=3==
Ootion
	 —-———ป —=ซ====
•Nfi.mhpr Model Facility x
INUlHDer. 	 : 	
of Net Income
l?ar>ilit;<ป: fx $1.000)
Total Assets
fx $1,000)
— i i J-—g— !B--^!^^^==SgJ— .
Baseline Return on Assets 2
Median
Lower
Quartile
Post-
Compliance
ROA3
Percent!
Change
ROA4
\ftpfl Meat Firrt Prnffviinf (Sithrfttepnrv A - D) 	 , 	 _ 	 1 	
IBATI
!fiAT2
IBATS
IBAT4
\R0d Mea
IBATI
flnATO
BEATS
BAT4
6



t Further P
12



$2,696.1



$50,870.6



5.3%



2.2%



5.30%
5.30%
5.29%
4.98%
0.00%
0.00%
-0.25%
-5.98%
recessing (Subcatesorv E - /) 	 	 	 — . 	
$7,650.9



$139,107.7



5.5%



1.3%



5.50%
5.50%
5.49%
5,33%
0.00%
-0.08%
-0.10%
-3.02%
	 r
IPSESl
Ipcpco
Ipopca

IP a/1 Mpfll

HP
-------
                Table 5-16 (cont.)
Impacts to Return on Assets Ratio: Upper-Bound Costs
          Meat Type and Process Classes
1
Option
PSES1
PSES2
PSES3
PSES4
Number • Model Facility *
of
Facilities
7



Net Income
(x $1,000)
$14,363.6



Total Assets
(x $1,000)
$261,155.9



Baseline Return on Assets 2
Median
5.5%



Lower
Quartile
1.3%



Post-
Compliance
ROA 3
5.46%
5.06%
5.20%
5.16%
Percent
Change
ROA4
-0.71%
-8.04%
-5.47%
-6.12%
Red-Meot fir.it Processing, Further Processing, and Rendering (Subcategory A-D)
BAT1
BAT2
BATS
BAT4
24



$29,321 'A



$553,233.8



5.3%



2.2%



5.30%
5.28%
5.29%
4.97%

PSES1
PSES2
PSES3
PSES4
17



$29,321.4



$553,233.8



5.3%



2.2%



5.27%
4.86%
5.02%
4.79%
0.00%
-0.36%
-0.21%
-6.23%

-0.51%
-8.35%
-5.19%
-9.54%
Poultry First Processing (Subcatesory K)
BAT1
BAT2
BAT3
BAT4
BATS
49




$12,333.9




$616,696.9




2.0%




-0.5%




2.00%
2.00%
1.92%
1.90%
1.89%

PSES1
PSES2
PSES3
PSES4
92



$12,321.9



$616,094.5



2.0%



-0.5%



1.98%
1.83%
1.86%
1.85%
0.00%
-0.24%
-3.80%
-4.94%
-5.41%

-0.78%
-8.68%
-6.96%
-7.33%
Poultry Further Processins (Subcategory L)
BAT1
BAT2
BATS
BAT4
BAT5

PSES1
PSES2
PSES3
PSES4
13




$4,676.4




$233,817.9




2.0%




-0.5%




2.00%
1.99%
1.90%
1.86%
1.85%

155



$4,062.7



$203,135.5



2.0%



-0.5%



1.96%
1.78%
1.83%
1.79%
0.00%
-0.41%
-5.14%
-6.91%
-7.63%

-1.90%
-11.25%
-8.38%
-10.54%
                        5-83

-------
                                        Table 5-16 (cont.)
                       Impacts to Return on Assets Ratio: Upper-Bound Costs
                                  Meat Type and Process Classes
                                                                                          Percent!
                                                                                          Change
                                                                                           ROA
                      Baseline Return on Assets 2
      Post-
Compliance
     ROA3
Model Facility'
         Total Assets
            (x $1,000)
Net Income
  x $1,000)
Poultry First and Further Processin
        First Processing and Renderin
         Further Processing and Rendering (Subcatego
  Poultry First Processing, Further Pocessing, and Rendering (Subcate?o
                                                 5-84

-------
                   Table 5-16 (cont.)
                        Baselin^RfiturnonAssets
                                    '
Percent
Change
 ROAJ
 -0.93%
-21.13%
-12.48%
-12.53%
               Model Facili
Netln^eTTotalAssete
                             Mejiian
                                2.0%
                                                          5.50%
                                                          5.48%
                                                          5.35%
                                                          —    -
                                                          5.07%
v,,rther Processing^
                                                           5.42%
                                                           _   —
                                                           4.86%
                                                                ••~
                                                           5.02%
                                                           4.78%
                                        Capital Costs)
                                5-85

-------
              Subcategory A through D:
              	     red meat first processing, further processing, and rendering
              —     red meat first processing and rendering

              Subcategory E through I:
              —     red meat further processing
              —     mixed further processing

              Subcategory J:
              —     rendering

              Subcategory K:
              —     poultry first processing
              	     poultry first processing, further processing, and rendering

              Subcategory L:
              —     mixed further processing
              —     poultry further processing
-2.6 percent
-0.2 percent
-4.6 percent

-0.5 percent
-0.1 percent
-2.8 percent

-0.7 percent
 -4.5 percent
 -3.8 percent
 -8.4 percent

 -4.8 percent
 -2.8 percent
 -5.1 percent
       Foi Indirect dischargers, the largest decrease in ROA takes place under PSES 2 in the poultry i^

processing, further processing, and rendering class.  The percentage change in ROA for this class is
negative 21 percent, followed closely by PSES 2 in the poultry first processing and rendering class with a

20 percent drop in the ROA.
        5.4.2.2  Upgrade Cost Financial Ratio Analysis


        Table 5-17 presents ROA impacts by meat type and process class using retrofit costs in place of

 new-equipment costs. The percentage change in ROA by class within each Subcategory are:
                Subcategory A through D:
                —     red meat first processing, further processing, and rendering
                	     red meat first processing and rendering

                Subcategory E through I:
                —     red meat further processing
                —     mixed further processing
  -1.6 percent
  -0.2 percent
  -2.8 percent

  -0.4 percent
  -0.1 percent
  -1.8 percent
                                                5-86

-------
                                         Table 5-17
                        Impacts to Return on Assets Ratio: Retrofit Costs
                                Meat Type and Process Classes
                                           Baseline Return on Assets
      Post-
Compliance
     ROA3
        Number,	ModelFac^
                               Total Assets
              of    Net Income
K,.d Meat First Processm
BAT1
 AT2
 AT3
BAT4
Red Meat n^her Processin

  \Recl Meat First ProcessinKandRenderinsJS^^

                                                 5-87

-------
                                     Table 5-17 (cont.)
                       Impacts to Return on Assets Ratio: Retrofit Costs
                               Meat Type and Process Classes

PSES1
Ipopco
Ipqpoa

.
Number Model Facility '
of
Facilities
7



Net Income
(x $1,000)
$14,363.6



Total Assets
(x $1,000)
$261,155.9

~ 	 — — 	
\RedMeat First Processing. Further Processing, anc
JBAT1
IRATJ
IJO ATTl
ID ATM
It

Upcpco
Llpopc-i

llP/iw/frv J
II R ATI
HBAT2
IRATS
I1RAT4
HBAT5
24



17



&ir$t Proce
49




$29,321.4



$29,321.4



	 3>33J,ZJJ.


Baseline Return on Assets 2
Median
5.5%


— .^ 	 	 	
I Rendering (Su
.3%
••
__

Lower
Ouartile
1.3%

	 ^__ —

^category A-L
9 9%



Post-
Compliance
ROA3
NA
NA
5.23%
5.19%
)
NA
NA
5.29%
5.20%

$553 233 8



5.3%



2.2%



NA
NA
5.12%
4.95%
Percent
Change
ROA"
NA
NA
-5.00%
-5.72%
	 	 ^1
NA
NA
-0.16%
-1.81%

NA
NA
-3.37%
-6.62%
j 	 ซ
ss/n# (Subcategorv K) 	 	 — , 	 	 	 1 	 • 	 1
$12,333.9




$616,696.9




2.0%




-0.5%
i-




NA
NA
1.95%
1.93*
N^
NA
NA
-2.49%
, -3.27%
, ' NA
 PSES3
HPSES4
               92
$12,321.9



$616,



                                                  2.0%
                                                -0.5%
 \Poultry Further Processing (Subcategory L)
 BAT1
 BAT2
 BAT3
 BAT4
I1BAT5
 PSES1
 PSES3
 IIPSES4
 13
155
$4,676.4




$4,062.7



$233,




$203



                                     2.0%
                                                               -0.5%
-0.5%
                                                                             NA
                                                                             NA
                                                                           1.87%
                                                                           1.86%
NA
NA
1.93%
1.91%
NA
NA
NA
1.83%
1.79%
NA
NA
r3.52%
-4.67%
NA
NA
NA
• -8.38%||
-10.54%|
                                             5-88

-------
              Table 5-17 (cont.)
Impacts to Return on Assets Ratio: Retrofit Costs
        Meat Type and Process Classes
Option
Number Model Facility J
of
Facilities
Net Income
(x $1,000)
Total Assets
(x $1,000)
Baseline Return on Assets 2
- Median
Lower
Quartile
Post-
Compliance
ROA3
Percent
Change
ROA4
Poultry First and Further Processing (Subcategory K)
BAT1
BAT2
BAT3
BAT4
BAT5

PSES1
PSES2
PSES3
PSES4
16





29



$11,952.9




$597,645.2




2.0%




-0;5%




NA
NA
1.95%
1.93%
NA

$11,894.4



$594,718.8



2.0%



-0.5%



NA
NA
1,90%
1.89%
Poultry First Processing and Rendering (Subcategory K)
BAT1
BAT2
BAT3
BAT4
BAT5,
17




$10,983.2


• .

$549,160.5




2.0%




-0.5%




NA
NA
1.92%
1.90%
NA
NA
NA
-2.55%
-3.61%
NA

NA
1 NA
-5.04%
-5.51%

NA
NA
-3.92%
-5.17%
NA

PSES1
PSES2
PSES3
PSES4
5



$11,156.4



$557,820.1



2.0%



-0.5%



NA
NA
1.77%
1.77%
Poultry Further Processing and Rendering (Subcategory L)
PSES1
PSES2
PSES3
PSES4
15



$8,897.7



$444,885.5



2.0%



-0.5%



NA
NA
1.91%
1.91%
Poultry First Processing, Further Processing, and Rendering (Subcategory K)
BAT1
BAT2
BAT3
BAT4
BAT5
6




$12,518.7




$625,934.1




2.0%




-0.5%




NA
NA
1.89%
1.87%
NA
NA
NA
-11.33%
' -11.67%

NA
NA
-4.30%
-4.67%

NA
NA
-5.49%
-6.57%
NA
                      5-89

-------
IPSESI
1PSES2
PSES3
IPSES4
         Number
              of
        Faculties
               12
                                       Table 5-17 (cont.)
                         Impacts to Return on Assets Ratio: Retrofit Costs
                                 Meat Type and Process Classes
                                            Baseline Return on Assets
   Net Income
     (x $1,000)
     $13,650.2
                                Total Assets
                                  (x $1.000)
Median
                                   2.0%
  Lower
Quartile
   -0.5%
      Post-
Compliance
     ROA3
        NA
                              NA
                            1.76%
                            1.75%
                                                                      ory L)
 BAT1
1BAT2
 BAT3
 BAT4
$4,510.3



$82,OC



                                                    5.5%
                                                               NA
                                                               NA
                                                             5.40%
                                                             5.31%
 \Rendering (SubcateRory J)
BAT1
IBAT2
BAT3
BAT4
2



                        $2,080.0
                  $104,001.6
                                                     2.0%
                                                                 -0.5%
 liPSESl
75
                        $2.076.0
                   $103,800.7
                                                     2.0%
                                                                 -0.5%
                                                                                NA
                                                                                NA
                                                                              1.88%
                                                                              1.85%


                                                                                 NA!
                                                                                 NA
                                                              1.81%
                                                              1.79%
                                       Percent
                                       Change
                                        ROA
                                                                           NA
                                                                                            NA
                                                                                       -11.99%
                                                                       -12.32%
                                                                                            NA
                                                                                         rl.77%
                                                                                         -3.46%
PSES1
PSES2
PESS
isis4
97

$4,510.3

$82,004.8

5.5%

1.3%
	
NA
NA
5.02%
4.78%
NA|
NA
-O.I L 70
• -13.03"%!
                                                                                             NA
                                                                                             NA
                                                                                             NA
                                                                                             NA
                                                                                          -9.63%
                                                                                         -10.50%
                                 s UJ certainty facilities is not applicable for these impacts             .
                                 the average of results for each class, discharge type and model fadhty size
                          S, Nonns and Key Bปb- Ratios, 1997-98.  Median and ,ower ..uartiie
                           Pos^ Annua.ized CoSB,,(To,a, Asse,s
   Calculated as: (Postcompliance ROA - Baseline ROA)/Baselme ROA.
                                                                     Costs,.
                                                5-90

-------
              Subcategory J:.
              —     rendering
              Subcategory K:
              —     poultry first processing
              —     poultry first processing, further processing, and rendering

              Subcategory L:
              —     mixed further processing
              —     poultry further processing
                                                                                      -0.7 percent
-3.0 percent
-2.5 percent
-5.5 percent
-3.3 percent
-1.8 percent
-3.5 percent
5.5    CORPORATE FINANCIAL DISTRESS

       The relevant decision making entity above the site level is the parent company, which may own
multiple sites that produce meat products. The corporate financial distress analysis identifies situations
where it might make financial sense to upgrade each individual site but the company a.s a whole cannot bear
the combined costs of upgrading all of its sites.  Using the methodology describes in Chapter 3, EPA
performed a preliminary Altaian Z' analysis based on responses to the detailed survey, information
presented in the industry profile (Chapter 2), and estimated facility level compliance costs.

        Table 5-18 summarizes the results of the preliminary Airman Z' analysis performed for the 20
companies with sufficient data available. In the table, first, the number of companies whose baseline
Airman Z' score falls into the "financially healthy" (Z' score greater than 2.9), indeterminate (Z' score less
than 2.9 but greater than 1.23), and "financially distressed" (Z' score less than 1.23) ranges are presented.
This is followed by the number of companies whose Z' score changes from one category to another as a
result of incurred compliance costs. Thus, for example, under BAT 1/PSES 1 compliance costs, the "-1"
 indicates that the Z' score for one poultry company that was "financially healthy" in the baseline fell below
 the 2.9 threshold, and the "+1" indicates that its Z' score moved into the "indeterminate" range; the zero
 indicates that no companies had Z' scores that moved into the "financially distressed" range due to the
 compliance costs. Although a change from "financially healthy" to "indeterminate" is considered an
 impact, it is not as significant in magnitude as a change from "financially healthy" or "indeterminate" to
 "financially distressed."
                                                5-91

-------
                                       Table 5-18
                                    Altaian Z' Results
                                           Number of Companies with Z' Score:
                                                     Less Than 2.9;
                                                   Greater Than 1.23
                                          Less Than 1.23
Greater Than 2.9
                  Post-Regulatory Incremental Change (Relative to Baseline)
BAT1/PSES1
BAT2/PSES2
BAT3/PSES3
 BAT4/PSES4
 BAT5/PSES4
 BAT3/PSES01
1 Compliance costs per pound of meat type are a weighted average of BAT costs for direct dischargers and zero
costs for indirect dischargers (i.e., the realistic scenario).
2 BAT 3 costs assigned to all facilities (i.e., the worst case scenario).
                                            5-92

-------
       EPA performed the Altaian Z' analysis on 9 red meat companies, 10 poultry companies, arid one
rendering company.  For the purpose of presenting the results of this analysis, rendering is included in the
red meat sector.

        In short, essentially one major red meat company has an Altaian Z' score that is in the
"indeterminate" region in the baseline, but is close to the "financially distressed" threshold.  Under
BAT2/PSES2, BAT3/PSES3, and BAT4/PSES4, this company is projected to become "financially
distressed." Furthermore, one major red meat company with a baseline Altaian Z' score in the "financially
healthy" range is projected to become "indeterminate" under BAT4/PSES4. There are no financial distress
impacts under the proposed option.

        Similarly, three major poultry companies have an Altaian Z' score that is in the "financially
healthy" region in the baseline, but is close to the "indeterminate" range. Under options BAT2/PSES2,
BAT3/PSES3, BAT4/PSES4 and BAT5/PSES5, all three of these companies are projected to move into
 the "indeterminate" region.  Under the proposed option two of the companies are projected to move into the
 "indeterminate" region, and under BAT1/PSES1 one company moves into the "indeterminate" threshold.

        Altaian Z' analysis was also performed to determine the impact of the proposed option if all
 facilities owned by each company were direct dischargers.  This was done by removing the indirect
 discharging model facilities from the production weighted averages used in the analysis.  Although this
 scenario is highly unlikely, it is useful as a worst-case scenario analysis. As observed in Table 5-18, the
 worst case scenario does not show any impacts significantly greater than the above analysis.
  5.6     MARKET AND TRADE IMPACTS

         The market model estimates the impact of compliance costs on the price and output of various
  meat products. The distinguishing feature of EPA's market model is that it explicitly incorporates cross-
  market impacts among .meat types into the analysis. The demand for meat products such as beef, pork,
  broilers, and turkey is closely related; a one percent increase in the price of pork, for example, may cause a
  0.7 percent fall in quantity of pork demanded, and a 0.2 percent increase in demand for beef.
                                                5-93

-------
                                                                      ......
themarketmodelapproach).
                     es
compliance costs per pou
                                                                                      co







    compliance costs.
                                                                      3 costs to
                                                                               direct dischargers and

                                                      5-94

-------
ป••
...




•a
• :O
. y3
0.
O
$
%
5"
J^
Q^



















,-



..o
C


as



?
C8
M
^
u







01
ofMeaiTyp
-a
o
h
a
Vf
S
ป^
w
1
"s
5
ri

. : -
PH
".':.': •<
•PQ




*w
^^ [T^
CQ r!f!
Q~
N*^

v> o
H 
•&e-
S



0053792
g


00
s
oo
en
o
o
o
fee


VO
T— 1
r— 1
(— ^
O
g

o-
5
v
c


































\
)
)
)
)


m
i —
S
o




r-
w-i
r-
^ [
0
o
0

s



0115712
o
fee-


0073207
o


en
r-~

\
)
)
>
OS

T— <
IT)
ฃ
O
g



OO
i — l
rM
0
O
o

1-H
en
vo

O
o
0123023
o


0107952
g


oo
t~~
•*!•
m
OJ

0
o
fee

t>
i
ป•— •
g
1 ฃ}
O4
































>
O4
g
O
g



o\
8
1 — 1
o
o
0
ซ/*
r— (
o
en
VI
O
o
0
o
0088346
o
fee-


1
g


ol
v
1— H
00
c-
o
o
fee

vc
OC


































>
>
)


^J~
01
^
o
g



en
O


o

g



0034718
o
feO-


.0032944
g


S
OJ
oi
• C
o
o
fee

oc
o>
OJ
^—


H
||

. ||





























                  11
                  3 ."
                  sฅ
                  Cu 35
                  >^ M

                   60
                    -

                g
                       "8-3
                        as
                       "w ซ-rf
                        3-8
S
P9
I
h


I
CQ
            I
•w a ?= if c
a?Iซ^

-------
VO
ON

-------
I






ฃ
V
2?
C5
a'S
ง.a
•g ฐ
!M .jj
SB U
ill
zii
^ -9 "SI
g g ง
... '.P*-.50 p

: "ss-t
S ,-ซ 2"
. o i ซM , ^^
' Vgl^
" '"ง11
lfu
a
il*i
dm • C
a * a
ง|^


ซ -o IH o
^ i x!i
Jii
•*-ป -*-* fee • If
2 5 ป 1
1ซJ
_ ฐ 0 -
?5 JS *- -52 .
i ซ 03 O
ซn ซ to ซ
a> 84 o d<
* ซ ซ ^
H -a 2 ts
a^*
8CQ jg

' งซป
.2 .2 a
•ง, S 8
a a u
sis
il*
ซ ••&
.ฃ, a.
Jo
1 '
ฃ




ง1 xi
IN


W >j.i-t '^>
1 H x. 1

p ^

zp


II

s>
5
i
5
1
^
s
C3


Hi
F
9
r-
i—
c
s

r-
r~
•n
o


c<
•^
o
• vr


T
r
o
c


V
c
c
V
c



<
s
<
<
(



-
•i
0
. -
(.
!
1
i
i ^





,
)
>
s-

cr
H ^~
1- ' ' ^
T c


T C<
i- c<
0 0
5 ?


r
0 0
si C


o c
(0 C
o c
d v
N <



D *
o
N
3N
~4


1 — (
&<3-


.1
1
m

r^
Kt-
9 .
o







3 ^
t. r-H
' 9


11
5 9


o S
^ 5


II



ซO i^SS
P.ฃ S^"
s >n
-C 9
3 9


*""' J^
"*• o
^ S

^ &
•4-1 O.
nj e
"a K
3) ^

i
\








>0
o
ฐi
* — i


CO
*— t
OO
00


S5
.00



VO

J5
o
s

c<
ON
•sf


VO
CN
oc
•*
CN


-












j















•g
• p
ฃ

ซ1



•<*
r-
ON



vo


OO
•&<;


4.
J
u
n
ff






















00
ON
OO
•*


"*
ON
r--
S"


IT)


r-
oo
VO
ON"



CN
ON
vq


00
c
•s>e

1
ซ
5
t
•4—
C/
fl









^
^
9


CO
t-H
9


ซ*ซ
o
o
o
























d





1
9
i
c5
•ae-

ON
r-
co


ON
i — 1
ON
1 !



I
00
T— I
9



00
-1 — 1
d


00
CO
d

r &
i 1
fC
i b^
H










3
j



1—














5
a
k




L^*



OO
.ON


ON
VO
d
•e/9


|

i









OO
r-
co


V~l
1 — I
ON






CN
ON



CO
ON


ON
vo
0
•s/3

C
ti
5
4>-
C/
&









ON
i — 1
O


9


o
o
o


g
d



g
9


VO
i — i
d

D
bO
"I
* C^*
J
>
4

ON

-------
projected to be somewhat lower. Under the worst case scenario, the largest impacts are again seen under
chicken; the price of chicken increases by 0.4 percent, domestic supply decreases by .0.2 percent, and
exports by almost 0.5 percent (see Table 5-21).
 5.7     IMPACTS ON OUTPUT AND EMPLOYMENT

        Changes in output and employment are directly proportional to costs of compliance, that is, higher
 costs lead to lower output and employment. The impacts resonate through the economy causmg a "npple
 effect.  EPA used the Department of Commerce's national final demand multipliers from the Regtonal
 mput-OutPutModelingSystemtoestimatetheseeffects(RIMSn;U.S.DOC, 1996).

         The methodology used for the input-output analysis is explained in Section 3.1.5.  The final
 demand output multipliers used here are4.96 for red meat and4.35 for poultry, which means that *or every
 $1 million of outputlost in the red meat and poultry industry, an additional $3.96 million and $3.35 muhon
 respectivelyislostthroughouttheU.S.economy. The employment multipliers are 46.93 for red meat and
 45 18 for poultry.  That is, for every $1 million in output loss in the red meat industry, 46.93 full-Ume
  equivalent (FTEs: 1FTE equals 2,080 hours and can be equated with one full-time job) jobs are lost m the
  U.S. economy (see Section 3.1.5.1 for more detail).

          The larger the compliance costs, the greater the output and employment impacts. This is the
  reason why the subcategories with the largest impacts will be the same as those with the largest costs
  presented in Section 5.1.1. Moreover, impacts estimated with the use of upper-bound costs will be lugher
  than those estimated with retrofit costs. Table 5-22 presents the output and employment impacts stemmmg
  from the various subcategories and discharge options using both upper-bound and retrofit costs.  As the
   table shows, for the direct dischargers with the use of new equipment costs, the largest impacts are seen
   UnderBAT4inSubcategoryAthroughD.  This option results inalossof $542 million per year in output
   (0 006 percent of 1999 U.S. GDP, $9,268.6 billion (U.S. DOC, 2001)) and a loss of 4,084 FTEs (0.003
   percent of 1999 U.S. employment, 128.9 million (U.S. DOL, 2002)) for the U.S. economy as a whole.
   These losses are spread over a wide variety of industries in addition to the meat products industry. Also
   note that the input-output methodology used for this analysis overestimates changes in output and

                                                 5-98

-------
          Table 5-22
Output and Employment Impacts
                                     Total Loss in Employment2
                                            ($Miilions)
                  "

                                              (244)
                                            (5,245)
                                            (3,332)
                                            (4,176]
                                             ($52)
                                            ($697)
                                            ($443)
                                            ($555J
SES1
SES2
SES3
SES4
.
Subcqteeo
BAT1
                                               (651)
                                              (3^534)
                                              (2,897
                                               3,815)
                                              ($86)
                                             ($469)
                                             ($385)
                                             ($507)
                      ($6
                     ($107)
                     ($128)
                     ($134)
 SES1
PSES2
 SES3
 SES4
 Subcate
 BAT1
 BAT2
  AT3
 BAT4
 BATS
                                                     0
                                                  (161)
                                                (1,612)
                                                (2,041)
                                                (2,203:
                                                      $0
                                                    ($19)
                                                   ($195)
                                                   ($247;
                                                    $266:
                    5-99

-------
                                          Table 5-22 (cont.)

                                   Output and Employment Impacts
                Pretax Annualized Costs

                       (SMillions)
Subcategory

and Option
-===
L Costs
-'
Retrofit


$117
$122
— ,. . '


$2
$3




$69
$87
=— — : =—========
Total L.OSS in Output *
($Millions)
Upper-Bound
($44)
($761)
($536)
($550)
Retrofit


($508)
($529)
Total Loss in Employment 2
($Mifflbns) _
Upper-Bound
(361)
(6,298)
(4,433)
(4,551)

$0
($D
($12)
($17)
($16)


($9)
($12)

0
(10)
(98)
.(144)
(128)

($61)
($424)
($300)
($379)
•-"


($299)
($378)
(508)
(3,5 10;
(2,485)
(3,137)
3=:^=S=S=K=5=
Retrofit


(4,199)
(4,379)



(73)
(101)




(2,475)
J 	 (3,129)
                              ฐ^^
 loss in output in the affected industry.
 ป Based on 47 jobs lost in the red meat industry and 45 in the poultry industry per $1 million change m output.
                                                   5-100

-------
employment because it does not allow for impact reducing substitutions between final products by
consumers or inputs by producers.

       The output and employment losses under the proposed options (BAT 3 for Subcategories A
through D, E through I, K, and L, and BAT 2 for Subcategory J), with the use of upper-bound costs are as
follows:              .
               Subcategory A through D:
               Subcategory E through I:
               Subcategory J:
               Subcategory K:
               Subcategory L:
$274 million
$3 million
$3 million
$195 million
$12 million
2,061 FTEs
   24FTEs
   19 FTEs
1,612 FTEs
   98 FTEs
        For tilt indirect dischargers, the largest impacts are seen under PSES 2 in Subcategory K. Undei
 this option, output losses total $761 million and employment losses equal 6,298 FTEs for the economy as a
 whole.

        Using retrofit costs, output and employment impacts are less severe.  For the proposed options, the
 impacts are as follows:
                Subcategory A through D:
                Subcategory E through I:
                Subcategory J:
                Subcategory K:
                Subcategory. L:
 $194 million
 $2 million
 $3 million
 $139 million
 $9 million
 1,463 FTEs
    19 FTEs
    19 FTEs
 1,148 FTEs
    73 FTEs
 5.8     NEW SOURCES

         EPA examined the possibility that the proposed rule may create incremental barriers to entry in the
 meat products industry. EPA used a variety of sources to estimate the entry rate of new firm into the meat.
                                              5-101

-------
products market. Using the U.S. Small Business Administration's "births and deaths" database (U.S.
SBA, 1998), EPA determined that over the 1995 to 1998 time frame, new establishments entered the meat
products industry ("births") at a rate of about 5.7 percent per year (i.e., the average ratio of new
establishments to existing establishments). Conversely, the same data show that existing firms have exited
the industry ("deaths") at a rate of 6.8 percent per year.3

        However, as reflected in the industry profile (Chapter 2), other sources indicate that the sectors
 composing the meat products industry are experiencing very different growth rates. Because the "births
 and deaths" database only tracks changes at the industry level (i.e., the 3 digit SIC level), EPA estimated
 the differential growth rates for the poultry and red meat sectors based on other data sources.  EPA used a
 published study of structural change in the poultry industry (Ollinger, et. al, 2000) based on Census'
 longitudinal database to estimate that ratio of new establishments to existing establishments over the 1967
 to 1992 period. Because the overall industry new establishment rate is a weighted average of the dtfferent
 rates in the poultry- and red meat sectors.EPA was able to calculate that the ratio of new establishments to
 existing establishments in the red meat sectors over the same time period.

       '  In summary, EPA estimated the ratio of new establishments  to existing establishments in the meat
  products industry as:

                 Overall industry average: 5.7 percent per year, which reflects a weighted average of the:
                 _     Poultry sector: 19 to 26 percent per year, and the
                 	     Red Meat sectors: 3 to 3.9 percent per year.

   Note that due to disparate data sources and time frames for these analyses, the rate of new entrants can
   only be interpreted as an approximate measure.
          A potential source of barriers to entry is the incremental capital costs the proposed rule may
   impose on an entrepreneur entering the meat products market.  If, in addition to the capital necessary
to
    data are consistent with the industry profile presented in Chapter 2.
                                                  5-102

-------
               „ invest consider* capM in
a decrease in the
                            e of return

 and would therefore act as a barrier to entry.
  wo
uld pay d. sanซ prices tt .abo,
                                                                                       on

   o.teincre.nen.a.capi.a.necessary.oen^ftemeatprcdu^Wus.ry.

















    impact analysis. EPA scalea toiai a                ^           r,raf1ctreet' s /nrfwsfry Norms and Key
                                                                                                 *
     entry into the meat products market.
                                                   5-103

-------
                                         Table 5-23
                             Ratio of Capital Costs to Total Assets
                                  40 CFR 432 Subcategories
                                                                                 Capital Costs
                                                                                     to Total
                                                                                  Assets Ratioll
                                                            Average
                                                        Capital Costs
                                                           x$l,000
Model Facility
  Total Assets
     x$l,000
Number of
  Facilities
Option
Suheateporv A through D
      BAT1
      BAT2
      BATS
                                                                 $0
                                                               $125
                                                              $4,161
                                                                       $535
                                                                    $10.409
                                                                     $7,670
                                                                    $10,046
PSES2
PSES3
                                                                         $8
                                                                       $129
                                                                      $1,683
       PSES1
       PSES2
  Subcatego
        BAT1
        BAT2
                                                                                  0.00%
                                                                                  0.00%
                                                                       $1,103
                                                                       $1,614
                                                                       $1,746
  PSES2
  PSES3
    ubcateporyK
         BAT1
         BAT2
         BAT3
         BAT4
         BATS
                                                                                   0.55%
                                                                                   0.62%
                                               5-104

-------
                                      Table 5-23 (cpnt.)
                             Ratio of Capital Costs to Total Assets
                                  40 CFR 432 Subcategones
                                                      Capital Costs
                                                           toTotal
    Average
Capital Costs
                                          Model Facility
                                             Total Assets
Numberof
       Ues
       138
        $307
        5,590
       $4,616
       $4,860
Option
      PSES
      PSES2
      PSES3
      PSES4
  ubcategor  L
       BATl
       BAT2
       BAT3
  other BAT options.
                                                   5-105

-------
capital costs compose an average of 3! 1 percent of facility assets. For direct dischargers, the largest impact
is 1.7 percent, which occurs under BAT 4 in Subcategory A through D.

        Under the proposed options - BAT3 for all subcategories except J, for which BAT 2 is specified
— the ratio of incremental capital costs to total assets for each subcategory is:
                                                                                       0.82 percent
                                                                                       0.08 percent
                                                                                       0.00 percent
                                                                                       0.42 percent
                                                                                       0.38percent
       •      Subcategory A through D:
       •      Subcategory E through I:
       •      Subcategory J:
       •      Subcategory K:
       •      Subcategory L:

The largest impacts thus occur hi Subcategory A through D.
         Table 5-24 presents the ratio of incremental upper-bound capital costs to total assets at the meat
 type and process class level. The largest impact is observed under PSES 4 in the mixed further processing
 class, where the capital costs compose an average of 4.24 percent of facility assets. For direct dischargers,
 the largest impact also occurs in the mixed further processing class, where incremental capital costs are 2.4
 percent of total assets under BAT 4.

         Under the proposed options the overall ratio of incremental capital costs to total assets at the
  subcategory level represents a range among the component classes of:
                 Subcategory A through D:
                 	     red meat first processing
                 	     red meat first processing and rendering
                 Subcategory E through I:
                 	     red meat further processing
                 	     mixed further processing
                  Subcategory J
                  —     rendering
                                                                                        0.82 percent
                                                                                        0.00 percent
                                                                                        1.35 percent
                                                                                        0.08 percent
                                                                                        0.01 percent
                                                                                        0.78 percent
                                                                                        0.00 percent
                                                 5-106

-------
            Table 5-24
Ratio of Capital Costs to Total Assets
   Meat Type and Process Classes
Option
Number of
Facilities
Total Assets
• fr $1,000)
Average
Capital Costs
(x $1,000)
Capital Costs
0 to Total
Assets Ratio
Red Meat First Processing (Subcategorv A - D)
BAT1
BAT2
BATS
BAT4
6



$50,870.6



$0
$0
$0
$801
0.00%
0.00%
0.00%
1.57%
Red Meat Further Processing (Subcategorv E-I)
BAT1
BAT2
BATS
BAT4
12



$139,107.7



$0
$4
$21
$1,058
0.00%
0.00%
0.01%
0.76%

PSES1
PSES2
PSES3
PSES4
168



$121,672,2



$236,
$1,231
$1,223
$1,720
0.19%
1.01%
1.00%
1.41%
Red Meat First and Further Processing (Subcategorv A - D)
PSES1
PSES2
PSES3
PSES4
28



$94,015.5



$274
$3,918
$3,783
$3,935
Red Meat First Processing and Rendering (Subcategory A - D)
BAT1
BAT2
BATS
BAT4
36



$553,233.8

/
,>

$0
$174
. $7,485
$8,694
0.29%
4.17%
4.02%
4.19%

0.00%
• 0.03%
1.35%
1.57%

PSES1
PSES2
PSES3
PSES4
15



$553,233.8



$663
$20,765
$14,013
$14,112
Red Meat Further Processing and Rendering (Subcategory E-I)
BAT1
BAT2
BATS
BAT4
4



$261,155.9



$0
$22
$66
$3,357
0.12%
3.75%
2.53%
•2.55%

0.00%
0.01%
0.03%
1.29%
               5-107

-------
                                        Table 5-24 (cont.)
                               Ratio of Capital Costs to Total Assets
                                 Meat Type and Process Classes
                                                                                     Capital Costs]
                                                                                          toTotall
                                                                                      Assets Ratio
                                                                                           0.20%
                                                                                           2'.03%'
    Average
Capital Costs
   (x $1,000
        $513
       $5,297
       $4,304
       $4,93
                                               Total Assets
                                                 (x $1.000)
Number of
  Facilities
         1
PSES1
PSES2
Red Meat Mrtf Processing Further Processing, and Rendering (Subcate^ry A - D
                                                                          $216
                                                                       $10,396
 Poultry First Processing (Subcategory K)
                                                                             $0
                                                                             $0
                                                                         $1,983
                                                                         $2,673
                                                                         $2,985
  Poultry Further Processing (Subcatego
                                                                              $0
                                                                             $11
                                                                            $838
                                                                          $1,183
                                                                          $1,363
                                                                            $235
                                                                           $1,527
                                                                           $1,303
                                                                           $1,754
                                                 5-108

-------
         Table 5-24 (cont.)
Ratio of Capital Costs to Total Assets
   Meat Type and Process Classes
Ootion
Number of
Facilities
Total Assets
(x $1,000)
Average
Capital Costs
(x $1,000)
Capital Costs
to Total
Assets Ratio
Poultry First and Further Processing (Subcategory K)
BAT1
BAT2
BATS
BAT4
BATS
16



'
$597,645.2




$0
$64
$2,359
$3,789
$4,233

PSES1
PSES2
PSES3
PSES4
29



$594,718.8



$0
$3,316
$4,006
$4,241
0.00%
0.01%
0.39%
0.63%
'0.71%

0.00%
0.56%
0.67%
0.71%
[Poultry First Processing and Rendering (Subcategory K)
BAT1
BAT2
BAT3
BAT4
BAT5
17




$549,160.5




$0
$27
$2,787
$3,594
$4,001
0.00%
0.00%
0.51%
0.65%
0.73%

PSES1
PSES2
PSES3
PSES4
5



$557,820.1



$0
$9,283
$5,813
$6,020
0.00%
1.66%
1.04%
1.08%
Poultry Further Processing and Rendering (Subcategory L) ' •
PSES1
PSES2
PSES3
PSES4
15



$444,885.5



$176
$3,045
$2,542
$2,708
Poultry First Processing, Further Processing, and Rendering (Subcategory K)
BAT1
BAT2
BAT3
BAT4
||BAT5
6




$625,934.1




$0
$0
$6,498
$6,688
$7,507
, 0.04%
0.68%
0.57%
. 0.61%

0.00%
0.00%
1.04%
1.07%
1.20%
                5-109

-------
                                        Table 5-24 (cont.)
                               Ratio of Capital Costs to Total Assets
                                 Meat Type and Process Classes
                                                                                    Capital Costs
                                                                                         to Total
                                                                                      Assets Ratio
                                                                                           0.11%
                                                                      Average
                                                                  Capital Costs
                                                                     (x $1,000)
Total Assets
   x$l,000
  $682,511.7
Option
PSES1
                                                       ercent in Subcategory L)
Mixed Further Processing (61 percent in Subcatego
                                                                            $6
                                                                          $641
                                                                        $1,948
 Rendering (Subcategory J)
 BAT1
                                                                             $0
                                                                         $1,154
                                                                         $1,304
                                                                            $47
                                                                         $1,103
                                                                         $1,614
                                                                         $1,746
                                                 5-110

-------
               Subcategory K:
               —     poultry first processing
               —     poultry first processing, further processing, and rendering
               Subcategory L:
               —     poultry further processing
               —     mixed further processing
0.42 percent
0.32 percent
1.04 percent
0.38 percent
0.36 percent
0.78 percent
5.9     SUMMARY AND OBSERVATIONS

        Table 5-25 presents a summary of the costs and impacts under the proposed options for the meat
products industry as a whole. Using upper-bound costs, total posttax annualized costs for the proposed
options under all subcategories are estimated at $68 million.  Of the total 209 nonsmall, noncertainty
facilities affected by the rule, 0.8 facilities are projected to close as a result of the rule. Compliance costs
exceed: 1 peiocnt of revenues for 18 facilities (8 percent of facilities), 3 percent of revenues for 4 facilities
(2 percent of all facilities), and 5 percent of cash flow for 22 facilities or 10 percent of facilities. Output
losses in U.S. are expected to total $487 million per year and employment losses are estimated at a total of
3,800 FTEs per year.  Including the 65 certainty facilities, costs and impacts increase by a margin of 8
percent. Total posttax industry compliance costs increase by $6 million and now equal $74 million.
Facility impacts include 1 facility closure and 24 facilities with compliance costs greater than 5 percent of
cashflow.

        With the use of retrofit costs instead of new equipment costs, total posttax  annualized costs for the
industry are $47 million.  The number of facilities projected to close as a result of the rule are 0.4. Five
percent or 12 facilities have compliance costs greater than 1 percent of revenues, 3  facilities have costs
greater than 3 percent of revenues, and costs for 16 facilities are greater than 5 percent of cash flow.
Annual output losses for the entire U.S. are estimated at $347 million and employment losses at 2,700
FTEs.  With the 65 certainty facilities, total posttax costs increase to $50.5 million, 0.4 facility closures are
projected, and for 17 facilities, compliance costs are greater than 5 percent of cash  flow.
                                                5-111

-------

-------











1

t-


1*
P
8 ซ
S"!
v> jg
C^ O
11
1— (
•s
ฃ•
I












1!H
•g

1
hi
&<
CO JS>
Se
|g
pfl
13 .S
| "1}
[2 o ป |
ง5 g ^
nS ^^
p
s*
z
*, Cta4
S e
i


Prbb.
Cash Flow
111
H S^
^ ?



^
1

c
P"

= i
ซ 0














W fl

CU S
bl
12 ^*
o
^

Less than
';,.;• .'/C'osts
11




N"""'
•g |
••g

•"S
cs
o ง
fio
|


















Subcategory L
oo

•*~s

^
1—1
oo
^

in

O
•n
o^



i-H
c5



r-
r-
o
oo
1-H
•e/3-





in
r-H





CO
ii
Q"t O
P m
m

> — '

r*
'*~/
en
^

oo
r—i
en
o
•n
1 r"'



C3



in
o
rn
i-H





i—t





m
1
1
Oil
^
r>H
OO
JS
^
oo
s
g


ซi
i-H
l-J
SfiS
r^*



00
o



z
06
so





ง
r*



1
i
•1
1
o\
i— i
T-l
S
G1
E
-s-




5
•
^t



ฐ5



Z
oo





ve
f^i



'TS ซ
1 Total Upper-Bou
Including 65
Certainty Faciliti
rT

t^
J3
f^1
S
g


V)

5
so




o



Z
<





S
c^




l_ 	
Total Retrofit1
0
Tf
ฐi.

ST
1
' 2


t>

S
V)




o



z
i





8
c^



^
1 Total Retrofit
Including 65
Certainty Faciliti





















1
ซ
o
JD
a
00
4n
^ ^3
a. e
8 g

S 3
1 fl
|J
ง* g

-------
5.10   REFERENCES


Dun & Bradstreet. 1998. Industry Norms and Key Business Ratios, 1997-1998. Desk-Top Edition.

Ollinger Michael, James MacDonald, and Milton Madison. 2000. Structural Change ™U-S^Chicken and
       ^Slaughter. Agricultural Economic Report No. 787. Washington, D.C, U.S. Department of
       Agriculture, Economic Research Service.

US Department of Commerce, Bureau of Economic Analysis. 1996. Regional input-output modeling
       system (RIMS II). Total multipliers by industry for output, earnings, and employment.
       Washington, DC.

U.S.DepartmentofCommerce,BureauofEconomic Analysis. 2001. Gross Domestic Product by
       Industry: 1947-2000. Downloaded on January 14,2001.

 U.S. Department of Labor.  2002.  Bureau of Labor Statistics Data.  Nonf arm EmploymenU991 - 2001.
        Available at: http://data.bls.g™/^-b™/survevmost. Downloaded on January 15, 2UU2.

 U S SBA ' 1998. Statistics of U.S. Businesses: Firm Size Data: Dynamic Data: Download US. industry
        toup data,  1990-1998 one year changes and 1990-1995 (U.S. Births, deaths, and job creatton by
        U.S. industry group, 1990 - 1998.) U.S. Small Business Administration, Office of Advocacy.
        Available at: http://www.sba.gov/advo/stats/data.html.
                                              5-114

-------
                                       CHAPTER 6
                INITIAL REGULATORY FLEXIBILITY ANALYSIS
 6.1     INTRODUCTION

        This chapter analyzes the projected effects of incremental pollution control costs on small entities.
 This analysis is requked by the Regulatory Flexibility Act (RFA) as amended by the Small Business
 Regulatory Enforcement Fairness Act of 1996 (SBREFA). The RFA acknowledges that small entities have
 limited resources and makes it the responsibility of the regulating federal agency to avoid burdening such
 entities unnecessarily. In response to the RFA, EPA has prepared an initial regulatory flexibility analysis
 (IRFA). Section 6.2 provides the initial assessment to determine if an IRFA is necessary.  Section 6.3
 describes the components of the IRFA. Section 6.4 presents the analysis of economic impacts to small
. businesses in the meat products industry, while Section 6.5 summarizes the steps EPA has taken to
 minimize small business impacts under the proposed rule.
  6.2     INITIAL ASSESSMENT

         EPA guidance on implementing RFA requirements suggests the following must be addressed in an
  initial assessment.  First, EPA must indicate whether the proposal is a rule subject to notice-and-comment
  rulemaking requirements. EPA has determined that the proposed meat products effluent limitations
  guidelines (ELG) are subject to notice-and-comment rulemaking requirements. Second, EPA should
  develop a profile of the affected small entities. EPA has developed a profile of the meat products industry,
  which includes all affected operations as well as small-businesses.  This information is provided in Chapter
  2. Chapter 5 of this EA presents the analysis of projected economic impacts to the industry as a whole,
  including both small and large businesses. Much of the information covered in these chapters applies to
  small businesses.  Additional information on small businesses in the meat products industry is provided in
  Section 6.4 of this chapter.  Third, EPA's assessment needs to detennine whether the rule would affect
  small entities and whether the rule would have an adverse economic impact on small entities.

                                                6-1

-------
                           costs for incre
                                          incremental pollution control as a
in Section 6.4.
                          ANALYSIS COMPONENTS
                u ,  „ TRFA must contain the following:
             ires mat ป MA mu
     REGULATORY I
      Section 603 of te EFA requires

           A, explanation of why the rute may be needed.

      '
                                          .
 .3.1  NeedforObJecttvesoftheR*
         6.3
„„—- —   -
                                                      blish effluent
                                                   ป*—ป•<ป
                                    6-2

-------
EPA to issue BPT effluent limitations guidelines. Section 304(b)(4) authorizes EPA to issue BCT
guidelines for conventional pollutants; Sections 301(b)(2)(E) and 304(b)(2) authorize EPA to issue BAT
guidelines to control nonconventional and toxic pollutants; Section 306 authorizes EPA to issue NSPS for
all pollutants; and Sections 304(g) and 307(b) authorize EPA to issue PSES and PSNS for all pollutants.
        6.3.2   Estimated Number of Small Business Entities to Which the Regulation Will Apply

        The RFA defines a "small entity" as a: (1) small not-for-profit organization, (2) small
 governmental jurisdiction, or (3) small business. EPA expects that the principal impact of the proposed
 rule will fall on small businesses in the meat products industry, rather than not-for-profit organizations or
 small governmental jurisdictions. Therefore, this analysis will focus on small meat products businesses.

        The RFA defines a "small business" as having the same meaning as the term "small business
 concern" under Section 3 of the Small Business Act (unless an alternative definition has been approved).
 The latter identifies a small business at the business entity or company level, not the facility level. The
 analysis, then, needs  to determine whether a facility is owned by a small business entity, not whether the
 facility itself may be considered "small."

         A small business is generally defined according to NAICS code by standards set by the Small
 Business Administration (SBA).  Under NAICS codes 311611, 311612, 311613, and 311615, a small
 business is defined as one with fewer than 500 employees.  Note that a facility may employ fewer than 500
 employees but not be considered "small" by this standard if it is owned by a larger parent company and
 total employment among all facilities that company owns exceeds 500 workers  (U.S. SBA, 2000).

         As stated above, it is important in determining the number of small business entities in the meat
 products industry to  differentiate between facilities owned by small businesses and small facilities owned
 by large businesses.  To make this differentiation, EPA used ratios of firms to establishments in the meat
 product industry derived from data compiled by the  U.S. Census Bureau for the Small Business
 Administration's Office of Advocacy (U.S. SBA, 1998). These ratios were calculated by dividing the
 number of firms within each NAICS code and employment class by the number of establishments in that
 code and class. EPA then applied this ratio tp model facilities in each meat type and process class
                                                6-3

-------
determined to have an employment range below 500 employees in order to estimate the proportion of
facilities that are stand alone small businesses relative to facilities owned by large businesses.1- 2

        In essence, EPA is assuming that within any NAICS code and employment class combination
Where the ratio of firms to establishments is less than one, establishments in excess of the number of firms
are all owned by large, multi-facility business entities. This is a reasonable assumption for the meat
products industry ffiFA.  EPA determined the employment ranges for each meat type and process class
based on its model facilities, which were matched to Census employment classes using annual production,
 estimated revenues, and other Census data (see Section 3.1.2.6 and Appendix B for details),  to tins
 matching process, EPA found:

                small model faculties invariably fell into employment classes with fewer than 10 workers;
                Sฃ of firms to establishments for the 1 to 4 and 5 to 9 employment classes is 1.0
                based on SB A' s database;
         .      medium, large, and very large model facilities (hereafter, "
                into employment classes with at least 250 to 499 workers (see Table B-6 for details)  and
                SL ^geHfto facilities employing between 250 and 499 employees are owned by a
                single company, that company in all likelihood would be a large business.

  For example, EPA determined there are 170 medium sized facilities in the red meat further processing
  class The medium sized facility was matched to Census data in the 250 to 499 employment range. The
  ratio of firms to establishments in this employment range and NAICS code is 0.825. Therefore, EPA
  assumes that 140 (= 170 x 0.825) of these facilities are small stand alone businesses; the remaining 30
  facilities are owned by large business entities.3
           > Clearly individual facilities employing more than 500 workers are large business owned, whether they
   are & stand alone business or owned by a larger entity.
                                                                                       in a
           2 EPA determined from publicly available sources that this
    (which is greater than the number of large businesses).
                                                  6-4

-------
       Tables 6-1 and 6-2 present the estimated number of stand alone small businesses, the number of
facilities that are owned by a large business, and the total number of entities in each model facility size
classification for the meat products industry. Table 6-1 provides the information by subcategory, while
Table 6-2 presents the information by meat type and process class.

        EPA estimates that a total of 5,174 out of 5,671 potentially affected facilities (91 percent) are
 small business owned under the 500 employee standard; an estimated 497 facilities (9 percent) are owned
 by large businesses.4  Subcategory E through I contains the most small business entities, 3,179 (98 percent
 of the subcategory), followed by Subcategory A through D with 1,065 (90 percent of the subcategory).
 Subcategory L is estimated to have 745 small businesses (94 percent). Seventy-three of the 119 facilities in
 Subcategory J (61 percent) are estimated to be small business owned. Subcategory K is the only
 subcategory in which less than half of facilities are estimated to be small (45 percent).

         By meat type and process class, facilities that perform poultry first processing operations, whether
  alone or in combination with other processes, tend to be owned by large business entities CTable 6-2). This
  tendency is not as strong among red meat first processors. Conversely, facilities that only perform further
  processing operations, whether for red meat or poultry, tend to be small stand alone businesses.


          6.3.3   Description of the Proposed Reporting, Recordkeeping, and Other Compliance
                 Requirements
          EPA has incorporated no incremental reporting or recordkeeping requirements in the proposed rule.
   Technical requirements are described in detail in the Development Document (U.S. EPA, 2002).  A brief
   summary of treatment technologies that will meet the effluent guidelines is presented in Chapter 4 of this
   document.
           * EPA determined from publicly available sources that the 65 certainty facilities (see Chapter 5) are all
    owned by large business entities.
                                                   6-5

-------
                                         Table 6-1
             Meat Product Industry Estimated Small Business Owned Facilities
                                 40 CFR 432 Subcategories
.-•-j — • 	 --' 	 ' 	 i 	 i 	 ~~T~
Model Facility Size
Subcategorv A through D
Small
Medium
Large 	
Very Large 	 	
Subcategory E through I
Small
Medium
Large 	
E/ery Large 	
jubcategory J 	
Small
Medium
Large 	
Very Large 	
Subcategory K 	
Small
Medium
Large 	
Very Large 	
Subcategory L 	
Small
Medium
Large 	
Very Large 	
Estimated Number of Facilities
Number of
Facilities*
Small Business Owned*

1,060
87
22
17
1,060
5
0
0

2,988
243
5
5
2,988
191
0
0

23
33
27
36
18
19
9
27

39
80
99
47
39
71
0
0

572
192
11
20
572
168
4
0

Small Total
Medium Tota
	 	 Large Total
	 Very Large Tota
	 Certainty Facilities
1 TOTAL
4,682
634
164
[ 125
65
5,67C
4,677
455
13
21
c
5,174
targe Business
...'• •••.'...;, •.->--:;.--OWjied*

0
81
22
17

0
52
5
5






0
9
99
47
|
o||
24|
6
20

5
179
150
' 981
) 65|
1 497]
* Numbers may not sum due to rounding.
Based on Screener Survey, Census Model Facilities, and SBA Special Tabulations.
Small business to large business owned ratio calculated from the Small Business Administration's establishment
and facility comparison data compiled by the U.S. Census Bureau.
Subcategories not multiplied by the ratio were those classified as having over 500 employees.
                                               6-6

-------
                                      Table 6-2
            Meat Product Industry Estimated Small Business Owned Facilities
                             Meat Type and Process Classes
                                                              Number of Facilities
                                                                          Large Business
                                                                                Owned*
                                                     Small Business
                                                          Owned*
Number of
Facilities*
                       Subcategory A - D)
edMeat First Processin
—~	:             I               282
mall	—-
 •  i-                       I                  6
 .edium	        1	•.—
ed Meat Further Processing (Subcatezory E -1)
V ^J.J J-*i*J.^p^'                             	
Red Meat First and Further Rendering (Subcategory A - D
  ed Meat First Processing and Rendering (Subcatego
  'ed Meat Further Processing and Renderin
  ..* **.ป F,ve, Pmr^in*. Further Free****: and Rendering (SubcatexoryA - D)
                            I               f-9f                    TT
 \Poultry First Processing (Subcatego
         Further Processin
                                              6-7

-------
                                       Table 6-2 (cont.)    ^.
              Meat Product Industry Estimated Small Business Owtปv
                              .  Meat Type and Process Classes
-:".
Model Faculty Size

lumber of
Facilities*
Estimated Number of iFac^
Small Business
Owned*
•-••;..-•;• 'Largetbv.
L ':.'.'•'•...•:• ownK
Poultry First and Further Processing (Subcategory K)
Small
Medium
Large
Very Large
20
17
6
22
20
15
0
0
0
2
6
22
Poultry First Processing and Rendering (Subcategory K)
Medium
Large
Very Large
9
10
3
8
0
0
1
10
3
Poultry Further Processing and Rendering (Subcategory L)
Small
Medium
Large
4
9
6
4
8
0
0
1
6
Poultry First Processing, Further Processing, and Rendering (Subcategory K)
Medium
Large
Very Large
5
10
3
4
0
0
1
10
0
Mixed Further Processing (59% Subcategory E- 1 and 41 % Subcategory L)1
Small
Medium
716
102
716
84
0
18
Mixed Further Processing and Rendering (59% Subcategory E - 1 and 41 % Subcategory L) '
Small
4
4
0
Renderer (Subcategory J)
Small
Medium
Large
Very Large
23
33
27
36
18
19
9
27
c
14
18
9

Small Total
Medium Total
Large Total
Very Large Total
Certainty Facilities
TOTAL
4,682
634
164
125
65
5,671
4,677
456
13
27
0
5,174
{
179
150
98
65
497
1 For nonsmall facilities, the allocation is 61% in Subcategory E through I and 39% in Subcategory L.
* Numbers may not sum due to rounding.
Based on Screener Survey, Census Model Facilities, and SBA Special Tabulations.
Classes with zero number of facilities were excluded from the table.
Small business to large business owned ratio calculated from the Small Business Administration's establishment
and facility comparison data compiled by the U.S. Census Bureau.
Classes not multiplied by the ratio were those classified as having over 500 employees.
                                                6-8

-------
       6.3.4   Identification of Relevant Federal Rules That May Duplicate, Overlap, or Conflict
               with the Proposed Rule

       The current meat products rule, 40 CFR Part 432, set effluent guidelines and limitations for the
beef and pork sectors of the meat products industry. These standards were set and revised over a number
of years, most recently in 1995 (see Table 1-1 for details).  The proposed rule revises the current industry
standards in existing subcategories and thus does not conflict with them. The proposed rule does set new
standards for facilities that perform poultry slaughter and processing operations. Prior to this proposal,
EPA had set no national effluent limitations guidelines or standards for poultry slaughterers or processors.

        Much of the water used by meat products industry establishments is for sanitation purposes.
Through contact with USDA's Food Safety and Inspection Service (FSB), EPA ensured that its proposed
rule would not conflict with food safety sanitation requirements. FSIS stated that water use is only one
way for facilities to comply with food safety regulations; alternative means to meeting the requirements are
 available.  In addition, if facilities do use water for sanitation purposes, operators have options for
 recycle/reuse or end of pipe treatment that will not affect compliance (citation needed).  Therefore, EPA
 has determined that the proposed rule does not conflict with FSIS food.safety regulations.
         6.3.5   Significant Regulatory Alternatives

         EPA took steps to minimize the regulatory burden associated with the rulemaking.  Fust, EPA
  categorized the industry based upon meat type (i.e., red meat or poultry), process class (i.e., slaughter,
  further processing, rendering), and facility size (small, medium, large, and very large based on production),
  then these categories were grouped into 40 CFR 432 subcategories. Both the meat type and process classes
  and the 40 CFR 432 subcategories differentiate between direct and indirect dischargers. All direct
  dischargers were costed for four sets of technology options regardless of meat type or processing stage;
  dkect dischargers that process poultry were costed for a fifth technology option. Similarly, all indirect
  dischargers were costed for four technology options regardless'of subcategory.  Indirect dischargers were
  costed for a different set of technologies than were dkect discharging facilities.  Thus, EPA's analysis
  provided significant flexibility for tailoring the proposed guidelines according to sector specific
                                                  6-9

-------
characteristics.  Finally, EPA also performed a small business analysis of all alternatives considered for
each subcategory.


6.4     SMALL BUSINESS ANALYSIS

        This section presents the projected economic impacts on small businesses resulting from the costs
of complying with the proposed ELG for the meat products industry. The impacts are estimated using the
methodology outlined in Chapter 3. Closure impacts, costs, and nonclosure impacts for small businesses
are presented at the subcategory level and the meat type and process class level by discharge type.

        Tables 6-3 and 6-4 provide the estimated number of small business owned facilities by both
discharge type and facility size according to subcategory and meat type and process class respectively.
Among both direct and indirect dischargers, the majority of facilities are owned by small business entities.
However, while just a little more than half of direct dischargers are small business  owned (56 percent), 95
percent of indirect discharging facilities are small business owned.

        In the discussion of small business impacts below, EPA adopts the following convention for
 referring to different establishment sizes. Essentially all establishments enumerated in the tables below are
 small businesses (i.e., independent business entities employing fewer than 500 workers). However, within
 this group of small business entities, EPA distinguishes small facilities from nonsmall facilities (i.e.,
 medium, large, or very large) based on facility production.5 EPA has set the following production
 thresholds to define small facilities in each subcategory:

         •        Subcategory A through D: facilities that slaughter less than 50 million pounds (live weight
                 kill) per year;
                 Subcategory E through I: facilities that produce less than 50 million pounds of finished
                 product per year. Because Subcategory E (small processors) is defined under the existing
         5 There is a single exception to the above rule. In Subcategory J (rendering), EPA determined that 5 small
  model facilities are owned by large business entities.  With that exception, all small model facilities are also small
  business entities.
                                                 6-10

-------
                                           Table 6-3
Meat Product Industry Estimated Direct and Indirect Discharge Small Business Owned Facilities
                                   40 CFR 432 Subcategories
1 ••• •-•••'
Model Facilitv Size
SS2=2=^^^=S=^=^Sa^^^=^=j=
Number of Facilities
Direct*
Indirect*
Direct Discharge
Facilities
Small
Business
Owned*
Large
Business
Owned*
Indirect Discharge II
Facilities
Small
Business
Owned*
Large
Business
'Owned*
!'nhrntf>cnrv A through D . 	 	 , 	 , 	
Small
Medium

Very Large
59
40
14
. 12
1,001
47
, 8
5
59
5
0
0
0
34
14
12
1,001
0
0
0
Subfnip-enry E through I 	 	 	 — 	 	
Small
Medium
|,arge
f ery Large 	

imail
Medium


lubcategory K
Small
Medium
Large
Very Large 	
Subcategory L
Small
Medium
Large 	 	
Very Large

Small Total
Medium Tola
	 — 	
Large Tota
II 	 	
Very Large Tota
TOTAI
48
17
1
1
2,940
226
4
4
48
10
0
0
0
7
1
1

6
7
6
8
17
26
21
28
5
4
2
6
1
3
4
2

0
32
38'
18
39
48
61
29
0
28
0
0
0
4
38
18

4
12
1
2
568
180
10
18
4
11
1
C
0
1
0
2

[ 117
I 108
I 6C
I 41
32ซ
4,565
527
I 104
8^
i 5,28C
116
58
3
6
1 183
1
50
57
, 35
143
2,940
181
0
0

13
15
7
21

39
44
0
0

568
158
4
0

4,561
398
• 11
21
4,991
0
47
8
5

0
45
4
4

4
11
.14
7

ฐ
5
61
29

0
22
6
18

4
130
93
63
290
 * Numbers may not sum due to rounding.
 Based on Screener Survey, Census Model Facilities, and SBA Special Tabulations.
 Small business to large business owned ratio calculated from the Small Business Administration's establishment
 and facility comparison data compiled by the U.S. Census Bureau.
 Subcategories not multiplied by the ratio were those classified as having over 500 employees.
 EPA did not distribute the 65 certainty facilities between direct and indirect dischargers.
                                                 6-11

-------
                                         Table 6-4
 Meat Product Industry Estimated Direct and Indirect Discharge Small Business Owned Facilities
                               Meat Type and Process Classes
                                                                       Indirect Discharge
                                                                           Facilities
                                               Direct Discharge
                                                   Facilities
                     Number of Facilities
                                                                                      Large
                                                                                   Business
                                                                                   Owned*
                                                                          Small
                                                                       Business
                                                                       Owned*
       Large
     Business
     Owned*
  Small
Business
Owned*
Model Facility Size
Red Meat First Processing (Subcateeory A- D
                           'Subcategory E-I
Red Meat Further Processin
 WMHBIIH               ^~^~^^^m
 mall
                                   'Subcategory A-D)
                                        674
Red Meat fir.it and Further Renderin
                              0
                              0
                                         28J	0
                                      Subcategory A-D)
 Red Meat First Processing and Renderin
 Red Meat Further Processing and Rendering (Subcategory E -1)
 Small
  Medium
  Large
                                          50
 **J Meat First Processing, Further Processing, and Rendering ^ Subcategory A
                              25
                              17
25
                                                        0
  Poultry First Processing (Subcategory K)
                               j)
                              17
                              25
50
0
0

0
12
5

   Poultry First and Further Processin
                                               6-12

-------
                                         Table 6-4 (cont.)
Meat Product Industry Estimated Direct and Indirect Discharge Small Business Owned Facilities
                                  Meat Type and Process Classes

Model Facility Size
Number of Facilities
Direct*
Indirect*
Direct Discharge
Facilities
, Small
Business
Owned*
Large
Business
Owned*
Indirect Discharge
Facilities
Small
Business
Owned*
Large
Business
Owned*
Poultry First Processing and Rendering (Subcategory K)
Medium
Large 	
Very Large
7
8
2
2
2
1
6
0
0
1
8
2
2
0
0
0
2
1
^Poultry Further Processing and Rendering (Subcategory L)
Ismail ,
[Medium 	
[[Large 	
0
0
0
4
9
6
0
0
0
0
0
0
4
8
0
0
1
6
fpoufrry First Processing, Further Processing, and Rendering (Subcategory K)
Medium
Large
Very Large
2
3
1
3
7
2
2
0
0
0
3
1
3
0
0
0
7
2
Mired Further Prnre.ssint> (59% Subcategory E- 1 and 41 % Subcategory L) '
Small
Medium
9
5
707
97
9
4
0
1
707
80
0
17
Mired Further Prnr.ex.iing and Rendering (59% Subcategory E- I and 41 % Subcategory L) 1
Small
0
, 4
0
0
4
0
tenderer (Subcategory J) 	 ,- 	 	 	 , 	
Small
[Medium
F 	
[[Large 	
Very Large

Small Total
Medium Total
1 Large Total
Very Large Total
TOTAL
6
7
6
8
17
26
21
28
5
4
2
6
1
3
4
2

117
108
60
41
326
4,565
527
104
84
5,280
116
59
3
6
184
1
49
57
35
142
13
15
7
21

. 4,561
397
11
21
4,990
i
11
14


L
130
93
63
290
1 For nonsmall facilities, the allocation is 61% in Subcategory E through I and 39% m Subcategory L.
* Numbers may not sum due to rounding.
Based on Screener Survey, Census Model Facilities, and SBA Special Tabulations.
Classes with zero number of facilities were excluded from the table.
Small business to large business owned ratio calculated from the Small Business Administration's establishment and
facility comparison data compiled by the U.S. Census Bureau.
Classes not multiplied by the ratio were those classified as having over 500 employees.
EPA did not distribute the 65 certainty facilities between direct and indirect dischargers.
                                                  6-13

-------
              guidelines as facilities that produce less than 6,000 pounds of finished product per day, all
              facilities in Subcategory E are by definition small;
              Subcategory J: facilities that render less than 10 million pounds of raw material per year;
              Subcategory K: facilities that slaughter less than 10 million pounds per year;
               Subcategory L: facilities that produce less than 7,000 pounds of finished product per day.

Based on median production, all small model facilities fall below these thresholds and are thus synonymous
with small producers; all other model facilities exceed the thresholds (see Appendix B, Table B-6 for
details).

        For each level of impact analysis, EPA first presents the results for small model facilities, then the
impacts for those nonsmall model facilities that EPA estimates are owned by small businesses.  The latter
group of facilities is a subset of the facilities analyzed in Chapter 5. Thus, impacts to nonsmall facilities
presented in Chapter 6 are not additional impacts of the proposed rule, but are a subset of those impacts
presented in Chapter 5.
         6.4.1   Total and Average Compliance Costs

         Tables 6-5 and 6-6 present total and per facility costs for small business owned meat products
 facilities. The tables include estimated capital costs, annual operating and maintenance (O&M) costs,
 pretax annualized, and posttax annualized compliance costs.6 Annualized costs are analogous to a
 mortgage payment that spreads the one-time investment of a home over a series of constant monthly
 payments. They are calculated as the equal annual payments of an annuity'that has the same present value
 as the stream of cash outflow over the project life and includes the opportunity cost of money or interest
  (see Section 3.1.1 of this document for more detail on cost annualization, and the Development Document
  (U.S. EPA, 2002) for details on the estimation of capital and O&M costs).
          5 EPA did not estimate retrofit costs for small model facilities. In Section 6.4, EPA will not present
  retrofit costs for medium, large, and very large model facilities owned by small businesses. These may be found by
  scaling results from Chapter 5 appropriately.
                                                 6-14

-------
       6.4.1.1 Total and Average Compliance Costs by Subcategory

       Small Model Facilities

       As seen in the Table 6-5A, estimated posttax annualized costs for small model direct dischargers
are less than $700 per facility under BAT 1. Small model indirect dischargers average from $24,000 in
Subcategory A through D to $42,100 in Subcategory L per facility under option 1. Option 3 is the highest
cost option per facility for direct dischargers (BAT 4 was not costed for small model facilities), and option
4 has the highest cost per facility for indirect dischargers (with the exception of Subcategory J). Per
facility costs for indirect dischargers exceed $137,000 under options 2, 3, and 4 for all subcategories.

       Under the proposed option (BAT 1) for small model facilities in subcategories K and L, posttax
annualized costs per facility are:

       •       Subcategory K:                                                                 NA7
       •   .    Subcategory L:                              •                                   $711

No option is proposed for small model direct dischargers  in subcategories A through J. No option is
proposed for small model indirect dischargers in any subcategories.


       Nonsmall Model Facilities                                         ,

       Table 6-5B provides costs for nonsmall model facilities owned by small businesses.  Under the
proposed option (BAT 3 in all subcategories except J; BAT 2 in Subcategory J) for nonsmall model
facilities that are owned by small businesses, posttax annualized costs per facility are:
               Subcategory A through D:
               Subcategory E through I:
 $6,756
$26,020
         BAT 1 is the proposed option for Subcategory K, but EPA 'currently estimates that there are no small
model facilities in the Subcategory.
                                               6-15

-------

-------

 I
c  -S
ll s

< "
so
 I




. o
•-V'l
:-.-••;'








'-

;-/5
; -0

?'• .'-

.'-. .- ;





: \.-:\
••-•- 1

f8
i;|
11
• C!

•--'•ง
O&MC
• . v.
O
'I
c
- "K
••U
Stf t
Q- ' '*•
"S *
11
• , c
11
's

• •••s
•-• .5
.-- 3*
:-o

•,.J2
'. 'O

"s
-. . c

• s
, - <
•<
1
O
3 ;SS- ซ

' (
en
en
o
S
in
rt-
so

*— i
m
se
o
in
VD
r-H
SO



vc
in
r-
so
c
oc
oc
se-

oo
en
in
so

oo
oo
vo
c
V

fy
PH
r-


oo
Os
OC3
SO
ON
ON
n
m
CN
VO

in
•n
m
cn
so
cn
30
VO
vo
\r
c-e-


VC
c-
in
C
SO
cn
00
oo
vo
en
vo
C
so
CN1
ON
O
en
vo
so

CN
en
If
en
VO
e^
se

CN
00
B
PH



8
ON
en
so
5
vo
SO

o
1
so
in
VO
rj^


%
in
r-
so
c
ON
oo
r-
so

VO
r-
ฃ
Cf
so

CN
ON

C
VO
en
se

ec
G



O
t-H
1
en
on
en
vo
so

00
vo
oo
•N
so
vo
O)
CN
rj^


oo
vo
oo
r-
r-
en
oc
ex
t--
-so

r-
ฃ
V

VO
VO
C""
ff
so

G
PH






























i
6*
1
-i
OO
n
8
en
VO
en
so

ป — i
0
ya
r-
vo



CN
VO
Tf
ON
SO
oo
V

en
en
VO
en
ON
SO

ON
c-
VO
s-

I



10
00
so
•a
m
se

o
en
cn
ON
se
in
ve
ON
S<=



en
en
r-
in
so
Vฃ
O
VO
SO

r-
oo
vo
en
so

ON
in
en
oo
in
o
o

o
G
PH



en
en
vo
so
en
0
ง
so

in
oo
ON
3N
SO
00
oc
vo
SO



e'-
en
vo
m
vo
so
r-
r-
c-
vc
vo
so

r
oc
e*-
so

i— i

C
in
vo
c

en
c/
0



o
o
en
oo
fir
VO
VO
ON
oo
so

in
oo
en
en
se-
ers
r-
oo
so



in
en
en
ON
ss
•n
r-
VO
ON
en
r-

en
ง
0
so

(->
ON
m
VO
OO
en
so

o
PC
V.






























1
j
s
H
-
•64
00


1
SO
S
vo
SO



CN
VC
CN
SO

o
en
so

oo
en
r-
so


en
(N
in
o


g
•*


n
vo
so
D
ฃ

ON
en
•ft
S
vo



•n
m
VO
CN
s*

r-
VO
oo
C-
so

en
en
vo
CN
SO


tf
m
CN
se


eN



•*
m
se
n
oo
m
m
so

ON
CN
en
vo
en
SO
••*
T— 1
O
in
00
se



oo
vc
so
ON
C
O
so

en
in
o
e<-
SO

O
r—
CN
oo
VO
se

en


































3
SO
O
oo
•60

VO
OO
OO
ป
CN
CN
OC



VO
oo
(N
ON
en
CN
se
ON
c-
VO
ON.
OO
0
r-
CN

vo
C
CN
OO
en
vc
ff*

VO
T— (
VO
en
f
&

V
0
oo
VO
in

i
oo
so
in
Y-H
VO
XI
SO

VO
oo
in
oo
en
en
vo
vc



ON
in
CN
ON
vo
ON
SO
r—
vo
in
vo
ee
o
so
C
o
CN
VO
VO
ftO

I—
•r~
r-
t—
vo
r-
sซ

CN
W
B
PH



O
3
ON
in
fir
OO
OO
VO
VO
SO

m
m
vo
ON
SO
vo
vo
ON
in
s=



oo
oo
en
oo
vo
CN
ON
SO
f-
cn
in
vo
VO
O

i
cn
f-
in
-so

S
-^*
CN
•^
r—
cn
so

cn
oo
g



in
vo
se
n
en
so

••*
S
en
in
vo
in
00
r-
so



in
oc
.ON
0
s<
ON
en
CN
c
vc
C
se
in
\r
f-
m
t-
in
so

in
r-
oo
in
vo
oo
in
se

0
a



;




















'ฃ

r

•o
VO

<
b


1
en
en'

ON
en
•*


S

in
V
cn



in.
C
ON
en
so

CN
en
V
so

r-
oo
CN
~


r-
00
*•—<
ON
cn
so

ff

1 — '

s
so
VO.
in
oo
so

1
oo
so
in
in
en
so



oo
ON
CN
CN
r—

^
VO
ON
ON
so

s
en
ON
m
ON


r-
oo
'~~'
ON
VO
en
so

CN
PC



g
ON
ON
vo
in
oo
in
SO;

ON'
VO
en
so
5
en
CN
so



en
en
oo
CN
in
r-
in
se
r-
en
in
CN
m
oc
VO
se

VO
oo
f-
t-
&•

^
r—
en
CN
SO

S


































3
CN
en
so
oo
vo
ON
en
so

vo
oo
en
CN
SO
CN
OO
CN
VO
m



in
VO
r-
-sf
ON
CN
ON
s<
en
oc
so
en
vo
o
in
c
&*
m
c-
> — '
oo
c-
f
so

c/
G
m
vo
m

CN
00
vo
fir
ON
n

oo
VO
CN
3N
SO
r-
VO
VO
VO
in
so



o
VO
00
VO
•vo
se-
en
in
ON
CN
ON
VC
se
en
oo
in
ON
CN
^
SO
ON
VO
V
oo
VO
oo
\f
so
CN
V
V.
PH



in
n
i — i
in
fir
OO
3
CN
VO
SO

in
oo
in
ON
SO
tj
S
SO



t~-
in
oc
c
ON
VO.
SO
VO
ON
ON
ON
en
SO
p-
oo
CN
O
c<
•SO
en
vj
r-
oo
cn
vo
ON
oc
o
so
en
oo
W
oo
PH



S
ON
VO
iO
b
o
oo
1 — 1
se-

en
so
in
in
in
in
r-
se



VO
T-H
in
OS
en
O
CN
r-
r-
SO
cn
m
en
CS
CN
CN
so
i
p~
00
\r
•so
^:
o
en
oo
CN
|
en
se
1




-------
00
•ป—H

VO

-------
Cs
^H

VO

-------

-------
               Subcategory J:
               Subcategory K:
               Subcategory L:
 $15,106
$215,386
$125,990
Estimated compliance costs for nonsmall model direct dischargers in the poultry subcategories are
significantly higher than for red meat and rendering subcategories. This may occur because red meat and
Tenderers are currently subject to effluent guidelines, but poultry establishments are not. No option is  •
proposed for nonsmall model indirect discharging facilities.
        6.4,1.2 Total and Average Compliance Costs by Meat Type and Process Class

        Small Model Facilities

        Table 6-6A presents estimated costs for small model facilities by meat type and process class.  The
 range of per facility costs within any given subcategory can cover a.wide variation among the meat type
 and process classes that compose that subcategory. For example, in Subcategory A through D, the average
 posttax cost per facility for BAT is $57,000; however, this reflects a range of per facility costs from
 $4,000 in the red meat first processing, further processing, and rendering class, to $119,000 in the red meat
 first processing class. The range of posttax annualized costs for small model facilities under the proposed
 option (BAT 1) within each subcategory is:
        •       Subcategory K:
        •       Subcategory L:
                — mixed first processing8

 No option is proposed for small model direct dischargers in subcategories A through J. No option is
 proposed for small model indirect dischargers in any subcategories.
      NA
     $711
         8 Throughout the remainder of this chapter, EPA will use the convention that if the results tor a single
 class are listed below a subcategory, then that is the only model size, class, and discharge type combination owned
 by small businesses in that subcategory.
                                                6-21

-------
s

-------
   *^j  w)




?==  i
1





ฃs
K
^


,









iJ
H
g














11
ซ "-3
p: ง
• a
a
11
•a

3
s.
O
ฃ
0
O
3
S
a
-11
S9
ฃ ง
a
,

1
^3


.K
'C
S

Q
Ca
'I
1
1
t
|
-=
"5

OS
^f



in
CN
i-H
69


S


p-
oo
69


^
1 — I
ft

O
o3



Os
VO
OS
69




CO
c<-
OC
€/=3
D
*—
E-
•4
PC
p-
-


g
in
69


vo
CO
"I
OS
69


T-H
co
00
69

O
CN
p~
oo
69


m
vo
od
Os
69

o
CN"
VD
69


CN
CN
P-
vo
•*
69




co
CO
CN
OO
^J-
69

CN
ฃ-
m




X
m



i — i
i — i
69

CO
P-
69

CN
VO
T— t
m
69


OS
CM
CM
en"
vo
69
l-H
OS
r-l
 OS
Os VO
t — VO
OO OS
oo •*
Os O
&* r4
69



n cs
vo O
-* m
CO CN
CO 00
vo C —
69 69
CO "*
on v:
W tฃ
en v
PH P-






















_>,
,
tq
|
I
o<
s


"S
"xi
s
a
&
K
Z
t
a.
!.
^
t
.=
^.
^
ซ
(^
m
vo
T— 4


OO
O
rh
69


P-
00
O
co
T 	 1
69

P-
T-H
OS
vo
69


CN
CO
OS
04
m
69

O
co
m
vo
69


OO
p-
oo
s




• CC
CO


&)
'
) V.
p-
: &
s a.
• OJ
ป 1-f
1
i
*
^*
p-
p-
co


o
Os
oo


OS
p-
o
o
l-H

m
p-
m


p~
OS
p-
1
69
P-
P-
VO
."
69

OS
m
CN
CO




CN
i — i
^!
r—
6"5
CN
V.
(L
•



O
ro
Os
CO
69

i— t
vo"
•51-
69

CO
O
VO
OS
69

O
00
in
69


P-
p-
i — i
co
69
P-
r-
p-
VT3

69

OS
CO
cn
p-
cn"
69



in
in
oo
T— 1
CN
in
i — i
69
CO
V,




OS
^
fo
 .if
g
^
^
•X4

CN
i — i
O
69


CO
OS
VO
>• vo"
69


^ 00
I 5
VO
69

r 5^
)• •*





i — i
p-
00
69


Ol
CO
T— I
1 — 1
69


m
oo
oo
i — i
i — i
69

S
T-H
CN
69


CO
00
Tt
in
co
•*
69

vo
vo
VO
O
P-
69


m
CN
ir
69




VO
OS
co
S
1—
^
1
V
p.
V



Os
p-
04
>n
fft


Os
CN
i — i
of
00
69


OS
cn
cn
cn
in
69

P-
cn
O
cn
p-
01
•69


OS
>n
m
OS
cn
VO
of
69
S
VD
O
•^j-
69

VD
CN
OS
VO
VD
VO
of
69



OO
01
OO
T-H
in
VD
cn
69
CN
V.
\ a
>
i



i-H
O
CN
in
fA


P~
o"
oo
69


00
in
69

cn
69


00
vo
Os
o
CM"
69
CN
VO
00
Os
cn
O
^>
69

in
oo
CM
CO
Os
in
of
69



O
VD
0
T-H
P-
co
f,
c<-
c/:
|
0-




co
co
OS
O
l-H
69

in
oo
co^
1-H
VD
T-H
69

P-
cn
01
OS
69

00
cn
o"
in
VD
•69


O
m
vo
vo
in"
69
OO
VO
CN
Os
VO
oo
69

OS
vo"
cn
VD_
69



CN
OS
1 — 1
in
04
69
OO
C/D




                                                                                                                                      en

                                                                                                                                      
-------
VO

-------

-------
       Nonsmall Model Facilities
       Table 6-6B provides costs for nonsmall model facilities owned by small businesses. Under the

proposed option (BAT 3 in all subcategories except J; BAT 2 in Subcategory J) for nonsmall model
facilities that are owned by small businesses, the range of posttax annualized costs per facility within each

subcategory is:
        .      Subcategory A through D:
               — red meat first processing

        .      Subcategory E through I:
               — red meat further processing:
               — mixed first processing:

        •       Subcategory J:
                — rendering

         •       Subcategory K:
                — poultry first and further processing:
                	poultry first processing, further processing, and rendering:

         •       Subcategory L:
                — mixed first processing:
                — poultry further processing:

 No option is proposed for nonsmall model indirect discharging facilities.
                                                                                            $6,756
 $26,020
  $5,985
 $91,709

 $15,106
$215,386
$174,281
$309,969

$125,990
  $91,709
$131,338
         6.4.2   Closure Impacts


         Facility level closure impacts are estimated using the site closure model described in Section 3.1.2

  and Appendix B.  The site closure model addresses the impact of compliance costs on the financial health
  of the individual facility.  In effect, the closure analysis estimates whether or not it makes economic sense

  for a facility to upgrade pollution controls, or if under these controls the facility would lose economic

  viability and therefore close.
                                                  6-26

-------

-------
3

-------
o

s

o

s


1
oo
CO
"!.
CO
VO
of
Ol
*— 1
0.
of
s
VO

ON
0
T— 1
ON
•60-
01
in
in
oc
oc
r-
V
,-
VI
1
3N
S
n
t-^
r-
co
ON
in
in
t—
f-
i — i
04
oo
ON
ป— 1
-S

-------
o
m

SO

-------
                                             t a cumulative probability function, the relative size
           combination class, and
a snbca^ory
                         and process
.4.2.2 does ซhe same by ป. ซype - process
                                          clas

                                                      w so wm foe taciemental probability of
                                                      ^

                                                          ofc


                                                                                  ill (OCBS on
       6.4.2.1



        Small Model Facilities
         in all snbcategories (te single excepuon
                                                6-31

-------
                 Table 6-7 A
Economic Closure Impacts: Small Model Faculties
           40 CFR 432 Subcategories
• Lmx. 	 "•"• ซp
Option
Subcategc
BAT1
BAT2
i"AT3




SES4

BAT1
BAT2
BATS

!SES1
SES2


Snbcateg
BAT1
BAT2
BATS

ISES1

SES3

Subcatei
PSES1

PSES3
JPSES4

of
Facilities
—•" " -1 .- 	 - - -L— ••^•i ปli^— •-—
Annualized
Compliance Costs
per Facility1
Pretax
Posttax
••••^•••gi^Mi^^^^^^^™ •*"^^^™^"*^™^^^^g
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flo\v
— g=p
Probability
Cash Flow
less Than .
Compliance
Costs3
Projected
Facility Impacts 4
Closures ]
>ry A through D 	 _ 	 —
59



1,001



orvEthrou
48



2,940



oryJ
6



17



wryK>
'39



' $494
$8,607
$72 828
$345
$5,739
$57,414
0.75%
15.82%
173.65%
0.63%
13.38%
147.23%
0.13%
2.74%
28.70%

$29,962
$162,234
$152,374
$172,616
ghl
$395
$5,955 1
$11,897
$24,298
$151,943
$141,591
$160,626
87.03%
544.23%
505.49%
569.76%
74.08%
463.24%
430.24%
484.88%

$332
$4,691
$9,586
1.12%
15.87%
32.44%
0.83%
11.67%
23.85%

$41,367
$148,447
$162,676
$180,014
$33,711
$137,169
$151,400
$168,731
114.05%
463.97%
512.14%
570.75%
83.86%
341.16%
376.57%
419.67%
15.97%
67.41%
67.01%
69.35%

0.14%
2.06%
4.39%

15.75%
57.02%
60.30%
63.18%

$0
$28,711
$295,816
$0
$22,510
$289,095
0.00%
159.92%
2053.90%
0.00%
56.44%
724.85%
0.00%
2.88%
34.00%

$47,547
$625,699
$446,441
$463,831
$41,033
$618,978
$439,720
$457,1 1C
291.52%
4397.57%
3124.02%
3247.57%
102.88%
1551.96%
1102.51%
1146.11%

$36,303
$154,481
$169,763
$189,66C
$31,268
$147,88]
$163,163
1 $183,06C
142.48%
1134.79%
1441.58%
> 1619.669!
80.33%
> 506.58%
, 611.01%
> 686.15%
5.26%
52.00%
45.22%
46.16%

17.64%
, 72.17%
, 72.22%
, 72.62%
0.1
1.7
17.0

160.0
674.8
670.8
694.1

0.1
1.0
2.1

463.2
1,676.6
1,773.1
1,857.8

0.0
0.2
2.0

0.9
8.8
7.7
7.8

> 6.9
, 28.2
•> 28.2
i 28.3
Employment

1
8
63

353
1,511
1,520
1,628|
8
0
2
4

979
3,545 1|
3,749
3,928

0
0
5

z
20
17
18

43
114~|
iTsl
Us]
                      6-32

-------
                                     Table 6-7 A (cont.)
                       Economic Closure Impacts: Small Model Facilities
                                  40 CFR 432 Subcategories
                                         Compliance Cost
                                         as a Percentage
                                        of Model Facility2
                                Probability
                                 Cash Flow
                                 Less Than
                                Compliance
                                    Costs3
                   Annualized
                Compliance Costs
                   per Facility *
                                     Projected
                                  Facility Impacts4
       Number
            of
      Facilities
                                                                         Closures I Employment
Option
 ubcategory L
 ATI
 AT2
 AT3
                                                                0.31%
                                                                2.71%
                                                               21.50%
              1.77%
             15.28%
             _^^_^^_^ซซi
            113.08%
              2.40%
             20.78%
            153.79%
              $711
             $6,139
            $45,447
  $846
 $7,770
$55,837
                                                                  37.49%
                                                                  67.47%
                                                                  66.48%
                                                                  67.92%
                                                   174.97%
                                                   683.34%
                                                   635.90%
                                                   704.63%
 418.51%
 ,^—^—^—^—
1597.97%
1486.97%
 $42,164
$170,856
$159,060
 $48,087
$178,615
$166,808
$184,357
 Total Excluding 65 Certainty Facilities
                                                                             683.8
                                                                            2,256.5
                                                                            2,861.1
                                                                            2,755.0
'SES1
'SES2
SES3
                                                                    umber of facilities in the
                                               6-33

-------
option 1 is less than 2.5 percent for all subcategories, although it becomes very high under option 3 (and

sometimes option 2) for all subcategories.


        Under the proposed option (BAT 1) for small model facilities in subcategories K and L, the ratio of

posttax compliance costs to net income, and the incremental probability of closure for each subcategory

are:
                Subcategory K:


                Subcategory L:
costs / net income:
probability of closure:

costs / net income:
probability of closure:
        NA
        NA

2.40 percent
0.31 percent
 EPA projects that no small direct discharging model facilities will close under the proposed option.  No
 option is proposed for small model direct dischargers in subcategories A through J. No option is proposed

 for small model indirect dischargers in any subcategories.
         Nonsmall Model Facilities


         Table 6-7B presents the closure analysis for nonsmall facilities by subcategory. Under the
 proposed option (BAT 3 in all subcategories except J; BAT 2 in Subcategory J) for nonsmall model
 facilities that are owned by small businesses, the ratio of posttax compliance costs, and the incremental

 probability of closure for each subcategory is:
                 Subcategory A through D:


                 Subcategory E through I:


                 Subcategory J:


                 Subcategory K:
 costs / net income:
 probability of closure:

 costs / net income:
 probability of closure:

 costs / net income:
 probability of closure:

 costs / net income:
 probability of closure:
 0.25 percent
 0.04 percent

 0.55 percent
 0.09 percent

 0.69 percent
 0.12 percent

 6.82 percent
 1.22 percent
                                                 6-34

-------
                                        Table 6-7B
          Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses
                                  40 CFR 432 Subcategories
                                         Compliance Cost
                                          as a Percentage
                                        of Model Facility2
   Annualized
Compliance Costs
  per Facility1
Probability
 Cash Flow
 Less Than
Compliance
  Costs3
   .Projected
Facility Impacts4
                                                                        ClosureslEmployment
                                      Net Income  Cash Flow
Subcatesory A through D
 ubcatego
  ATI
 Subcategory J
                                $15,106
                               $175,269
   ubcatesory K
                                               6-35

-------
                                          Table 6-7B (cont.)
           Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses
                                      40 CFR 432 Subcategories
                       ^^^^^^=^s^^^.
                        Annualized
                     Compliance Costs
                       per Facility
                                              ^---—•—=^^- ,
                                              Compliance Cost
                                               as "a Percentage
                                              of Model Facility-
Probability
 Cash Flow
 Less Than
Compliance
  Costs3
..'•',. Projected
Facility Impacts *
                                                                                  Closures
                                                                                            Employment
                                                                           0.00%
                                                                           0.08%
                                                                            0.89%
                                                                            1.26%
                                                                            1.45%
                                                                                        0.0
                                                                                        0.0
                                                                                        0.1
                                                                                        0.1
                                                                                        0.1
                                                                                                       16
                                 16
                                                                                                       16
0.30%
1.94%
1.41%
1.81%
0.4
3.1
2.3
3.0
70U
548 1
416|
5221
Total Excluding 65 Certainty Facilities
                                                                                                       size
|PSES4                     NA	NA|        JNA|         JN^.|          ""i      "-'        •"'—
AH impacts presented in this table are sum of the average of results for each subcategory, chscnarge type and model facility si
mmMnntion-weiehtedbv the number of facilities in each subcategory.               	_   _  .,.,„_
AH impacts presentea in uus iauic ui" ซ* "•- u..ซi-&~ -ซ	
combination, weighted by the number of facilities in each subcategory.               ^   ff  -,--   •  ,u / loco
'T^talannualized compliance costs for subcategory and discharge class divided by number of facilities m that class.
1 Ratio of posttax annualized compliance costs to net income and cash flow.
ป SS net income or cash flow less than posttax annualized compliance costs minus probability net income or cash flow

ซฐ QosTre^robability cash flow less than annualized compliance costs multiplied by the number of facilities in the
subcategory  Employment: employees per model facility multiplied by the number of projected closures.
'Option BAT 5 fs only found in Poultry operations. Subcategory L includes poultry further operations and mixed further
opVradons The count for BAT 5 is for poultry further operations only and hence, the number of facilities is smaller than for
other BAT options.
                                                    6-36

-------
               Subcategory L:
costs / net income:
probability of closure:
4.87 percent
0.89 percent
EPA projects that 0.4 nonsmall direct discharging model facilities will close under the proposed option,
with an associated employment loss of 107 workers. As would be expected, given the pattern of
compliance costs in Section 6.4.1, these impacts are projected among poultry processing establishments.
No option is proposed for nonsmall model indirect discharging facilities.
        6.4.2.2  Projected Closure Impacts by Meat Type and Process Class

        Small Model Facilities

        Table 6-8A provides closure impacts for small model facilities by meat type and process class.  In
 this particular case, the closure impacts at the meat type and process class mirror the pattern at the
 subcategory level.  Almost without exception, the ratio of compliance costs to net income for indirect
 dischargers .exceeds 100 percent under options PSES 2, 3, and 4. The ratio for most direct dischargers  is
 much smaller, but still substantial under options BAT 2 and 3.

         Under the proposed option (BAT 1) for small model facilities in the following subcategories, the
  range for the ratio of posttax compliance costs to net income within each subcategory is:
                 Subcategory K:
                 Subcategory L:
                 — mixed further processing
  costs / net income:
  costs / net income:
        - NA
  2.40 percent
  The incremental probability of closure due to the proposed rule is 0.31 percent in the mixed further
  processing class. No option is proposed for small model direct dischargers in subcategories A through J.
  No option is proposed for small model indirect dischargers in any subcategories.
                                                  6-37

-------
                 Table 6-8A
Economic Closure Impacts: Small Model Facilities
        Meat Type and Process Classes
1 1
Option
Red Meat
BAT1
BAT2
BATS

PSES1

PSES3
PSES4
Red Meat
BAT1
BAT2
•BATS

IPSESI
|pSES2

|PSES4
•MiiuiuiiigigiiiBigiiBBS *
Number •
of
Facilities
Annualized
Compliance Costs
perFacility1
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash FloW
^"T™^^*^^^Bg^^^=^= ™
Probability
:CashFlow
JLiCSS 1 nan
Compliance
Costs3
Projected
Facility, Impacts 4
Closures ]
Employment
Fir-ft Pracessine (Subcateeorv A-D) 	 , — _ 	 , 	
17



265



$0
$10,492
$128,400
$0
$8,225
$119,051
0.00%
29.68%
429.50%
0.00%
25.26%
365.64%

$25,331
$161,620
$150,996
$167,480
$20,652
$152,271
$141,647
$158,130
74.51%
549.35%
511.02%
0.00%
63.43%
467.67%
435.04%
0.00%
0.00%
5.13%
65.70%

13.49%
70.23%
69.28%
0.00%
0.0
0.9
11.2

35.8
186.1
183.6
0.0
0
2
24

77
403
397
oil
Further Processing (Subcategory E-I) 	 : 	 , 	 ,___ 	 1
43



2,489



$339
$5,731
$6,470
$285
$4,512
$5,158
0.96%
15.27%
17.46%
0.71%
11.23%
12.83%
012%
1.98%
2.27%

$40,967
$143,871
$162,635
$179,795
$33,411
$132,625
$151,388
$168,548
113.06%
448.80%
512.30%
570.37%
83.13%
330.00%
376.69%
419.39%
15.61%
56.12%
60.34%
63.19%
0.1
0.9
1.0

388.5
1,396.9
1,501.8
1,572.8
\Rpd Meat First and Further Processing (Subcategory A-D) 	 	 	 , 	
JPSESl

IPSES3
JPSES4 	 .
674



$33,490
$171,105
$158,480
$175,760
$27,320
$161,756
$149,131
$166,410
98.56%
583.57%
538.02%
600.36%
83.91%
496.80%
458.03%
511.10%
18.17%
70.81%
69.99%
71-03%
122.5
477.3
471.7
478.7
\Red Meat First Processing and Rendering (Subcategory A-D)
BAT1
BAT2
BATS

PSES1
PSES2
PSES3
IPSES4_
17



12



$1,215
$9,536
$114,841
$849
$5,792
$74,308
1.83%
12.50%
160.42%
1.54%
10.50%
134.65%

$0
$11,271
$138,106
$156,316
$0
$6,695
$90,043
$104,56';
0.00%
14.45%
194.39%
225.74%
0.00%
12.13%
. 163.17%
189.48%
0:31%
2.18%
31.70%

0.00%
2.53%
38.42%
44.22%
0.1
0.4
5.4

0.0
0.3
> 4.6
, 5.3
Oil
2H
2

'821
2,951
3,173
3,323

265
• 1,033
1,021
1,036

1
3
36

0
2
30
35|
                      6-38

-------
                                      Table 6-8A (cont.)
                        Economic Closure Impacts: Small Model Faculties
                                Meat Type and Process Classes
                                         Compliance Cost
                                          as a Percentage
                                        of Model Facility2
   Annualized
Compliance Costs
  per Facility1
                                                             Probability
                                                              Cash Flow
                                                              Less Than
                                                             Compliance
                                                                  Costs3
         Number
              of
        Facilities
                                                               Employment
Red Meat Further Processing and Renderin
                                                                  62.51%
                                                                 A-D)
                                                                   0.09%
                                                                   1.50%
                                                                   1.50%
                      and Rendering (Subcatego
Red Meat First Processing, Further Processi
                                          113.97%
                                          112.30%
                                                21.79%
                                                46.03%
                     $80,797
                    $161,385
 Poultry First Processing (Subcatego
                                                                   70.52%
                                                                   71.25%
                                   493.29%
                                   552.37%
   oultry Further Processins (Subcategory L)
                                                                   59.73%
                                                                   73.93%
             $48,389
            $179,208
            $166,752
            $184,331
  SES1
  SES2
  'SES3
  SES4
                                          2740.20%
                                          2549.75%
$182,331
$169,875
$187,454
 920.28%
1017.29%
  \Poultfy First and Further Processing (Subcategory K)
                                                                   73.27%
                                                                   73.93%
                                                                   73.93%
                       1348.41%
                       2002.76%
  $134,102
  $150,479
   'oultry Further Processing and Rendering (Subcategory L
                                              2.17%
                                                                     5.57%
0.0
0.2
0.2
0.2
o|
3|
3|
_3J
                                               6-39

-------
                                            Table 6-8A (cont.)
                           Economic Closure Impacts: Small Model Facilities
                                      Meat Type and Process Classes
Option

Number •
of
Facilities
Annualized '
Compliance Costs
per Facility1
Pretax
Posttax
Compliance Cost
as a Percentage ,
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
Mired Fnrthrr Prncessine (59 vercent Subcategory E-1,41 percent Subcategory L)
BAT1
BAT2
BATS

PSES1
PSES2
PSES3
PSES4
9



707



$846
$7,770
$55,837
$711
$6,139
$45,447
2.40%
20.78%
153.79%
1.77%
15.28%
• 113.08%
, 0.31%
2.71%
21.50%

$45,484
$175,729
$164,322
$181,785
$36,937
$164,483
$153,076
$170,539
124.99%
556.61%
518.01%
577.11%
91.91%
409.27%
380.89%
424.34%
17.33%
• 62.59%
60.66%
63.47%
0.0
0.2
1.9

122.6
442.5
428.8
448.7
0
0
4

259
935
906
948
\Mixed Further Processing and Rendering (59 percent Subcategory E-1,41 percent Subcategory L) ง
IPSES1
PSES2
P3ES3
PSES4
4



$19,860
$145,065
$139,317
$163,117
$12,687
$93,893
$90,534
$106,507
7.91%
58.57%
56.48%
66.44%
6.21%
45.94%
44.29%
52.11%
1.19%
9.28%
8.93%
10.60%
0.0
0.4
0.4
0.4
Oil
60
6JJ
6
Renderine (Suhcateeorv J) 	 	 . 	 ,- 	 , 	
BAT1
BAT2
BATS

PSESl
PSES2
PSES3
PSES4
6



17



$0
$28,711
$295,816
$0
$22,510
$289,095
0;00%
159.92%
2053.90%
0.00%
56.44%
724.85%
0.00%
2.88%
34.00%

$47,547
$625,699
$446,441
$463,831
$41,033
$618,978
$439,720
$457,110
291.52%
4397.57%
3124.02%
3247.57%
102.88%
. 1551.96%
1102.51%
1146.11%
5.26%
52.00%
45.22%
46.16%
0.0
0.2
2.0

0.9
8.8
7.7
7.8
0
0
5

2
20
17
18
Total Kxrludine 65 Certaintv Facilities _, 	
BAT1
BAT2
BATS

PSESl
PSES2
JPSES3

117



4,565



NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA

NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.2
3.0
21.9

843.8
. 2,771.3
, 2,857.1
L 2,786.3
1
10
74

1,796
5,966
6,165
6,065 1
f\li lllll/uvia UJLCoClliCU. ill "iio IMX/AV mw v*ปw ซ-.—*—.j-j—	       -       w*                     -
weighted by the number of facilities in each subcategory.
1 Total annualized compliance costs for subcategory and discharge class divided by number of facilities in that class.
2 Ratio of posttax annualized compliance costs to net income and cash flow.
3 Probability net income or cash flow less than posttax annualized compliance costs minus probability net income or cash flow
less than zero.
4 Closures: probability cash flow less than annualized compliance costs multiplied by the number of facilities in the
subcategory. Employment: employees per model facility multiplied by the number of projected closures.
                                                      6-40

-------
       Nonsmall Model Faculties
       Table 6-8B presents the closure analysis for nonsmall facilities by class.  Under the proposed

option (BAT 3 in all subcategories except J; BAT 2 in Subcategory J) for nonsmall model facilities that are

owned by small businesses, the range for the ratio of posttax compliance costs to net income within each
subcategory is:
               Subcategory A through D:
               — red meat first processing

               Subcategory E through I:
               — red meat further processing
               — mixed further processing

                Subcategory J:
                — rendering

                Subcategory K:
                                                     costs / net income:
costs / net income:
costs / net income:
 costs / net income:
                	poultry first and further processing
                — poultry first processing, further processing and rendering
                Subcategory L:
                — mixed further processing
                — poultry further processing
                                                      costs / net income:
                                  0.25 percent
0.55 percent
0.09 percent
2.03 percent

0.69 percent
 6.82 percent
 5.03 percent
 8.94 percent

 4.87 percent
 2.03 percent
 5.31 percent
  The largest incremental probability of closure occurs in the poultry first processing and rendering class:

  1.61 percent.  No option is proposed for nonsmall model indirect discharging facilities.
          6.4.3   Facility Nonclosure Impacts


          EPA estimated enclosure impacts for small business owned facilities affected by the proposed

  effluent guideline. These impacts include:


                 ratio of pretax annualized compliance costs to model facility revenues,

                 ratio of pretax annualized compliance costs to model facility EBIT,

                 ratio of posttax annualized compliance costs to model facility net income,

                                                  6-41

-------
                              Table 6-8B
Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses
                      Meat Type and Process Classes
r
Option
Red Meat
BAT1
BAT2
BATS
BAT4
RcdMea
BAT1
IBAT2
BATS
BAT4

PSES1

PSES3
PSES4
Poultry
BAT1
BAT2
BATS
BAT4
BATS

PSES1

PSES3
PSES4
Poultry
BAT1
BAT2
BATS
BAT4
BATS

PSES1


IPSES4

Number -
of
Facilities
•••••^••••••^^B*"^^"*''^^^^^^^^^^^*1*^^^^^^
Annualized
Compliance Costs
per Facility1
Pretax
Posttax
ggj^^sgs^^^= ^^gr^gg^
Compliance Cost
as a Percentage
of Model Facility 2
Net Income
;'•;.';,•';,; ,;<.<
Cash Flow
Probability
Cash Flow
Less .1 nan
Compliance
Costs3
Projected |
Facility Impacts 4 |
Closures
f First Processine (Subcatesorv A-D) 	 , 	 ,_
e



$0
$0
$11,374
$184,589
$0
$0
$6,756
$121,398
0.00%
0.00%
0.25%
4.50%
0.00%
0.00%
0.21%
3.74%
0.00%
0.00%
0.04%
0.77%
0.0
0.0
0.0
0.0
Further Processing (Subcatesorv E- 1) _! 	 	 , 	 r
8




132



First Proct
15





29



$0
$8 812
$9,683
$238 353
$0
$5,207
$5,985
$156,186
0.00%
0.08%
0.09%
2.48%
0.00%
0.07%
0.08%
2.07%

$73,445
$291,379
$278,156
$355,323
$46,494
$189,083
$181,172
$233,980
0.74%
3.00%
2.87%
3.71%
0.62%
2.50%
2.40%
3.10%
0.00%
0.01%
0.02%
0.40%

0.12%
0.49%
0.47%
0.60%
0.0
0.0
0.0
0.0

0.2
!_ 0.6
0.6
0.8
>ssinp (Subcatesorv K) 	 	 	 — _ — , 	 -,
$0
$27,256
$338,382
$438,186
$478,754
$0
$15,917
$222,567
$289,884
$317,915
0.00%
0.46%
6.42%
8.36%
9.17%
0.00%
0.35%
4.84%
6.30%
6.91%

$70,879
$778,694
$612,338
$644,743
$45,886
$508,810
$404,760
$427,592
1.32%
14.68%
11.68%
12.34%
1.00%
11.06%
8.79%
,9.29%
0.00%
0.08%
1.12%
1.47%
1.61%

0.23%
2.61%
2.06%
2.18%
0.0
0.0
0.2
0.2
0.2

0.1
0.8
0.6
0.6
Further Processine (Subcategory L) 	 	 	 , 	
10





123



$0
$19,361
$199,583
$265,637
$290,048
$c
$11,62C
$131,338
$175,902
$193,172
0.00%
0.47%
5.31%
7.11%
7.81%
0.00%
0.39%
4.43%
5.93%
6.52%

$70,838
$421,630
$300,77'7
$373, 97€
$45,63?
$273,64$
$198,40f
, $248,99!
) 1.92%
i 11.44%
> 8.34%
3 10.459?
1.629?
9.639?
7.029'
) 8.799
0.00%
0.08%
0.98%
1.32%
> 1.45%

3 0.359?
, 2.159?
3 1.569?
9 1.969?
0.0
0.0
0.1
0.1
0.1

) 0.4
> 2.6
3 l.S
•> 2.&
Employment

0
0
0
0

0
0
0




212
282

0
ฐ
??]
75
75

38
300
225
225

0
0
16
16
16

64
440
I ' 327

                                   6-42

-------
                            Table 6-8B (cont.)
Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses
                      Meat Type and Process Classes
Option
•:::- ;--:,,
Number '
of
Facilities
.Annualized
Compliance Costs
\per Facility *
Pretax
Posttax
Compliance Cost
as a Percentage
of Model Facility2
Net Income
Cash Flow
Probability
Cash Flow
Less Than
Compliance
Costs3
Projected
Facility Impacts 4
Closures
Employment
Prmltry Firxt and Further Processing (Subcategory K)
BAT1
BAT2
BAT3
I*"AT4
AT5

3ES1
3ES2
SES3
SES4
5




$0
$23,570
$266,052
$404,854
$447,263
$0
. $14,224
$174,281
$266,944
$296,461
0.00%
0.41%
5.03%
7.70%
8.55%
• 0.00%
0.31%
3.79%
5.80%
6.44%
0.00%
0.07%
0.88%
1.35%
1.50%

10



$6,180
$425,123
$405,896
$434,570
$3,609
$270,556
$266,675
$286,611
0.10%
7.81%
7.69%
8.27%
0.08%
5.88%
5.79%
6.23%
0.02%
1.37%
1.35%
1.45%
0.0
0.0
0.0
0.1
0.1

0.0
0.1
0.1
0.1
0
0
0
. 38
38

0
38
38
38
oultry First Processing and Rendering (Subcategory K)
ATI
AT2
AT3
BAT4
BATS

PSES1
PSES2
PSES3
I3ES4
6





2



$0
$30,172
$300,827
$386,186
$419,683
$0
$18,121
$200,158
$257,861
$282,176
0.00%
0.78%
8.61%
11.10%
12.14%
0.00%
0.66%
7.29%
9.40%
10.28%
0.00%
0.14%
1.61%
2.09%
2.29%

$11,662
$994,704
$563,692
$585,408
$6,927
$646,925
$375,887
$391,659
0.30%
27.84%
16.18%
16.86%
0.25%
23.57%
13.70%
14.27%
0.05%
5.40%
3.07%
3.20%
0.0
0.0
0.1
0.1
0.1

0.0
0.1
0.1
0.1
0
0
16
16
16

0
16
16
16
mltry Further Processing and Rendering (Subcategory L)
3ES1
3ES2
SES3
SES4
8



$22,134
$259,228
$219,656
$240,482
$13,843
$166,342
$142,691
$156,805
0.40%
4.80%
4.12%
4.52%
0.30%
3.61%
3.10%
3.41%
0.07%
0.84%
0.72%
0.79%
0.0
0.1
0.1
0.1
- 0
38
38
38
oultry First Processing. Further Processing, and Rendering (Subcategory K)
ATI
AT2
ATS
rAT4
BAT5

PSES1
PSES2
PSES3
pcpc/t
\9ฃฃ-L.
2





3



$0
$54,880
$471,217
$500,227
$547,238
$0
$32,050
$309,969
$330,176
$362,899
0.00%
0.92%
8.94%
9.53%
10.47%
0.00%
0.70%
6.73%
7.17%
7.88%
0.00%
0.16%
1.57%
1.68%
1.85%

$56,672
$992,301
$642,103
$663,330
$36,465
$637,387
$423,427
$438,841
1.05%
18.39%
12.22%
12.66%
0.79%
13.85%
9.20%
9.53%
0.18%
3.30%
2.16%
2.24%
0.0
0.0
0.0
0.0
0.0

0.0
0.1
0.1
0.1
0
0
0
.0
0

0
38
38
38
                                    6-43

-------
                                         Table 6-8B (cont.)
           Economic Closure Impacts: Nonsmall Model Facilities Owned by Small Businesses
                                   Meat Type and Process Classes
 Compliance Cost
 as a Percentage
of Model Facility
                      Annualized
                   Compliance Costs
                     per Facility1
Probability
 Cash Flow
 •-Less Than
Compliance
   Costs3
    Projected
Facility Impacts 4
                                                                                          Employment
All impac s presented in this table are sum of the average of results for each class, discharge type and model tacuuy size
•r Probability cash flow less than annualized compliance costs multiplied by the number of facilities in the subcategory.
Employment: employees per model facility multiplied by the number of projected closures.
* Option BAT 5 is only found in Poultry operations.
                                                   6-44

-------
        nltioofpos,B,annualizedcomplianceccstt,omodeHadliซyeashfloW



        numteoffacmtiesซxpecซa,



        5 and 10 percent of revenues, and
       .
        and 10 percent of cash flow.


impacts is described in Section 3.1.3.
     6.4.3.1 Nonclosure Impacts by Subcategory
     SmaU Model Faculties
  exceed that threshold9 do exceed that threshold.
       ;^^I^,ซp^^^--^-to--11^<"'ฐ'
  the impact analysis.

                               6-45

-------
cilities
 ll
  o o
 S 8s
3=3 ง
2 s-5
o
3

S
Wฃ

55 cs

-------
r

-------
       Under the proposed option (BAT 1) for small model facilities in subcategories K and L, the ratio of
pretax compliance costs to revenues, and the number of establishments incurring costs exceeding 1 percent
of revenues and 3 percent of revenues are:
               Subcategory K:
               Subcategory L:
costs / revenues:
exceeding 1 percent:
exceeding 3 percent:
costs / revenues:
exceeding 1 percent:
exceeding 3 percent:
        NA
        NA
        NA
0.20 percent
0.2 facilities
0.1 facilities
EPA projects that about 0.2 small direct discharging model facilities will incur costs exceeding 1 percent of
revenues under the proposed option. Also note that the ratio of posttax compliance costs to cash flow is
 1.77 percent for small direct dischargers .in Subcategory L. No option is proposed for small model direct
dischargers in subcategories A through J.  No option is proposed for small model indirect dischargers !-
any subcategories.
                                      rs in
        Nonsmall Model Facilities

        Table 6-9B presents a summary of nonclosure impacts for nonsmall model facilities by
 Subcategory, discharge type, and technology option. For nonsmall model facilities, the impacts in terms of
 the ratio of costs to revenues and cash flow are relatively much smaller than impacts to small model
 facilities for any given option in any given subcategory.  In only one case, (Subcategory J, PSES 4) do
 average compliance costs exceed 2.5 percent of model facility average revenues, or 10 percent of model
 facility average cash flow (Subcategory K, PSES 2). To the extent that impacts under the proposed option
 for nonsmall model facilities exceed impacts to small model facilities, it is'because a higher option is
 proposed for nonsmall model facilities.

         Under the proposed options (BAT 2 for Subcategory J; BAT 3 for all other subcategories) for
 nonsmall model facilities, the ratio of pretax compliance costs to revenues, and the number of
 establishments incurring costs exceeding 1 percent of revenues and 3 percent of revenues is:
                                                6-48

-------

-------

-------

-------
               Subcategory A through D:
               Subcategory E through I:
               Subcategory J:
               Subcategory K:
               Subcategory L:
costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:

costs / revenues:
exceeding 1 percent:
exceeding 3 percent:
0.02 percent
0.0 facilities
0.0 facilities

0.07 percent
0.2 facilities
0.1 facilities

0.17 percent
0.5 facilities
0.1 facilities

0.58 percent
5.9 facilities
1.2 facilities

0.55 percent
2.2 facilities
0.4 facilities
FPA projects that about nine nonsmall direct discharging model facilities will incur costs exceeding 1
percent of revenues under the proposed option. No option is proposed for nonsmall model indirect

discharging facilities.
        6.4.3.2 Nondosure Impacts by Meat Type and Process Class


        Small Model Facilities


        Table 6-10A presents nonclosure impacts for small model facilities by meat type and process class.
 Under the proposed option (BAT 1) for small model facilities in subcategories K and L, the range for the

 ratio of pretax compliance costs to revenues within each subcategory is:
                Subcategory K:

                Subcategory L:
                — mixed further processing
 costs / revenues:

 costs / revenues:
          NA

  0.20 percent
                                                6-52

-------
CO
in

vo

-------

-------
r

-------
VO

-------
No option is proposed for small model dkect dischargers in subcategories A through" J. No option is

proposed for small model indirect dischargers in any subcategories.
        Nonsmall Model Facilities


        Table 6-10B presents nonclosure impacts for nonsmall model facilities by meat type and process
 class.  Under the proposed options (BAT 2 for Subcategory J; BAT 3 for all other subcategories) for

 nonsmall model facilities, the range for the ratio of pretax compliance costs to revenues is:
                Subcategory A through D:
               — red meat first processing

                Subcategory E through I:
                — red meat further processing
                -  ซnixed further processing

                Subcategory J:
                — rendering
costs / revenues:
costs / revenues:
costs / revenues:
                Subcategory K:                       costs / revenues:
                — poultry first and further processing
                — poultry first processing and rendering
                Subcategory L:
                — mixed further processing
                — poultry further processing
                                                      costs / revenues:
                                                                                      0.02 percent
0.07 percent
0.01 percent
0.27 percent

0.17 percent
                                  0.58 percent
                                  0.37 percent
                                  1.00 percent

                                  0.55 percent
                                  0-27 percent
                                  0.59 percent
  No option is proposed for nonsmaU model indirect discharging facilities.
  6.5    REGULATORY FLEXIBILITY ANALYSIS


         Based on the results presented in Tables 6-5 through 6-10, EPA has chosen to minimize economic
  impacts to small business establishments in the meat products industry by tailoring its proposed guidelines

  to differences in subcategory, discharge type, and facility size. Specifically, EPA is:
                                                 6-57

-------
r-
d
'"H
(N
d
i — i
ON
r-H
rH
VO
d
o
r-^
>n
^
d
!>
en
*— <
i — i
cs
in
d
'OO
in

-------

-------
1
   o
   2
 s
 ง
I
m
jliance Costs
)f Cash Flow
3 i
curring Co
iPercentag
s m
Facilities]
Greater Th
fi
J C
Compliance
itageofReve
ies Incurring
r Than Percei
11
Percent |
ace Cost as a
,!odelFaci!it
r
|
1


10 Percen
.M
1
•ti
a
1
to
10 Percent
5 Percent
3 Percent
1
Cash Flow
1
Revenues
Of
Faculties
c
"e
o
o
o
o
o
0
o
o
o
o
o
0
o
o
c
VO
0
o
c
o
CO
ty
(I
r/i
0)

m



CO
o
in
0
0
OS
f
CO
oo
OS
o
c<
CO

ty;
r/

O
ro

in
1—4
ป— t
O
CO
O
1*
o
r-
0
V
0-
oo
oc

CO
ty

o
CO

in


cs

0
o
o
CO
o
c
c
1
o
oo
cy

o
VO

-^
CO
o
o
ฐ
oo
o
s
VC
CO
c
v<
o

CO
C/3
n
3
o
CO
1 — 1
CO
CO

o
0
VO
o
c
r
c
o

CO
ty

o
in

VO
CO
o
o
^
CO
o
o
CO
c-
o
CO
o

c








. — _
e (Subcategor
andRenderin
'Processing,
i
.:
!
c
a,

o
o
o
o
0
o
o
o
o
o
0
o
o
c
o
0
o
c
o
CO
i
3
O
i— i
O
,_,
o
o
o
0
o
o
o
o
o
o
o
\r
O
o
O

a
t
0
oo
o
rO

O
O
o
o
V
0
CO
C-
VO
\r
VO
VO
0

co
E-
pc

o.
oo
o
CO


*^
O
V
O
r-
r-
r
vc
O
r-
o

f
pc

o
Os
O
Tt-

o
^•^
o
VO
O
oo
oo
r-
VO
m
r-
o

in
E-
pc
















o
^H
o
CO
o
o
^"^
o
c>
o
o
OS
O
c-
V
o
o
O
CO
ty
p


— i
CO
CO
CO

^""^
o
o
in
oc
C'
t~-
Os
c

a
c
pt,
30
o
VO
T-H
CO
CO
o
^" "^
0
c-
c-
^
co
ON
n
o

5
o
CO
O
0
VO
V
r-
c-
o

pc
"-J
o
oo
O
CO

O
CO
o
co
o
o
VC
c
c
o

pq

















oo
CO
C--

o

OS
o
o
CO
0
o
o
CO
oc
o
0
O!
ty
ฃ
c

CO
1— i
~*
in
CO
O

CO
CO
>n
OS
f
co
f
c-
vo
c

CO
t/
(I
c

oo.
VO

m


in
CO
vc
O
o
r-
!
co
oc
O

CO
cc
[I
ty;

CO
^

CO

CO
CO
m
OS
vq
>r
co
CO
r-;
co
OS
VO
o
o

c
PH

-------
r

-------
               not proposing new effluent limitations and guidelines for indirect dischargers in any
               subcategory;
               proposing to exclude small producers (i.e., small model facilities) from revisions to
               effluent guidelines for subcategories A through D, E through I, and J (red meat and
               rendering subcategories);
               proposing to set less stringent guidelines (BAT 1 instead of BAT 3) for small producers
               than for nonsmall producers in subcategories K and L (poultry subcategories).

EPA presents its estimate of the number and model size of small business owned facilities that will be
affected by the proposed rule in Tables 6-11 and 6-12.  Table 6-11 presents the estimates by subcategory;
Table 6-12 presents them by meat type and process class.

        By not proposing new guidelines for indirect dischargers, EPA excludes 96 percent of all small
business entities (4,990 out of 5,174 small business owned facilities) in the meat products industry from
additional regulatory burden. By excluding low production volume facilities in subcategories A through D,
E though I, and J,  112 of 140 small business entities in the red meat and rendering subcategories incur no
new costs under the proposed rule. Finally, by proposing a lower option — based on current performance
— for low production facilities in subcategories K and L, EPA minimizes potential regulatory costs to
those 72 affected small business establishments.  Thus, EPA anticipates that 1.4 percent (72 of 5,174) of
small business owned facilities in the meat products industry will incur costs under the proposed rule.

         Table 6-13 summarizes projected impacts to 71 small business owned meat products facilities that
 are expected to incur compliance costs.10 The four small model faculties are expected to incur total posttax
 annualized compliance costs of $2,600, about $700 per facility.  Average projected costs exceed 0.2
 percent of model facility revenues; about two of these facilities are projected to incur costs statistically
 exceeding 1 percent of revenues.

         For the 67 nonsmall model facilities owned by small businesses, posttax annualized compliance
 costs total $8.0 million,  about $119,000 per facility. However, the overall average is somewhat
 misleading. Twenty-seven establishments in subcategories A through J are projected to incur about
         10 Small differences appear in facility counts due to rounding (e.g., Table 6-11 shows 72 affected small
  business establishments, Table 6-13 shows 71).
                                                6-62

-------
                                                                 Facilities
                                                              Affected SmaU
                                                                    Busine|
                                                                    Ownfid*
                                                          SmaU Business
                                                                Owned*
      Facility Size
Subcategory A throughฎ
 mall
  edium
 subcateg
  mall
     •*
   edium
  ubcategQ
     Small Total
  Medium Total
          Total
Very Large Total
        TOTAL
    not sum due to rounding.
                                               d SBA s eciai Tabulations
                                                               '
                                                6-63

-------
                             Table 6-12
     Meat Product Industry Estimated Direct and Indirect Discharge
Affected Small Business Owned Facilities by Meat Type and Process Classes
^P^amummiiii,,!, 	 HITT nBMlllllff?^M!IV 	 '•••' mh 1 	 n 	 ^-ISI^f^S^Sa^m
Model FacUitv Size
i-.ii 	 	 • -a^ ^^^^^S—Bg^ggggT"1"1!!! '?"
Direct Discharge Facilities
•'• ': ':••- :": "'-"'' •
Small Business
Owned*
Affected Small
' Business
Owned*
Indirect Discharge Facilities
Small Business
Owned*
Affected Small
Business
Red Meat First Processing (Subcategory A-D) 	
Small
Medium
17
• 5
0
5
Red Meat Further Pry>ซ?.ซzn? (Subcategory E- 1) 	
Small
Medium
43
8
0
8
Red Meat First and Further Rendering (Subcategory A - D) 	 	
_ ,, 1 o
0
Red Meat First Pmcessing and Rendering (Subcategory A -D)
Small
17
0
265
0

2,489
132

674

12-
Red Meat Further Processing and Rendering (Subcategory E- 1) — ,
Small

Small
0
0
32
Further Processing, and Rendering Subcategory A-D)
25
0
Poultry First Processing (Subcategory K) 	 .
(5mall
Medium
	 —
Poultry Further Processing
Small
Medium
Large 	 .
0
15
0
15
50

19
29
(Subcategory L) , 	 .
0
9
1
0
9

272
119
4
0
0




0

0





0
0

0
0
0
Poultry First and Further Processing (Subcategory K) 	 	 	 	 , 	 . 	 fl
Small
Medium
0
5
C
5
20
10
Poultry First Prr>r.p..iaing and Rendering (Subcategory K) 	 	
Medium
6
e
, 2
OH
	 ; 	 : 	 II
o|
— 1
0
Poultry Further Processing and Rendering (Subcategory L) 	 ' 	
Small
Medium
(
C
c
(
I • -A
I ฃ
1- 0
\ 0
|p_..r,n, First prp^ecmp Further Processing, and Rendering (Subcategory K) 	 1
Si™'
/-
^
3 OJ
                                  6-64

-------
                                     Table 6-12 (cont.)

                 Meat Product Industry Estimated Direct and .Indirect Discharge

            Affected Small Business Owned Facilities by Meat Type and Process Classes
                                                            Indirect Discharge Facilities
                               ct Discharge Facilities
                                                                            Affected Small

                                                                                 Business

                                                                                 Owned*
                                         Affected Small
                                                          Small Business

                                                               Owned*
Small Business
Model Facili
                                    E-1 and Subcategory L) [
ixed Further Processing (Subcatego
                                                 E-1 and Subcate
  ixed further Processing andRenderin

  mall	

                     J
enderer (Subcatego
             Medium Total
          Very Large Total
 *Numbers may not sum due to rounding.

 rF^f^
 facilities the allocation is 61% in Subcategory E through I and 39% in Subcategory L.
 Basedon Screener Survey, Census Model Facilities, and SB A Special Tabulate,
 Classes with zero facilities were excluded from the table.     •
 EPA didTnot Sribute the 65 certainty facilities between direct and indirect dischargers.
                                             6-65

-------
vo

-------
$18 000 in compliance costs per facility, while the remaining 40 facilities in the poultry subcategones (K  .
and L) incur an average of about $187,000 in costs. This disparity fes presumably because there are
      tly no guidelines for poultry processors. Even in subcategories K and L, average compliance costs
    pose less than 0.6 percent of facility revenues, and about 9 of the 67 potentially affected small
businesses are statistically projected to incur costs exceeding 1 percent of revenues.
curren
com
 6.6     REFERENCES
        Protection Agency, Office of Water.
            1998  Statistics of U S  Businesses: Firm Size Data: U.S. Data: Classified by employment; size
           ^iSSSTsS -  1998 all industries data." U.S. Small Business Administrate, Office
         uf Advocicy  V-iilnM- ซ- ^•//www.sha.gov/advo/stats/data.html.
 U.S.SBA- 2000. Small Business Size Standards Matched to                      s
         Svstem (NAICS) Codes. U.S. Small Business Administration, Office of Size Standards.
         Available at: http-//www.sba anv/size/indextableofsize.html. December.
                                                 6-67

-------

-------
                                       CHAPTER?
                           ENVIRONMENTAL BENEFITS
7.1    BENEFIT VALUATION METHODOLOGY

       The proposed meat products industry effluent limitations guideline will reduce emissions into the
waters of the United State, The reduction in emissions will reduce the levels of fecal coliform and
biological oxygen demand and improve other indicators of water quality.  As water quality improves waters
may become suitable for increasingly demanding human uses. A primary benefit of the regulation is the
restoration of waters to conditions conducive to fishing and swimming.

        Eachusecategorycanbedefmedintermsofasetofwaterqualityindicators. If the indicators
 meet or exceed all of the criteria for a given use, then the water body can be used for that use. Vaughan
 (1986) developed a water quality criteria ladder which describes the type of recreational use that a water
 body can support (none, boating, fishing, or swimming). For example, a water body with a biologlcal
 oxygen demand (BOD) between 3 and 4 mg/1 is suitable for boating and fishing but not for swimmmg. All
 of the indicators must achieve the prescribed level for the water body to support a given level of use. Thus,
 if a water body had BOD between 3 and 4 mg/1, but a fecal coliform count greater than 2,000 per 100 ml,
  it would be classified as not boatable because of the high coliform count. The overall use category „ the
  least demanding use supported by any of the water quality indicators.

          Once the use of the water body is defined by the Vaughan ladder, the public willingness to pay for
  changes in use category can be estimated. Mitchell and Carson (1986) conducted a national contingent
  valuation survey which sought households' willingness to pay for improvements in the quahty of the
  nation's waters in terms of a use ladder. This survey characterized households' annual willingness to pay
   for improvements in freshwater resources from their baseline conditions to fishable and swimmable
   conditions. The survey sought to estimate the value of discrete changes from one use category to another
   corresponding to the Vaughan water quality ladder.
                                                 7-1

-------
database of nvers and streams.
                    andCa.son s^dy is
due to the regulation.
 wheป U .ซ. I. a change tan on. use category
 toit
  are ascribed ,o i,. ThU
  711  A Continuous Approach to Valuation
   '
   index as an in
                                                                                     on
                                                              households elseปh^ EPA

                                                                          ^^

                                               7-2

-------
Quality Index (WQD combines information ten four water quality measures rather than using only the
Ming lowest quality criterion to define use category.  For this benefit valuation, EPA used NWPCAM to
compile a WQI ten turbidity, BOD, fecal conforms, and dissolved oxygen indexes; this WQI is based on
work by McClelland (1974). Vaughn's breakpoints on the water quality ladder can be translated mto the
WQI as shown in Table 7-1. However, the translation results in almost all reaches falling into the top use
category in the baseline, tha, is, their WQI was greater than 76.19. This demonstrates the difference
between applying a limiting qualuy rule among four criteria and using a single aggregated measure.  Some
 criteria are apparently more difficult to achieve than others. Merely achieving the WQI represented by the
 values in the Vaughan criteria misses the fact that any one criteria that is no, satisfied can reduce the use
 ,evel. An alternative mapping from WQI to me Mitchell and Carson WTP values  is necessary for the
 results to be comparable with prior benefit valuations.

         Since the baseline distribution of use categories is well understood and generally accepted, it is
  desirable  for the distribution based on WQI to match the existing distribution of use categories m the
  baseline  EPA derived WQI values to represent the breakpoints on the water quality ladder based on
  empirical observation of the WQI distribution among us* categories in the baseline data.  EPA calculated
  the mean and standard deviation of WQIs for the reaches in each use category in the baseline populatton of
  reaches.  If reaches are normauy distributed within each use category, 84 percent of observed WQI for each
  category should be less than the mean WQI plus one standard deviation (SD).  The Mean + SD value
  serves as the criterion for the boundary with the next higher use category.  Table 7-2 shows the calculation
  and the resulting criteria.

           Table 7-3 shows how applying this set of criteria to the baseline NWPCAM data predicts baseline
   use category.  The first column indicates the use category using the standard most restrictive cntenon
   method The second column indicates the distribution of use categories assigned using the Mean + SD
   criteria  given the baseline use category. Shaded rows indicate agreement between both methods.  Srxty-
   four percent of reaches fall into the same use category using this method as in the most restricts use
   method (- 19 0 + 7.4 + 14.9 + 22.4). About 88 percent of reaches fall into use  categories the same or
   lower than their category in the baseline. Clearly, the two methods frequently agree and, exceptforthe
   lowest  category, the Mean + SD criteria usually places the reach in a lower category.
                                                  7-3

-------
                                           Table 7-1
                       Applying WQI to Vaughn's Use Category Criteria
Characteristic
Fecal Coliforms
Dissolved Oxygen
BOD - Max -day
Turbidity
Measure
#/100ml
percent
mg/1
JTU
Weight
0.314
0.333
0.216
0.137
No Use to Bbatabfe
Criteria
2000
45
4
100
Weighted
2.388
3.267
2.376
1.474
Beatable to Fishable
Criteria
1000
51
3
50
Weighted
2.562
3.526
2.534
1.646
Fishableto
Criteria
200
83
1.5
10
Weighted
3.559
4.475
2.643
1.810

Product/Implied WQI

27.337

37.668

76.190
Source: Weights: Bondelied, 2001; Values: Vaughan, 1986; Values were scaled by eye from graphs in McClelland,
1974, Appendix A.
                                               7-4

-------
[
                                                  Table 7-2
                          Empirical Calculation of Criteria from the Baseline Scenario
                                                                                          Rate, R
                                                                                       ($/WQI, 1999)
                                                            WTT va • fom HVA. .001,
       Source: EPA analysis of Baseline
                                                  U.S. freshwater bodies from baseline quality to the next
                                                                         waters to use category 3,
                                                         7-5

-------
                                          Table 7-3
           Comparison of Baseline Scenario Categorization under Most Restrictive Use
                                   and Mean + SD criteria
Use Category
by Most
Restrictive Use
0
0
0
1
1
1
2
2
2
2
3
3
3
3
Use Category
by -•• •".'.'--
Mean + SD
0
1
2
0
1
2
0
1
2
3
0
1
2
3
- .-,. ---. ' • .. • "•-.; .; : ' •/,;
:'••; \:-^-' •'•.-• '•• ::Vr;v
, .Number, of
Reaches in Category
125,727
49,110
758
8,161
49,107
12,416
5,468
89,383
98,320
16,031
103
6,759
50,942
147,994
: : Percent of '
yibst Restrictive Use
r-."r ";>'"•" -; 'Category
71.6%
28.0%
0.4%
11.7%
70.5%
17.8%
2.6%
42.7%
47.0%
7.7%
0.1%
3.3%
24.8%
71.9%
,.".'". '' - " -' ' •
• • ' ~ • " • "
.'-:- -: .;..-:• j. }:*<•:•'••.
Percent of All
: Reaches
19.0%
7.4%
0.1%
1.2%
7.4%
1.9%
. 0.8%
13.5%
14.9%
2.4%
• 0.0%
1.0%
7.7%
22.4%
Source: EPA Analysis of Baseline Access database, 10/2/2001
                                              7-6

-------
       The Mitchell and Carson willingness to pay values were updated to 1999 values for the recent
Concentrated Animal Feeding Operations (CAFOs) regulation benefit assessment to account for changes in
income and the value of the dollar. The CAFOs assessment, however, valued only changes in use
categories. The continuous WQI method requires that the Mitchell and Carson willingness to pay values be
converted to continuous measures of benefits. This rate of change for each use category is calculated so
that the total willingness to pay at each breakpoint is equal to the total in the Mitchell and Carson benefit
ladder (as adapted to 1999 values for the CAFOs benefits assessment). The resulting rates are shown m  ^
column 5 of Table 7-2.  The not boatable category is arbitrarily spread over the whole range from 0 to 79.'
No value is associated with improvements above the swimmable level, which is a very small range.  The .
 result is a linear approximation of an increasing marginal benefit curve, f(W0, W,), as shown in Figure 7-1.
• With each step, the rate of increase in benefits is roughly four times higher than the previous step.  As the
 rate of increase in willingness to pay per household increases with use category, the tendency of the WQI
 mean + SD breakpoints to categorize reaches lower than they would have been under the most restncttve
 use criterion  will cause the  benefits to be conservatively valued. However, a method which values any
 change in WQI will most likely generate higher values than a method which only  includes changes in use
 categories.                       .

         EPA used the NWPCAM model to estimate changes .in water quality indicators. NWPCAM
  produces a Microsoft Access database for the baseline  conditions and each regulatory scenario. Each
  database is then processed to generate weighted estimates of household willingness to pay. For each reach,
  the model calculates the household willingness to pay for a national change in water quality between the
  reach's baseline WQI (W0) and its WQI in the regulatory  scenario (W,) and scales it by the length of the
  reach, k;.
                                   B,, = ks [f (Wu) - f(Woi)]                                  CD
   where- f(W) is the average household benefit of a change in water quality from W0 to W, at the national
   level and k is the length of reach i. This yields a mileage  weighted benefit measure, Bni, for each reach, i, m
   each state, n.
           ' Mitchell and Carson described non-boatable waters in graphic terms so their value for the changeonay
   be an overestimate. However, few water bodies approach a zero WQI, so much less than the full value for the
   imi
iprovement to boatable can ever be attributed to the regulation.
                                             7-7

-------
                       Figure 7-1
    Cumulative Willingness to Pay for Changes in WQI, f(W)
          Cumulative WTP for WQI Changes
700
                                                       100
                           7-8

-------
       Waters closer to one's home are easier to access and use, so it might be expected to command a
higher value. Mitchell and Carson asked respondents to apportion their willingness to pay between
proving lซa, waters, i.e. in-state, and proving more distant waters. On average, respondents aUocated
two-thirds of ft* WTP to in,,a,e waters. So, benefits are calculated on a state-by-state basis in terms of
benefits to me stag's households frora in-state and out-of-state improvements. For to-state benefits, S,, the
mileage weighted vaiue is divided by the total stream miles within the sMe, L., and multiplied by two-
thirds to essence, the WTP value is weighted by the proportion of in-state waterways affected and me
 proportion of the total household value for in-state water qualify improvements.  This quantify multrpaed
 by the number of households' in the state, H., yields the value of the fa**, changes in water qnahty to
.state households.
                                             0.67
                                                         B,,,
                                                                                              (2)
         Households in every state also value the improvement in water quality in other states.  T*e sum of
  WTP weighted by mileage for states other.than the home state is divided by the sum of reach mileagemall
  other states, L,.' One third of this sum multiplied by the number of households ta the state yields the
  willingness of one state's households to pay for improvements in distant states.
                                                 0-33    B
                                                            _ni
                                                                                               (3)
                            ;-of-state values is the total willingness to pay of all households within the state
The sum of in-state and out-
for the water quality improvements of the scenario. The sum
                                                            of state values is the national benefit estimate.
    "      - -    ..    • i_ *  1 nnn ,,ซ^^nc- 1OQ8
 UCll.wJ.AA-i^'*-1-* * ป" ซ•"	  t
 households nationwide in 1999 versus 1998.
    double counting.
                                                    7-9

-------
       7.1.2   Use Category Approach to Valuation  =

       As a comparison, EPA also estimated the benefits of the proposed regulation using the change in
use category method as in previous benefits assessments.  The 4 use categories (none, beatable, fishable,
and swimmable) were labeled from 0 to 3. There are 6 possible positive changes in use categories.
Changes in category from a more demanding use to a less demanding one are possible but were ignored in
this estimate. Table 7-4 shows the possible changes and the annual WTP values per household ascribed to
each change in national water quality from the Mitchell and Carson WTP values as updated to 1999
values. Larger changes are valued more highly.

       Each reach in the database was placed in one of these categories of use change or a no change
category.  The assumption that two-thirds of value applies in-state and one-third applies out of state is
maintained.  So two-thirds of the household's value would have been achieved if all of the state's
waterways made the identified change. As only k^ miles are estimated to make change, j, the total length in
each category in state, n,  is divided by the total length of rivers in the state, Ln, to weight the WTP value.
                                    Sn  =
                                      n
(4)
        Out of state values are estimated similarly with all of the out of state mileage in each category
weighted by the total out of state mileage, L_n.
                                                                                                (5)
 As in the continuous method, state values are summed to yield national benefit estimates.
                                                7-10

-------
              Table 7-4
WTP Values for Changes in Use Category
      No Use to
     Swimmable
Boatable to
 Fishable
Boatable to
Swimmable
Fishable to
Swimmable
                     7-11

-------
7.2   .BENEFIT VALUATION RESULTS

       Benefits of the proposed regulation are modeled based on 97 (36 direct dischargers) meat
processing plants for which data were available nationwide. These plants provided a sample set of impacts
for evaluation.  The mileage affected by the changes is small. The most effective scenarios result in net
upgrades in use categories on less than 45 river miles. Table 7-5 shows the number of river miles that
change use category in each scenario. Many of these changes occur in states with relatively small
populations, e.g., Nebraska, so the benefits generated from in-state improvements are also small.  Table 7-6
summarizes the valuation results by scenario and compares the continuous WQI method of assessing
benefits with the change in use category method used in CAFOs. The continuous method generates a
higher estimate of the dollar value of benefits. However, counting from lowest to highest benefit values, the
two methods place the scenarios in essentially the same order. This indicates that the change in category
approach may have been capturing the significant effects of the water quality change on a national basis
though perhaps missing detail at the state level.

        Tables 7-7 through 7-10 show the state level changes and values. Table 7-7 shows the mileage
 that changes from one use category to another by state as well as the number of households and number of
 households per river mile. Waters in only 6 states change use categories. The Mitchell and Carson WTP
 results place a premium on in-state waters. Both methodological approaches generate higher benefit values
 for states with greater population per river mile. Arkansas, Iowa, and Nebraska are geographically large
 states with small populations so they generate fewer benefits per river mile improved. On the other hand,
 Maryland is a small state with a large population and so generates disproportionately high benefit totals.
 Improvements in Wisconsin water quality affect less mileage but result in use categories increasing more
 than one step. One reach in Wisconsin increases from no use to swimmable.

         Table 7-8 indicates which states will experience the largest changes in WQI under the proposed
 Scenario 7. Wisconsin, Iowa, Illinois, and Minnesota show large total mileage changes in WQI indicating
 large changes in the water quality of many water bodies. Wisconsin, Texas, and Minnesota have large
 average changes in WQI. Reaches in these states will be improved to a greater extent than reaches which
 will be improved in other states.  Note that while the WQI scale ranges from 0 to 100, it is not a ratio scale
  so an average change of 14 cannot  be interpreted as a 14 percent change.  Nevertheless a 14 point change is
                                                7-12

-------
                                        Table 7-5
            Reach Use Category Changes from Alternative Scenarios (97 Facilities)
                                      (Reach Miles)
Scenario
-i i   !•
Scenario
—       •
Scenario
_^^—^^-^^-ซ
Scenario
            BAT3 (M&P)+BAT2
D Scenario
r^—
B Scenario 7
1	   	
|Scenario8
Source: EPA Analysis of NWPCAM results databases, 1/10/2002.
                                               7-13

-------
                                          Table 7-6
                         Summary of Monetized Benefits (97 Facilities)
               (Willingness to pay for changes from baseline water quality, $ 1999)

Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Scenario 6
Scenario 7
Scenario 8
BAT2 Only
BAT3 Only
BAT4 Only
BAT2 + PSES1
BAT3+PSES1
BAT4 + PSES1
BAT3 (M&P)+BAT2

Total Monetized Benefits
Continuous
$15,469,000
$15,578,000
$15,615,000
$15,919,000
$16,029,000
$16,066,000
$15,578,000
$16,029,000
Use Change
$1,032,000
$1,115,000
$1,115,000
$1,806,000
$1,890,000
$1,890,000
$1,115,000
$1,890,000
Rank Order of Scenarios
Continuous
1
2
4
5
6
8
2
6
Use Change
1
2
2
5
6
6
' 2
6
Source: EPA Analysis of NWPCAM results databases, 1/10/2002.
                                              7-14

-------
                                          Table 7-7                                .
       Households and River Mileage Affected by State, Proposed Scenario 7 (97 Facilities)
                                (Miles, unless otherwise noted)
Maryland
————^™—
Nebraska
,ซ,^^—^—ซ^^—
Wisconsin
Source: EPA Analysis of NWPCAM results databases, 1/10/2002.
                                                 7-15

-------
                                             Table 7-8
          Households and Changes in WQI by State, Proposed Scenario 7 (97 Facilities)
State
Alabama
Arkansas
Florida
Georgia
Illinois
Iowa
Kansas
Kentucky
Louisiana
Maryland
Minnesota
Mississippi
Missouri
Nebraska
Oklahoma
South Dakota
Tennessee
Texas
Virginia

Households
(Thousands)
1,720
1,003
6,083
2,941
4,590
1,141
1,033
1,548
1,654
1,971
1,852
1,031
2,161
658
1,332
287
2,172
7,357
2,668
2041
Households per River
.-.:,• .."".. Mile;;—v>
119.4
78.1
926.7
190.9
383.5
73.4
60.6
123.0
158.1
634.2
111.2
87.7
121.5
41.4
88.2
15.6
171.4
155.9
• 217.3
163.9
Total Mileage
Change in WQI
290.0
.52.6
1.0
41.2
1,255.3
1,964.9
3.9
1.0
12,6
46.0
977.7
35.2
123.1
76.1
2.7
1.0
4.8
107.0
4.8
3.699.0
Average Change ii
WQI
1.9
0.6
1.0
0.7
4.0
4.7
1.0
1.0
1.0
2.7
9.0
0.8
0.9
1.7
0.1
0.5
1.0
11.9
0.4
14!7
Source: EPA Analysis of NWPCAM results databases, 1/10/2002.
Note: Total Mileage Change in WQI is the sum of the differences between WQI under Option 7 and WQI in the
baseline for each'reach that changed in the state multiplied by the length of the reach, i.e., for each state,
ฃ (W,, - W0,)&r The average change in WQI is this value divided by the total length of rivers in the state that are
affected by the proposed option.  Thus, the average refers only to the average among water bodies affected, not all
waters in the state, and is weighted by the length of water bodies affected.
                                                  7-16

-------
                                         Table 7-9
            Total Benefits by State, by Use Category Change Method (97
          (Willingness to pay for changes from baseline water quality, thousand $1999)
State
Alabama
• i      ••
Arizona
• i     •
Arkansas
California
•   i    ""
Colorado
 — i  -i    -
 Connecticut
 	
 Delaware
 __
 District of Columbia
 Florida
 Georgia
     -*•"
 Idaho
  ^^—"^^
  Illinois
  ^~——~~~
  Indiana
  MB^~M
  Iowa
  Kansas
  ^—™—
  Kentucky
  Louisiana
  _^™^^-^—
  Maine
  M^^^MW^
  Maryland
   Massachusetts
           —
   Michigan
   Minnesota
   .
   Mississinoi
   •     •—
   Missouri
   _.
   Montana
12
 111
 70
  35
 184
 16
150
        11
                   35
184
                                                7-17

-------
                               Table 7-9 (cont.)
(Total willingness to

-------

-------
              distressed position but no disproportionate effects on a particular region or segments of the
              private sector (Chapters 5 and 6);
              Section 202(a)(3)(B) — disproportionate effects on local communities. EPA projects one
              meat products site to close as a result of the costs of the proposed combination of options
              and one large company to move into a financially distressed position but no
              disproportionate effects on local communities (Chapter 5).
              Section 202(a)(4) — estimated effects on the national economy (Chapter 5);
              Section 205(a) — least burdensome option or explanation required (this Chapter).

The preamble to the proposed rule summarizes the extent of EPA's consultation with stakeholders including
industry, environmental groups, states, and local governments (UMRA, sections 202(a)(5) and 204).
Because this rule does not "significantly or uniquely" affect small governments, section 203 of UMRA does
not apply.                      ,

        Pursuant to section 205(a)(l)-(2), EPA has selected the "least costly, most cost-effective or least
burdensome alternative" consistent with the requirements of the Clean Water Act (CWA) for the reasons
discussed in the preamble to the rule. EPA is required under the CWA (section 304, Best Available
Technology Economically Achievable (BAT), and section 307, Pretreatment Standards for Existing
Sources (PSES)) to set effluent limitations guidelines and standards based on BAT considering factors
listed in the CWA such as age of equipment and facilities involved, and processes employed. EPA is also
required under the CWA (section 306, New Source Performance Standards (NSPS), and section 307,
Pretreatment Standards for New Sources (PSNS)) to set effluent limitations guidelines and standards based
   Best Available Demonstrated Technology. EPA determined that the rule constitutes the least
 on
 burdensome alternative consistent with the CWA.
 8.3     REFERENCES
 Katzen  1996 Economic Analysis of Federal Regulations Under Executive Order No. 12866.
         Memorandum for Members of the Regulatory Work Group from Sally Katzen, Ad, OIRA.
         January 11,1996.
                                                8-3

-------
8.2     UNFUNDED MANDATES REFORM ACT ANALYSIS

       Title E of the Unfunded Mandates Reform Act of 1995 (Public Law 104-4; UMRA) establishes
requirements for Federal agencies to assess, the effects of their regulatory actions on State, local, and tribal
governments as well as the private sector.  Under Section 202(a)(l) of UMRA, EPA must generally
prepare a written statement, including a cost-benefit analysis, for proposed and final regulations that
"includes any Federal mandate that may result in the expenditure by State, local, and tribal governments, in
the aggregate or by the private sector" of annual costs in excess of $100 million.2 As a general matter, a
federal mandate includes Federal Regulations that impose enforceable duties on State, local, and tribal
governments, or on the private sector (Katzen, 1996).  Significant regulatory actions require Office of
Management and Budget review and the preparation of a Regulatory Impact Assessment that compares the
costs and benefits of the action.

        The proposed meat products industry effluent limitations guidelines are not an unfunded mandate
on state, local, or tribal governments because industry bears the cost of the regulation. The pretax cost
estimate to industry ranges from $80.0 million per year to $112.1 million per year, while posttax costs —
costs out of industry's pocket — range from $50.5 million (retrofit costs) to $73.8 million (upper-bound   •
costs). Thus, it is not clear that the proposed rule is an unfunded mandate on industry.  EPA, however, is
responsive to all required provisions of UMRA. In particular, this Economic Analysis (EA) addresses the
 requirements of UMRA:

                Section 202(a)(l) — authorizing legislation (Chapter 1 and the preamble to the rule);
                Section 202(a)(2) — a qualitative and quantitative assessment of the anticipated costs and
                benefits of the regulation, including administration costs to state and local governments
                (Chapters 5 and 7);
         •       Section 202(a)(3)(A) — accurate estimates of future compliance costs (as reasonably
                feasible; Chapter 5);
                Section 202(a)(3)(B) — disproportionate effects on particular regions or segments of the
                private sector. EPA projects one meat products site to close as a result of the costs of the
                proposed combination of options and one large company to move into a financially
     2 The $100 million in annual costs is the same threshold that identifies a "significant regulatory action" in Executive Or|
                                      .8-2

-------
                                       CHAPTER 8
                        COST-BENEFIT COMPARISON AND
               UNFUNDED MANDATES REFORM ACT ANALYSIS
8.1    COST-BENEFIT COMPARISON

       The pretax annualized costs of the proposed rule range from $80.0 million (retrofit costs) to
$112.1 million (upper-bound costs). The pretax cost is aproxy for the social cost of the regulation because
it incorporates the cost to industry (posttax costs), and costs to State and Federal governments (i.e., lost
income from tax shields).1 In other words, the cost part of the equation is well-identified and estimated.

        The estimated quantified and monetized benefits of the rule range from $1.1 million (use category
 change method) to $15.6 million (continuous method). These benefits estimates reflect only the 94 plants
 (36 direct dischargers) actually analyzed for water quality improvements. The corresponding annualized
 costs for these facilities are $33.7 million. If the ratio of costs to benefits for these facilities is the same as
 the ratio of costs to benefits for all facilities, the total (continuous) benefits of the rule would be $37.0
 million. This, however, is an underestimate because EPA can fully characterize only a limited set of
 benefits to the point of monetization.  Chapter 7 focuses mainly on the public's willingness to pay for
 improvements in the recreational use of water bodies (e.g., boating, swimming). However, other benefits
 may accrue due to the proposed rule that are not included in these monetized values.  Water withdrawn for
  municipal or industrial uses may need less pretreatment. The value of waterfront property may be
  increased if water quality is improved.  The benefits estimates do not include improved POTW operations
  and reduced costs at POTWs.  Finally, the proposed regulation will generate improvements in habitat and
  ecosystem services which are valued for their existence. Therefore, the reported benefit estimate
  understates the total benefits of this proposed rule.
     1 All sites are currently permitted and permits are reissued on a periodic basis, so incremental costs
  administrative costs of the regulation are negligible.
                                                 8-1

-------
the rule would be $37.0 million.  There is less than a $1 million difference between the least and most

beneficial scenarios.
 7.3     REFERENCES

 Bondelied, Timothy (RTI). 2001. Personal Communication with Will Wheeler, EPA, and Drew
        Laughland, ERG, September 28, 2001.

 Carson RichardT and Robert Cameron Mitchell.  1993.  The Value of Clean Water: The Public' s
        WuSgness to Pay for Boatable, Fishable, and Swimmable Quality Water. Water Resources
        Research, 29(7 July):2445-2454.

 McClelland, Nina I.  1974.  Water Quality Index Application in the Kansas River Basin. Prepared for U.
        S. EPA-Region VE.

 US EPA 2001 Environmental and Economic Benefit Analysis of Proposed Revisions to the National
        PollutantScharge Elimination System Regulation and the Effluent Guidelines for Concentrated

        SX^
        Benefits of Achieving Recreational Use levels. Washington: EPA/Office of Water, EPA 821 R
        01-002.  January, 2001.

  U S EPA 2002  Environmental Assessment of Proposed Effluent Limitations Guidelines
        td Standards for the Meat and Poultry Products Industry Point Source Category. Washington.
         EPA/Office of Water, EPA-821-B-01-008.

  Vauehan William J  1986. The RFF Water Quality Ladder, Appendix B in Robert Cameron Mitchell and
         Richard T Carson, The Use of Contingent Valuation Data for Benefit/Cost Analysis  in Water
         Pollution Control, Final Report. Washmgton:Resources for the Future.
                                                7-22

-------
substantial. In several states, only a small number of water bodies will be affected by the proposed
regulation so both total and average WQI changes are quite small. The conversion from change in WQI to
monetized benefits is non-linear as changes in some use categories are more valued than others. Thus, a
rank ordering of states from Table 7-8 may not match the rank ordering of states by total monetized
benefits.

        The difference between the two methods is much more pronounced at the state level than at the
national level. Table 7-9 shows the state totals for the sample plants using the change in use category
method. All states show a benefit from the proposed rule because their residents value the change in out of
state water quality. The largest benefits accrue to Maryland households. Maryland has a large population
relative to the mileage of streams in the state and a larger proportion of river miles affected by the
regulation than other states. Georgia, for example, has 5.5 miles of streams changing categories because of
the regulation compared to 5 miles in Maryland.  However, Maryland has three times the number of
households per river mile and generates almost three times the value of benefits from similar mileage
affected.

        Table 7-10 presents the total benefits by state using the  continuous method. Many more states are
shown to generate benefits from the regulation.  Illinois and Wisconsin generate markedly greater benefit
values because water quality improvements that do not generate use category changes are included. The
difference in results from each method depends on the number of water bodies that were near one of the.
breakpoints on the Vaughan water quality ladder. The rate of accrual of benefits changes at the
breakpoints under the continuous method but there is no substantial reward for crossing a breakpoint. The
use category change method only rewards crossing the breakpoints.

        In addition, states with large populations generate greater benefits for improvement in out of state
waters. California and New York together now generate almost $1 million in benefits even though few of
the water quality changes are near their waters.

        The monetizable benefits from the proposed rule, Scenario 7, for the 97 sampled plants are $15.6
million by the continuous method and $1.1 million by the use category method. If the ratio  of costs to
benefits for all facilities is the same as the ratio of costs to benefits for these facilities, the total benefits of
                                                7-21

-------
                                    -10 (cont.)
(Total wfflinguess *>1
VState

ft     "
BMontana^



|Nebraska


1
BNevada

r

IM^W Hampshire,



 |New Jersey


      rMexico ^_



  JNew York_


  JNorthCarolina,



  |Nr>rth Dakota_



  lohig,


   loklahoma


   I	
   lOregon,
   w

   |pennsYlvania_


   S"	
    itthode Island

    I	
    SsouthCarohna_


    1—'—
    knuthDakota_



    (Tennessee
    |__^


     |Texaฃ_

     I
     lUtah^
      Ivirginia^



      BWashington _



      SwP.st Virginia


      H    '
      8Wisconsin_
      it
                      17
                     ซ^^™



                      34
                     • ••



                      32
                      ^^MM>



                      21
                      • •••



                      138
                      _•ป



                       30
                      —^—"—



                      321
                      —  -



                      137
                       		



                        12
                       •  —



                       203
                       •——•



                        61
                       ซซ•'•



                        62
                       .^——•



                        217




                         17
                        • "••



                         68
                         ^ซซ™



                          13
                         ซซv



                         101
                        —  •-



                         585
                         _ —



                          32
                          .•——



                           11
                          —P——



                          128
                         —  ••



                          106
                          _——



                           34
17
• •



46
•——•



 32




 21
•i <•



139
  —^



  30
  — —



 324
 • i —



  138
  • i •



   12
  .•—



  205




   64
   i——•



   63
   ——•



   219




    18
   _ซ•



    69




    13
    ^MMซ



   102
   _.



    589
    __



     32
     ——-



     11
     ——'



     129
     ซ-ซ^*-



     107
      H^^



      34
17
M^^



46
•^•H*



32
• •—



 21
• -



140
 ^M^M



 31
 • i —



 325
 — —



 139
  _ i —



  12
  • —



 206
  ^•i^^



  70
  •n i-



   63
  ซ——



  220




    18
   ปMW



    69
    ซ^^



    13
    ซ••>



   103
   .—•—•



    590




     33




     11
     ปป•••



     130
    •——



     107
      >^^MI



      34
^Scenario


S-p-i

  TTT_i8


  STL58

  irrii^4-


   22jL__22

  144T   145
     -~

    31
   •v—^

   334
   143
    • •



     12
    •ซ—^—



    212
     ,•—^—



     84
     — —



      65
     • '•-



     226




      18
      ^v>



      71
      ซซM



      14
      ——•



      105
      ••  ••



      600
      ซ——



       33
       •—^—



        11
       •^•MV



       143
      •  ••



       110




        38
 32
• '••



338
••  •



144
 •—i—



  12
 i^ซ^—



 214
  iป^—



  87
  ^^^



   66
  _



  228




   18
   • i—



   71




   14
   ——•



   W7




   604
   • ••



    34




     11
    _ —



    145




    111
     • *•



     38
                         =P^HB
                             9 4	T. .. ,- . Li c 919  $16,029
   ir-	            isisleTfesTS^^

   ^^^s^^^^^
                                            7-20
6




  18
  —•



  58
  _  —



  34
  -i •-



   22
  ซ-.—•—



  146
   •i ••



   32
   ^•^^^



  339
  •  -•



   145
    .ปป•



    12
    •i •-



   215




    93
    1^1 —



    66
    ——^



    229
     ^^^



     18
     ^—~~



     72
      !• •



      14
      ซซ-••



     107
     i^_M^



     605




       34
       • •



       11
       • •



      146
      .^——•



       112
       • •



        38
       ^^•MV



     3349.




         9
    _	 —


    $16,066

       	—•""
7




  17
  ^~—



  46




  32
  •vซ



  21
  ^^—



  139
   .•_••



   30
   W>



  324
  M—^^™



   138
    _ซ•



    12
     —



   205
    .——



    64
    •B^*"



     63
    ^—~



    219
     • ••



     18
     M^^



     69
      •—•



      13
      •^^"



     102
     •  •••



     589
      —^



       32
       ซ•_••



       11
       ^^^^



      129
      — i  —



       107




        34
       ^•^^^



     3,344
8



  18
  • ••



  58
    —



  34
  — —



   22_




  145_




   32'
   — -•



   338
   •  •••



   144
    ^^•^^



    12
    ,——•



   214^
    --.•••"



     87
     •i -



     66
     • ••



    228_




     18
     ,•••—



     71
      .——



      l_4j




     107
        —



     604
      •.i ••



      34
      ^•^^



       11
      ^_v



      145,




      111.



       38
       •^-—•



     334ง_



         9
         ——


    |$16,029
                                               _9



                                          "$15,578
                                              ::——"*

-------
                                 Table 7-10
             Benefits by State, by Continuous Method (97 Faculties)
(Willingness to pay for changes from baseline water quality hi state, thousand $1999)

State
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky 	
Louisiana
Maine
Maryland 	
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
. Scenario
1
344
84
52
555
75
58
13
10
284
218
22
4,301
105
1,360
48
71
80
23
845
110
175
717
57
131
2
346
85
67
561
76
58
13
11
287
220
22
4,328
106
1,360
52
72
80
23
846
111
177
718
57
132
3
346
85
72
563
76
59
13
11
288
222
22
4,328
107
1,361
52
72
81
23
846
111
177
719
57
140
V..4-"
347
88
72
581
79
60
14
17
295
275
22
4,312
110
1,446
50
74
83
24
865
114
182
726
64
149
5
349
89
87
587
79
61
14
17
298
278
23
4,338
111
1,447
54
75
84
24
866
115
184
727
64
150
6
349
89
92
589
80
61
14
17
299
280
23
4,339
111
1,447
54
75
84
24
866
116
184
727
64
158
'. •!•,.'••
346
85
67
561
76
58
13
11
287
220
22
4,328
106
1,360
52
72
80
23
846
111
111
718
57
132

8
349
89
87
587
79
61
14
17
298
278
23
4,338
111
1,447
54
75
84
24
866
115
184
727
64
150
                                      7-19

-------
United States
Environmental Protection
Agency
            Office of Water (4303)
            Washington, DC 20460
EPA-821-B-01-006
February 2002
Economic Analysis of Proposed
Effluent Limitations Guidelines
and Standards for the Meat and
Poultry Products Industry:
Appendices

-------

-------
                                       APPENDIX A
                           COST ANNUALIZATION MODEL
       Figures A-l and A-2 provide an overview of the cost annualization model as used for analysis of
the proposed rule, and as will be used for analysis of the final rule respectively. Inputs to the model differ
in each analysis because for the analysis of the final rule, data from the 2001 Meat Products Industry
Survey detailed questionnaire will be used in addition to other data from the proposal analysis.  The inputs
for proposal include the capital and operating and maintenance (O&M) costs for incremental pollution
control developed by  EPA, and a variety of secondary sources. The cost annualization model calculates
four types of compliance costs for a site:                                          ,

        •       Present value of expenditures — before-tax basis
        •       Present value of expenditures — after-tax basis
        •       Annualized cost — before-tax basis
        •       Annualized cost — after-tax basis

        There are two reasons why the capital and O&M costs should be annualized. First, the initial
 capital outlay should not be compared against a site's income in the first year because the capital cost is
 incurred only once in the equipment's lifetime. That initial investment should be spread over the
 equipment's life. Second, money has a time value. A dollar today is worth more than a dollar in the future;
 expenditures incurred 15 years from now do not have the same value to the firm as the same expenditures
 incurred tomorrow.

        The cost annualization model is defined in terms of 1999 dollars because the latest year for which
 financial data will be available from the detailed survey is 1999.  Pollution control capital and O&M costs
 are estimated in 1999 dollars and used to project cash outflows.  The cash outflows are then discounted to
 calculate the present value of future cash outflows in terms of 1999 dollars. This methodology evaluates
 what a business would pay in constant dollars for all initial and future expenditures. Finally ,J the model

                                               A-l

-------
Data Sources      Inputs
                                                       _. Outputs
Engineering
Incremental
Pollution Control
Costs
Secondary
Sources
Capital Costs
                   O&M Costs
Cost Deflator to
$1999
                   Depreciation Method-
                   (MACRS)
                   Federal Tax Rate


                   State Tax Rate
                   Discount Rate  	
                   (OMB)

                   Taxes Paid
                   (Limitation on Tax
                   Shield; Modeled from
                   Census data)

                   Tax Status
                   (by Assumption)
                           Cost Annualization
                                Model
                                              8**.,
 Present Value
of Expenditures
                                                   Annualized
                                                      Cost
                                      Figure A-l

                  Cost Annualization Model for the Proposal Analysis
                                          A-2

-------
 Data Sources      Inputs
                                                         Outputs
 Engineering  .
 Incremental
 Pollution Control
 Costs
 Secondary
 Sources
Capital Costs
                    O&M Costs
Cost Deflator to
$1999
                    Depreciation Method
                    (MACRS)
                    Federal Tax Rate
                    State Tax Rate
2001 Meat Products   Discount Rate
Industry Survey
                    Taxes Paid
                    (Limitation on Tax
                    Shield)


                    Tax Status
                    (Corporate or Personal)
                           Cost Annualization
                                 Model
                                                       ^t ^^ป\&,KX. .
 Present Value
of Expenditures
                                                    Annualized
                                                       Cost
                                       Figure A-2

                     Cost Annualization Model for the Final Analysis
                                            A-3

-------
calculates the annualized cost for the cash outflow as an annuity that has the same present value of the cash
outflows and includes the cost of money or interest.  The annualized cost is analogous to a mortgage
payment that spreads the one-time investment of a home into a defined series of monthly payments.

        Section A.I discusses the data sources for inputs to the cost annualization model for the proposal
analysis as well as the final analysis. Section A.2 summarizes the financial assumptions in the model.
Section A.3 presents all steps of the model with a sample calculation.
A.I    INPUT DATA SOURCES

        A.1.1 "EPA Engineering Cost Estimates

        The capital and O&M costs used in the cost annualization model are developed by EPA's
engineering staff.  The capital cost is the initial investment needed to purchase and install the equipment; it
is a one-time cost.  The O&M cost is the annual cost of operating and maintaining the equipment.  O&M
costs are incurred every year of the equipment's operation. For proposal, EPA estimated average
compliance costs for a series of model facilities based on subcategory, size, and discharge type (for details
see Development Document, U.S. EPA, 2002). For the final rule, EPA will use model facilities developed ,
from detailed questionnaire data.
       A.1.2  Secondary Data

       The cost annualization model is developed in terms of constant 1999 dollars. Hence, as necessary,
all costs are deflated to 1999 dollars for the cost annualization model using a cost deflator.  As mentioned
above, engineering cost estimates are already in 1999 dollars. However, in the proposal analysis, income
measures and the variance of their distributions were derived from Census data in 1997 dollars and need to
be adjusted. EPA calculated the implicit price deflator for Food and Kindred Products from Bureau of
Economic Analysis' Gross Domestic Product by Industry data (U.S. DOC, 2000). For analysis of the final
rule, income measures and other survey data will be in 1999 dollars.
                                             .A-4

-------
        The depreciation method used in the cost araiualization model is the Modified Accelerated Cost
Recovery System (MACRS).  MACRS allows businesses to depreciate a higher percentage of an
investment in the early years and a lower percentage in the later years.

        Tax rates are determined by the Federal tax rate plus the national average state tax rate. Table A-
1 presents the Federal tax rate for corporations and individuals (CCH, 1999b). The Federal tax rate is
calculated from a graduated system with a tax rate for each level of taxable income. Table A-2 lists each
state's top corporate and individual tax rates and calculates national average state tax rates (CCH, 1999a).
The cost annualization model uses the average state tax rate because of the complexities of the industry; for
example, a site could be located in one state, while its corporate headquarters are located in a second state.
Given the uncertainty over which state tax rate applies to a given site's revenues, the average state tax rate
— rounded to three decimal points — is used in the cost annualization model for all sites (i.e., 6.6 percent
corporate tax rate and 5.6 percent personal tax rate).

        For the proposal analysis, taxable income — earnings before interest and taxes (EBIT) — is
derived from Census data: Derivation of EPA's estimate of EBIT for model facilities is discussed in more
detail in Appendix B. For the final analysis, EPA will use the value of EBIT reported in the survey. The
value of EBIT determines the tax bracket for the site.

        The cost annualization model incorporates variable tax rates according to the level of income to
address differences between small and large businesses. For example, a large business might have a
combined tax rate of 40.6 percent (34 percent Federal plus 6.6 percent State). After tax shields, the
business would pay 59.4 cents for every dollar of incremental pollution control costs.  A small business,
say a small sole proprietorship, might be in the 20.8 percent tax bracket (15 percent Federal plus 5.8
percent State). After tax shields, the small business would pay 79.2 cents for every dollar of incremental
pollution control. The net present value of after-tax cost is used in the closure analysis because it reflects
the long-term impact on its income the business would actually experience.

        The discount rate is the minimum rate of return on capital required to compensate debt holders and
equity owners for bearing risk. It is also called the marginal weighted average cost of capital or the
                                               A-5

-------
                                            Table A-l
                                       Federal Tax Table
Corporate Tax Rate
Taxable Income
($1,000)

$0-$50
$50 - $75
$75 - $100
$100 - $335
$335 -$10,000
$10,000 - $15,000
$15,000 - $18,333
More than $18,333
Average
Effective
Tax Rate

15%
25%
34%
34% *
34%
35%
35% *
35%
: Individual Tax Rate
Taxable Income
($1,000)

$0 - $25.75
$25.75 - $62.45
$62.45 - $130.25
$130.25 -$283. 15
More than $283. 15
Average
Effective :
Tax Rate

• 15%
.28%
31%
36%
40%'



Source: CCH, 1999b. 2000 U.S. Master Tax Guide. Chicago, IL: CCH.
* For the $100,000 to $335,000 taxable income range, the actual tax rate is 38% and for taxable income between.
$15,000,000 and $18,333,333, the actual rate is 39%. However, these rates were temporarily imposed to phase out
certain benefits and hence, are not used here.
                                               A-6

-------
      Table A-2
State Income Tax Rates
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas ;
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri '
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
' • \ • - ''--• • '. ; .
Corporate Income
Tax Rate
5.00%
9.40%
8.00%
6.50%
6.65%
4.75%
7.50%
. 8.70%
5.50%
6.00%
6.40%
8.00%
4.80%
3.40%
12.00%
4.00%
8.25%
8.00%
8.93%
7.00%
9.50%
2.20%
9.80%
5.00%
6.25%
6.75%
7.81%
0.00%
8.00%
7.25%
7.60%
7.50%
7.50%
10.50%
8.50%
6.00%
?H V Basis for States ' /:''-'•;•.-'' -:•-:•-
With Graduated
Tax Tables

$90,000+

$100,000+



%


$100,000+



$250,000+

$250,000+
$200,000+
$250,000+




$10,000+


$50,000+



$lMillion+


$50,000+
$50,000+

Personal Income Tax
- Upper Rate
5.00%
0.00%
5.04%
7.00%
9.30%
4.75%
4.50%
6.40%
0.00%
6.00%
8.75%
8.20%
3.00%
3.40%
8.98%
6.45%
6.00%
6.00%
8.50%
4.80%
5.95%
4.40%
8.00%
5.00%
6.00%
11.00%
6.99%
0.00%
0.00%
6.37%
8.20%
6.85%
7.75%
12.00%
7.30%
7.00%
With Graduated
Tax Tables
$3,000-1

$150,0004
$25,000+
$47,000

$10,000+
$60,0004

$10,000+
$40,000+
$20,0004


$52,000+
$30,0004
$8,000+
$50,0004
$33,0004
$3,0004


$50,0004
$10,0004
$9,0004
$71,000+
$27,000+


$75,000+
$42,0004
$20,0004
$60,000+
$50,0004
$200,0004

        A-7

-------
                                          Table A-2 (cont.)
                                       State Income Tax Rates
State
Oregon
Pennsylvania
Rhode Island *
South Carolina
South Dakota
Tennesee
Texas
Utah
Vermont *
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Corporate Income
Tax Rate
6.60%
9.99%
9.00%
5.00%
6.00%
6.00%
0.00%
5.00%
9.75%
6.00%
0.00%
9.00%
7.90%
0.00%
Basis for States
With Graduated
Tax Tables








$250,000+





Personal Income Tax
Upper Rate
9.00%
2.80%
10.40%
7.00%
0.00%
0.00%
0.00%
'. 7.00%
9.45%
5.75%
0.00%
6.50%
6.77%
0.00%
With Graduated
Tax Tables
$5,000+

$250,000+
$12,000+



$7,500+
$250,000+
$17,000+

$60,000+
$15,0004


Average: I 6.58%

5.59%

Source:  CCH, 1999a. 2000 State Tax Handbook. Chicago, IL: CCH.
Basis for rates is reported to nearest $1,000.
* Personal income tax rates for Rhode Island and Vermont based on federal tax (not taxable income).
+ Tax rates given here are equivalents for highest personal federal tax rate.
                                                A-8

-------
weighted average of debt and equity rates.  The discount rate is used to calculate the present value of the
cash flows. As recommended by the Office of Management and Budget (OMB), for the proposal analysis,
a real discount rate of 7 percent is used to represent the opportunity cost of capital (OMB, 1996). For the
final analysis, the discount rate for each site will be obtained from the survey data. For sites that do not
report a discount rate, EPA will assign the median discount rate as the opportunity cost of capital.

        Average taxes paid is used to limit the tax shield to the typical amount of taxes paid in any given
year.  For the proposal  analysis, it is calculated as the amount of tax paid in 1999 by the model facility (see
Appendix B for more detail). In the final analysis, average taxes paid will be calculated from the 1997,
1998, and 1999 taxes paid by the site.

        Corporate structure is used for the purpose of estimating tax shields on expenditures. A C
corporation pays federal and state taxes at the corporate rate. An S corporation or a limited liability
corporation distributes  earnings to the partners and the individuals pay the taxes. For the purpose of the
proposal analysis, EPA assumes that all model facilities pay  federal and state taxes at the corporate rate.
In the final analysis, EPA will distinguish corporate structure based on detailed survey data. The tax rate
for S corporations and limited liability corporations will be presumed to be zero.1 All other entities will be
assumed to pay taxes at the individual rate.
A.2    FINANCIAL ASSUMPTIONS
        The cost annualization model incorporates several financial assumptions:
   1 The effect of this assumption is to assume there is no tax shield for S corporations and limited liability
corporations (LLCs). S corporations and LLCs will see no change in tax shield benefit because they do no.t pay
taxes. The persons to whom the income is distributed, however, will see the change in earnings due to incremental
pollution control costs; there is no tax shield benefit.
                                                A-9

-------
       •       Depreciation method is the Modified Accelerated Cost Recovery "System (MACRS)-2
               MACRS applies to assets put into service after December 31, 1986. MACRS allows
               businesses to depreciate a higher percentage of an investment in the early years and a
               lower percentage in the later years.

       •       There is a six-month lag .between the time of purchase and the time operation begins for
               the pollution control equipment. A mid-year depreciation convention may be used for
               equipment that is placed in service at any point within the year (CCH, 1999b, 
-------
would be 15-year property. According to IRS requirements, pollution control equipment can be
depreciated, but the total cost of the equipment cannot be subtracted from income in the first year. In other
words, the equipment must be capitalized, not expensed (CCH, 1999b, 1991; and RIA, 1999,  Section 169).
A.3    SAMPLE COST ANNUALIZATION SPREADSHEET
                         •'.,••      ~-
       In Table A:3, the spreadsheet contains numbered columns that calculate the before- and after-tax
annualized cost of the investment to the site. The first column lists each year of the equipment's life span,
from its installation through its 15-year depreciable lifetime.

       Column 2 represents the percentage of the capital costs that can be written off or depreciated each
year.  These rates are based on the MACRS and are taken from CCH (1999b). Multiplying these
depreciation rates by the capital cost gives the annual amount the site may depreciate; which is listed in
Column 3. Depreciation expense is used to offset annual  income for tax purposes; Column 4 shows the
potential tax shield provided from the depreciation expense—the overall tax rate times the depreciation
amount for the year.-

       Column 5 is the annual O&M expense. In this example, Year  1 shows six months of O&M
($10,000 -f 2 = $5,000).  Year 1 and Year 16 show only six months of O&M expenses because of the mid-
year convention assumption for depreciation. For Years 2 through 15, O&M is a constant amount.
Column 6 is the potential tax shield or benefit provided from expensing the O&M costs.
                                                                         a
       Column 7 lists a site's annual pre-tax cash outflow or total expenses associated with the additional
pollution control equipment. Total expenses include capital costs, assumed to be incurred during the first
year when the equipment is installed, plus each year's O&M expense.

       Column 8 is the adjusted tax shield. The potential tax shield is the sum of the tax shields from
depreciation (Column 4) and O&M/one-time costs (Column 6).  If the  potential tax shield for any year
exceeds the 3-year average taxes paid, the tax shield is limited to the average taxes paid by the facility. In
Table A-3 example, the potential tax shield in Year 2 is $2,052 plus $2,160 = $4,212. This exceeds the
                                              A-ll

-------
1
a

o
K
ฃ -o i 3
S '? 3 3
z i .2 5
5} ซ -a o


















c\
o\
OS









INPUTS
Survey ID ft

1
1





























i-
O
1
z
i
a
O

3"
c\
i>
ividualTaxTab
•a
B
H- >
•g
•s
o




CJ
3
53
p
H
1
•g
w
O
13
c\

•— '










8
8
i;
H.
o
1
1
_u
1
H

u
&0
>
^




u
1
ฃ— *



g
cT
o





S?

CO
5
1
u
3
[S

o
0
o
f
.> 1
t5 PS
sa g
6
o
o


fl> flj
^ *-
11
W f-




u
1




g
cf

g
o
U
8
c
S
c
Annual Operation & Mai






















=3
r —








Real Discount Rate:



o




0
o
i/i





0
ซ*














Corporate Tax Structure



o
s
in
V*

Q
o
in





0
o"
^o-



g
CO
8







ซ



o
o
vo


o
C5
S





O
O
>n
r-



m
CO
c^f





1


1999 Taxes Paid



o *o
CO CO
0 O
o co
CO OO
— C-)

0 0 Q
p p p
CO CO CO





000
p p p
o >n" o"
*— ( CO O
Wr iff ,-J
s

O VO
u-i vo






8

Marginal Income Tax Rz
Federal
State






















vq
ป— i








Combined

-------
















P— 4 '
0
O- S
ฃj ri
ll
*? .*

Q
o ^?
rt ^^
ro " '"
H
Q

c3

I*


~

Q
3
5



"^" VO ^O \D VO VO VD VO VO
oo vo vo VD vo VD VD vo vo
ง &fc6 fcfefc&fc

Ococnencncocoeoeo
voenenenencococo'co


^5. ^ej. ฃ/5- ^/^. ^5. &%. f-ft $fr ^15.





888888888
o o o o o o o o o
tnoooooooo
^ lfฑ yj ^Q ^g. ^ y^ ^g. ^
'•6^-

ooooooooo
oo vo VD vo vo VD vo vo vo





ooooooooo
ooooooooo
o o o o o o o o o
1000000000
jyj ^j ^ ^ ^ ^ ^ ^

SS5SSซSS!=
^g^^^ซe^ซe^




ooooooooo
to O^ oo r-— vo vo io vi 10
^45. ^/j. ^. ^. ^. ^^. ^5. ^5. ^/^.



^? vo *O I*** ON O4 ON ON ON
>o ON oo f-- vo vo *o *o vo

• .

•~^ o^ en ^f in vo t^ oo ON













VO \O vo vo
vo vo vo VD
fc fc ^^

co en co en
m en m co
CO CO CO CO

-6O- V^ -feO- -&O-





ง000
888
o o o o
Q^ ^ ^ ^


VO vo vo vo
8 S S a




o o o o
ง000
O 0 O
o o o o
^ ^ ^g. ^.

gg-gg
^ W 40- V^




o o o o
" o\ ON O\ ON
10 v^ • to vo
-9*3. ^15. ^yg- ^(5.



ON ON ON ON
vo vo vo vo



0 -H 




O O
ON ON
vo vo
•W- -W-



ON ON
"n in



Tt m













en I-H
oo vo
eป -*
en en
es

^
^H
t—

^5-





o o
o o
O 0
 %j-
>— i r-i
V3- en




8 8
O 0
-in o"
&$ >n
ซ*
r- o
en o
VD VD
S




O 0
,i
•^^ o
^
CS C^- VO
VO VO VD
CO *O C * *O
o" S **. o"
C<1 J3 O1 C-l
fee- 00 — > fr
s *^
H
1
&
%
as"
S

m -a — i o
C• s
s e
ฃ o

t,., u
0 g
g O
"=3 "S
> _S
c '"3'
Jg
G







CO
ฃ
3
•3 •
1
0
•a
•"^H
1
*O ••'"
g^.o
0) "S
^
*J o
1 fe
= ,
P o3
I &
on ^
E? ง
g ^
8^
ฃซง
0 ซ
8 •ซ
•S '3
s o
U ™
'T? ^^
o is
0 ซ
1 ^
lง
C3 S
tซ f— 1
•S c2
(S S
C

ซ

-------
average taxes paid over the last three years ($2,333) and hence, the tax shield for Year 2 is $2,333. This
approach is conservative in that the limit is applied every year when a company may opt to carry losses
forward to decrease tax liabilities in future years.  An alternative approach is to limit the present value of
the tax shield to the present value of taxes paid for the 15-year period. Should the first approach appear to
overestimate cost impacts, the second approach may be examined as a sensitivity analysis.

        Column 9 lists the annual cash outflow less the adjusted tax shield (Column 7 minus Column 8);
a site will recover these costs in the form of reduced income taxes.  The sum of the 16 years of after-tax
expenses is $250,000 (1999 dollars), i.e., the sum of the capital expense ($100,000) and 15 years of O&M
($150,000).  The present value of these payments is $194,267. The present value calculation takes into
account the tune value of money and is calculated as:
Present Value of Cash Outflows =
                                                            cash outflow, year.
                                                    ^,   - ^—.
                                                   i=i   (1  + real discount rate/"
The exponent in the.denominator is i-1 because the real discount rate is not applied to the cash outflow in
Year 1. The present value of the after-tax cash outflow is used in the closure analysis to calculate the post-
regulatory present value of future earnings for a site.
        The present value of the cash outflow is transformed into a constant annual payment for use as the
annualized site compliance cost. The annualized cost is calculated as a 16-year annuity that has the same
present value as the total cash outflow in Column 9.  The annualized cost represents the annual payment
required to finance the cash outflow after tax shields. In essence, paying the annualized cost each year and
paying the amounts listed hi Column 8 for each year are equivalent.  The annualized cost is calculated as:
        Annualized Cost = Present value of cash outflows x
                                                                  real discount rate
                                                             1  - (real discount rate +  l)"n
                                              A-14

-------
where n is the number of payment periods. In this example, based on the capital investment of $100,000,
O&M costs of $10,000 per year, a tax rate of 21.6 percent, and a real discount rate of 7 percent, the site's
annualized cost is $20,565 on a pre-tax basis and $ 18,110 on a post-tax basis.3


       The pre-tax annualized cost is used in calculating the cost of the regulation. It incorporates the
cost to industry for the purchase, installation, and operation of additional pollution control equipment as
well as the cost to federal  and state government from lost tax revenues.  (Every tax dollar that a business
does not pay due to a tax shield is a tax dollar lost to the government.)  Post-tax annualized costs are used
to shock  the market model because they reflect the cost to industry.
A.4    REFERENCES

CCH.  1999a.  Commerce Clearing House, Inc. 2000 State Tax Handbook. Chicago, IL.

CCH.  1999L.  Commerce Clearing House, Inc. 2000 U.S. Master Tax Guide.  Chicago, EL.

U.S. Department of Commerce, Bureau of Economic Analysis. 2000. Grass Domestic Product by Industry
        for 1997-1999. Survey of Current Business. Washington, D.C. December.

OMB.  1992. Guidelines and Discount Rates for Benefit-cost Analysis of Federal Programs. Appendix A.
        Revised Circular No. A-94. October 29. Washington, DC:  Office of Management and Budget.

RIA. 1999. the Research Institute of America, Inc. The Complete Internal Revenue Code.  New York,
        NY. July 1999 Edition.

U.S. EPA.' 2002. Development Document for the Proposed Revisions to the Effluent Limitations
        Guidelines for the Meat Products Industry.  EPA-821-B-01-007. Washington, DC: U.S.
        Environmental Protection Agency, Office of Water.
    3 Note that post-tax annualized cost can be calculated in two ways. The first way is to calculate the annualized
 cost as the difference between the annuity value of the cash flows (Column 7) and the adjusted tax shield (Column
 8). The second way is to calculate the annuity value of the cash flows after tax shields (Column 9). Both methods
 yield the same result.

                                              A-15

-------

-------
                                      APPENDIX B
                            FACILITY-LEVEL ANALYSIS
       EPA used publicly available information to project facility-level impacts under the proposed
rule. EPA based its facility-level analysis on the U.S. Census Bureau's 1997 Economic Census of the
following four industries: Animal (Except Poultry) Slaughtering (NAICS 311611), Meat Processed From
Carcasses (NAICS 311612), Rendering and Meat Byproduct Processing (NAICS 311613), and Poultry
Processing (NAICS 311615). The Census provides detailed revenue and cost information by
employment class, which EPA used to build model facilities. To analyze facility-level impacts based on
the Economic Census data, EPA compared estimated compliance costs with four measures of income:

       •       Average establishment revenues
       •       Average establishment earnings before interest and taxes (EBIT)
       •       Average establishment net income
       •       Average establishment cash flow

Each level of analysis more closely approaches the goal of using estimated compliance costs to draw
strong inferences about definable impacts on the establishment, but each level of analysis requires
additional assumptions to generate the test data. Thus, each level of analysis presents a tradeoff. For
example, the relationship between facility net income and the impact of compliance costs is much more
clearly defined than the relationship between facility revenues and compliance cost impacts. Estimating
average facility net income requires more assumptions than estimating average facility revenues,
however, and that increases the uncertainty about the baseline benchmark against which impacts are
measured.
        Section B.I presents an intuitive overview of the strategy EPA used to develop model facilities
and measures of their income. Average facility values and the variance of those values are discussed in
sections B.2.1 through B.2.4 below — one section for each of the four proposed levels of analysis.
Section B.3 describes Issues concerning sabcategorizing the proposed model facilities and matching
those facilities with the engineering model facilities. Section B.4 examines a question concerning the

                                             B-l

-------
probability that some facilities may be projected to have negative income in the baseline. Section B.5
outlines some qualifications and limitations of the methodology used to model meat product facilities.
B.I    GENERAL MODELING STRATEGY
       For each level of analysis, EPA's strategy was similar. First, average revenues, net income, or
cash flow was estimated for model establishments of different sizes. EPA based its size classification for
developing model establishments on facility employment, taking advantage of the detailed information
the Census Bureau provides by employment class. Table B-l presents the number of establishments by
employment class within each industry. The number of employment classes within each industry is large,
providing a good level of detail, and the number of observations within each employment class is  '
generally large. Thus, the average facility income measures should not be skewed by a small number of
atypical observations.                                                           •

       Using average income alone as the basis for projecting economic impacts on model
establishments imposes a limitation on the analysis.  Simple comparison of average compliance costs
with die model facility's average income generates an all-or-nothing result: all facilities represented by a
particular model incur impacts identical to those of the model facility. For example, if the model facility
is projected to close because it incurs compliance costs exceeding cash flow, then all facilities
represented by that model are projected to close. In reality, however, incomes of the actual facilities that
the model represents compose a.distribution around the mean income (i.e., the model facility's income)
for that group of facilities. Actual facilities that are smaller than the average, therefore, may be negatively
impacted by the proposed rule even if the model facility appears unaffected. Conversely, larger-than-
average facilities may be unaffected by the rule even if the model facility is  affected.

       To deal with this limitation, EPA estimated the distribution of facility income around the model
facility mean. In order to do this, EPA obtained from the Census Bureau a special tabulation of the
variances and covariances of important income components around their respective mean within each
employment class (U.S. Census Bureau. 2001).  Combining this information with the assumption that
these observations are normally distributed around the mean, EPA constructed a distribution of
revenues, EBIT, net income, and cash flow for the group of facilities represented by each model. Given
                                              B-2

-------
                                            Table B-l
               Number of Establishments by Industry and Employment Class, 1997
Establishment Size by
Number of Employees
Ito4
5 to 9
10 to 19
20 to 49
50 to 99
100 to 249
250 to 499
500 to 999
1,000 to 2,499
2,500 or Greater
Total
Number of Establishments in NAICS Industry:
311611:
Animal
Slaughter -
507
275
225
141
79
64
33
21
39
9
1,393
311612:a
Meat Processed
From Carcasses
293
176
206
246
140
143
68
25
0
0
1,297
311613:b
Rendering
27
30
40
81
62
0
0
0
0
0
240
311615:
Poultry :
Processing
54
18
15
35
. 34
. 67
79
97
70
5
474
Source: U.S. Census Bureau, 1997a through 1997d.
a Due to disclosure issues, the 500-to-999-employee establishment size for NAICS 311612 (Meat Processed From
Carcasses) includes data for 2 facilities with employment between 1,000 and 2,499 and 1 facility with employment
greater than 2,500.
b Due to disclosure issues, the 50-to-99-employee establishment size for NAICS 311613 (Rendering) includes data
for 10 facilities with employment between 100 and 249 and 1 facility with employment between 250 and 499.
                                                B-3

-------
the large number of observations within each employment class (see Table B-l), the assumption of a
normal distribution around each mean should be acceptable.

       Having generated a distribution around the model facility mean, EPA compared estimated
compliance costs with an appropriate benchmark for each model in order to project the number and
percentage of facilities estimated to close under the effluent guideline. Suppose, for example, that a
model facility has an average cash flow of $100,000. That model facility represents an entire class of
facilities, some of which will earn cash flow less than $100,000. If compliance costs are estimated to be
$40,000 for the model facility, then the model facility itself would not be projected to close, but other
facilities in the same class with cash flow of $40,000 or less would be expected to close. Given the mean
and variance of cash flow for that model class, the probability that facilities in that class earning less than
$30,000 in cash flow can be readily calculated. Multiplying that probability by the number of facilities in
the class results in the projected number of closures for that class.  Multiplying the projected number of
closures by the average number of employees per facility in the employment class results in an estimate
of employment impacts.        •      •

       This methodology is illustrated in Figure B-l. The curve represents the cumulative distribution
function for cash flow around the model facility average of $100,000.  For the purpose of this
illustration, EPA set the standard deviation of the distribution equal to 100,000,  and EPA assumed cash
flow is normally distributed.1 The vertical line marking the estimated average annualized compliance
costs of $40,000 determines the probability of closure. Reading from the point on the graph where the
distribution function intersects the compliance cost marker, the probability that a facility earns cash flow
that is less than $40,000 per year is about 28 percent. Note, however, that the distribution function also
shows that about 16 percent of facilities in this class already have cash flow less than zero before the
regulation is promulgated (the point where the distribution crosses the $0 value).  Therefore, the
incremental probability that a facility hi this  model class will close due to the regulation is about 12
        1 The standard deviation of a distribution is equal to the square root of the variance of the distribution.  Thus,
standard deviation and variance are equivalent ways of measuring the dispersion of a distribution around its mean
value. A larger variance for a given mean value reflects a more dispersed distribution; the curve in Figure B-l would
be flatter.
                                               B-4

-------
                                 Figure B-l
                     Baseline Distribution Function for
                      Model Establishment Cash Flow
   1.00
   0.75
t
   0.50
   0.25
                            I
• Cash Flow
 Corap. Costs
    0.00
    -$200,000 -$100,000    $0    $100,000 $200,000  $300,000 $400,000
                             Cash Flow
                                      B-5

-------
percent (28 percent minus 16 percent).2 Multiplying this incremental probability of closure by the
number of establishments in the model class results in EPA's projected number of closures due to the
proposed rule.

       To employ this modeling strategy, EPA must develop measures of several parameters used to
create the models. First, EPA must develop estimates of average model facility income in each class to
be examined. Second, it must estimate the variance — or dispersion — of income for each class. EPA
used a variety of publicly available data sources to develop its estimates of these parameters.  Third, EPA
must estimate how income is distributed (i.e., the shape of the cumulative distribution function) in each
class. As described above, EPA assumes that facility income is normally distributed in each class.
Finally, EPA must match its model facilities developed from economic and financial data to the model
facilities used to estimate compliance costs based on engineering data. Each of these components in
EPA's modeling strategy is examined in detail in the sections to follow of this Appendix.3
B.2     FACILITY INCOME

        B.2.1  Facility Revenues

        The Census Bureau publishes the value of total shipments by employment size for each NAICS
code, along with the number of facilities in that size class. The value of total shipments includes the
value of primary and secondary shipments as well as resale, contract, and other miscellaneous receipts.
This makes the value of total shipments a reasonable proxy for total revenues. EPA calculated average
facility revenues by employment class within each industry as the value of total shipments divided by the
number of establishments in each class. EPA obtained from the Census Bureau the variance of the value
        2 EPA cannot evaluate the effect of the regulation on facilities with negative cash flow in the baseline
("baseline closures"). As discussed in Section 3.1.2, the basis for EPA's closure analysis is that an establishment
must have positive earnings prior to the regulation, and negative earnings after regulation. If an establishment has
negative earnings prior to the regulation, then it may very well close even if the regulation is never promulgated.
Thus, closure of such an establishment should not be considered an impact of the regulation.
        •* EPA explored the implications of using different data sources to estimate the variance of income
distribution, as well as alternative assumptions concerning the distribution of income within each class. The
sensitivity analyses are presented in Appendix E.
                                               B-6

-------
of shipments around the mean within each employment class. Table B-2 presents the mean and standard
deviation of revenues for each employment class in the affected NAICS codes.
       B.2.2  Facility EBIT

       B.2.2.1 Average Facility EBIT by Employment Class

       Next, EPA estimated average model facility EBIT in each class. EBIT is calculated by subtracting
cost of goods sold, general, sales, and administrative costs (GS&A), and depreciation and amortization
from total revenues.  EBIT then becomes the basis for calculating model facility net income and cash
flow.

       As with revenues, EPA compared estimated compliance costs and the distribution of EBIT by
employment class to project the number and percentage of facilities expected to incur costs exceeding
specified percentages of EBIT. There are no clearly defined thresholds for measuring impacts relative to
EBIT, as there are for income measures- like cash flow. Although clearly a facility would be projected to
close  if its pretax annualized compliance costs exceeded its EBIT, a facility would also be projected to
close  if its compliance costs were some fraction of EBIT (i.e., if the facility also had to pay taxes and
make interest payments on loans out of EBIT to remain open). Nonetheless,  using EBIT as a  benchmark
against which to compare compliance costs is an improvement over using revenues alone as an income
measure, since the latter make no allowance for facility operating costs.

       EPA used 1997 Economic Census data to estimate model facility EBIT and its variance by
employment class within each NAICS industry (U.S. Census Bureau, 1999a - 1999d). Facility revenues
were  estimated by value of shipments. The Census Bureau provides most of the significant categories of
operating costs that would be included in EBIT. For each of the four meat product  NAICS industries, the
Bureau provides:

       •       Payroll and material costs directly attributed to the employment class level
       •       Benefits, depreciation, rent, and purchased services attributed at the industry  level
                                              B-7

-------
                            Table B-2
Model Facility Income Mean and Standard Deviation by Employment Class
NAICS
Establishment
Employment
Size Class
Income Measure (x $1,000)
Revenues
Net Income
Cash Flow
Standard Deviation (x 1,000)
Revenues
Net Income
Cash Flow
NAICS 3 1 1611: Animal (Except Poultry) Slaughtering
Ito4
5 to 9
10 to 19
20 to 49 '
50 to 99
100 to 249
250 to 499
500 to 999
1,000 to 2,499
a 2,500
•$440
$1,265
$2,655
$8,413
$22,490
$69,474
$160,914
• $262,734
$677,948
$1,426,054
$28
$46
$64
$336
$1,303
$2,696
$4,005
$4,983
$29,321
$9,934
$33
$55
$86
$382
$1,438
$3,248
$4,714
$6,924
$33,489
$18,50r
292
842
1766
5598
14964
46227
107069
174819
451095
948872
56
89
147
617
2260
5211
8024
10403
53662
31988
56
89
• 147
617
2260
5211
8024
10403
53662
31988
NAICS 3 11612: Meat Processed From Carcasses : v ; ,
Ito4
5 to 9
10 to 19
20 to 49
50 to 99
100 to 249
250 to 499
500 to 9991
1,000 to 2,499
* 2,500
$413
$1,393
$2,845
$7,452
$19,049
$52,075
$105,066
$172,089
NA
NA
$30
$152
$160
$462
$1,823
$4,510
$6,308
$14,364
NA
NA
$40
$181
$204
$562
$2,045
$5,450
$7,555
$16,840
NA
NA
381
1286
2626
6877
17581
48062
96969
158827
NA
NA
81
320
367
1079
3819
9936
13266
31591
NA
NA
81
320
367
1079
3819
9936
13266
31591
NA
NA
NAICS 311613: Rendering -;:<:,?
Ito4
5 to 9
10 to 19
20 to 49
50to992
100 to 249
250 to 499
500 to 999
1,000 to 2,499
S 2,500
$860
$3,818
$6,476
$11,681
$17,108
NA
NA
NA
NA
NA
$14
$510
$608
$1,879
$2,406
NA
NA
NA
NA
NA
$40
$572
$730
• $2,244
$3,069
NA
NA
NA
NA
NA
1155
5128
8697
15688
22976
NA
NA
NA
NA
NA
311
794
1047
3199
4476
NA
NA
NA
NA
NA
• 311
794
1047
3199
4476
NA
NA
NA
NA
NA
                               B-8

-------
                                       Table B-2 (cont.)
           Model Facility Income Mean and Standard Deviation by Employment Class
NAICS ^r
Establishment
Employment
Size Class '
Income Measure (x -$1,000) _' *
?
"Revenues
Net Income
Cash Flow -
Standard Deviation (x 1,000)
'< , -/
Revenues *
Sf ~^ "
Net Income
Cash Flow
NAICS 311615: Poultry Processing
Ito4
5 to 9
10 to 19
20 to 49
50 to 99
100 to 249
250 to 499
500 to 999
1,000 to 2,499
> 2,500
$258
$759
$3,292
$11,721
$14,881
$29,999
$71,300
$117,768
$182,579
$321,884
$7
$23
$453
$2,428
$1,463
$2,324
$3,466
$13,362
$17,045
$1,072
$18
$40
$484
$2,564
$1,618
$2,745
$4,602
$14,784
$20,179
$7,856
158
. 465
2017
7184
9120
18386
43698
72177
11.1898
197275
28
70
631
3266
2225
3966
5956
20658
29094
4551
. 28
70
631
3266
2225
'3966
5956
20658
29094
4551
1 Due to disclosure issues, data for 2 facilities with 1,000 < employment < 2,499, and 1 facility with 2,500
employment combined in lower category for NAICS 311612.
2 Due to disclosure issues, data for 10 facilities with 100 < employment < 249, and 1 facility with 250 < employment
< 499 combined in lower category for NAICS 311613.
                                              B-9

-------
In addition to payroll and material costs, the Bureau provides capital expenditures and value added
directly attributed to the employment class level.
level:
'EPA used a additional assumptions to distribute industry-level costs to the employment class


 •       Employment benefits were assumed to be proportionate to payroll.
 •       Depreciation was assumed to be proportionate to capital expenditures.
 •       Rent payments were assumed to be proportionate to capital expenditures.
 •       Building repairs were assumed to be proportionate to capital expenditures.
 •       Equipment repairs were assumed to be proportionate to capital expenditures.
 •       Communications were assumed to be proportionate to the value of shipments.
 •       Legal services were assumed to be proportionate to the value of shipments.
 •       Accounting services were assumed to be proportionate to the value of shipments.
 •       Data processing services were assumed to be proportionate to the value of shipments.
 •       Advertising services were assumed to be proportionate to value added.
 •       Refuse removal was assumed to be proportionate to material costs
Using capital expenditures to distribute depreciation, rent, and repair costs to the employment class level
is based on the implicit assumption that capital expenditures are proportionate to capital stocks. For
example, expenditures on building repairs are presumably a function of buildings owned; because that
information is not available, EPA used an additional assumption that in general, capital stocks by
employment class are proportionate to capital expenditures by employment class.

       EPA thus calculated model facility EBIT as the average value of shipments (payroll, material
costs, benefits, depreciation, rent, and all specified purchased services) within each employment class.
Because revenues, payroll, and cost of materials are the most significant components of EBIT, the error
introduced by distributing industry-level data among employment classes should be small. Table B-3
presents Census data used to estimate EBIT at the employment class level. For NAICS 311613

                                             B-10

-------
V)
3
 ro
 95

 U
ho.
 05
 U
 \O
 CO
 o
  -2
   CB
     •s  a

     fc 0
a  -
S  05
S  o
Dollars

Millions
   cu


   R



   U
               ro
               ON
               CN

               CN
               VO
               V;
               cc
                   ~te
                   •Os
                      VO
                       vn
                       VO
                       ON
                       10
                       CN
                       VO
                       •6Q-
                          00
                          •6/3-
                          0\
                           CS
                           10
                          •ee-
                           CO
                           vo
:->
                               '
                               ON
                                  •&e
                                  fee
                                  (X)
                                  fee
e
B
                                      10
                                      -&e-
                                      fee-
                                         10
                                          tM
                                          VO
                                          00
                                          ON
                                          fee-
                                              tN
                                              CO
                                              vo-
                                              tN
                                                 >0
                                                 VO
                                                 fee
                                                  •
                                                     CN
                                                     fee
                                                     CO
                                                         tN
                                                             VO
                                                             fee-
                                                             fee
                                                              t>JO
                                                                0)
                                                                •oe
                                                                CN
                                                            bfi
                                                            'co

                                                            1
                                                            •a
                                                                    fee
                                                                    VO
|
p

I
                                                                        SR
                                                                        CN
                                                                        >0
                                                                         
-------
(rendering), payroll and material costs make up over 86 percent of estimated costs (where estimated
costs equal the sum of payroll, material costs, benefits, depreciation, rent, and purchased services). For
NAICS 311611 (slaughter), 311612 (processing), and 311615 (poultry), payroll and material costs exceed
90 percent of estimated costs.

       Table B-4 presents a sample calculation of average establishment EDIT by employment class
within each industry using these assumptions.  With few exceptions, EBIT increases monotonically with
establishment size. For animal slaughtering establishments (NAICS  311611) and poultry processors
(NAICS 311615), EBIT for the largest employment class is smaller than EBIT for many other classes.
This might indicate that some of these very large establishments are cost centers for larger business
establishments.
       B.2.2.2 Variance of EBIT by Employment Class

       Although the variance of revenues (value of shipments) is directly provided by the Census
special tabulation, the variance of EBIT needs to be estimated. EBIT is a linear function of its revenue
and cost components. Thus, the variance of EBIT can be estimated using the standard statistical
relationship where the variance of a linear function is itself a linear function of the variance and
covariance of its constituents.

       To estimate the distribution of EBIT for each model facility, EPA used the variance and
covariance of the value of shipments (R), payroll (P) and material costs (M) for each employment class
provided by Census.  Given that mean EBIT, lcE, for an employment class is:
                                      XE   XR    XP   XM
where xs denotes the mean value of revenues, R, payroll, P, and material costs, M. EPA computed the
variance of EBIT, oE2, as:
                        ฐE2 = ฐR2  ^  V  ''  ฐM2  ' 20RM - 2CRP  :  2ฐPM
                                             B-12

-------
CO
CO
CS
I
 I
a
 en
 C3








'"'+*
-'- rt
• *ง
' -O
/3
• ,2
1 W
tM

S3
•f
' -C


















/,i
H ซ
HH CO

W ฎ

00 ^-s
3 j.co
.S :CC
Jj/0

ปrt • ^
CO ^ ^5
CS " V r
E: J2
g
•- •ซ 2
ซ^ ^3 •;*•>
^H " -. r/)
- ^^^


:;f 1
!!ง ^
'o> J3
:|%
•S :,i|

o w S
Ti o 2
.a •ฃ
' •..:. .•.•"r-'
Vs1
--ง s

tฃL O
s ^
'•s'S-s
-si! S
S3 g
3 , co *•>
^ -
-M ^
S 8 '
ill
•jQ r^ ?
^(Sj -^ ^^
W cc 'S










- '




?"r
43
hi
3

•CO
1
.;;_>•
f
•"ซ
;^
j^
i—t
•--*•••<
en
a



.:'•?








n



in



r-









flN

V5-


r-
e<^





en



ง





o

SO
S



eN
x>


in
00
i — i








00
ON



0
<=H




C-'
r~



t-





ON
Q

en
n
fee-


0



en
S








ON
Os



OC
CS
•**




ON
m
SO



c-




ON
C
r~>

en
Os
3N
W


00
s


o
0
•6^-







(N
en
i— i
*


CS
ON
V*




OC
fe



3




ON
2
r-i

en
n
ฃe-


0
ง


CN
CN
fee-







n
Os
S
	 )
ฃf\






O
ON
ง


i
ON




49,473.0
^—
fee-

en
en




c
c
0

^
in
fee

•^j-
8
oo
fee-

in
so
Os
ON
en







o
en
•&e


r~;
CN
R
fee



44,054.0
o


cs




Cs
ฃ
1

OS
s
*n
fee-

•^•
en"
fee

m
I
oo







O
oc
oc


ON
^r
fee-



CD
o
^C
fee

en




Os
ON
O
|
'~H

oo
N


SO
X3
ON
t-T


^H
OO
o
en
r~

feO-





o
1
ON
fee-

oq
oo
fee-



O
el
rr

fee
ON



&
a
s
x
o














,-" T
;"
,CA

. Jป
fn

.
:l

...ซ
c





rt
ON-
2
8
in
                                                                                                                                            cn

-------


















(2
O
*i
g
1
"E<
a
WB


>•ป
>o
fl
5 ง*
g T-t

^f
ซ ซ
1 *
1
•2
"S
1
S
o
a
M
0)
>
 "*"*
Q ^

'urchased
Services'
thousands)
^H XJB.
-^^

u
terial Costs
thousands)
03
^5* ^5*
^

1 i
Sis ซ
&•! s
f5!
w is


"si
o 0
9 es
ฃM CO

> o

fiErf "***
^^ fj^,
N-*-

CM _
O A

Q >O n
JC a>
it
3 Jl
rfa ^ g
I=S CO *0
























ซ
'ง
"S
?

CO
VO
T— <
T— 4
CO
CO
s
-525






















o
f*^.
CO
fee-







p






,— 1
CO
fee-



m
ป



CO
o
oo
fee



oq
06
fee-







^









2
VO
CO

ON
fee-






in
in
fee-





VO
oo
in
fee-



o
o



00
CO
H
fee-


in
vd
CO







o
CO









o
T-H
ฃ^
*^J.
*-H
1— 1
fee





•^j-
co
i — r
fee^




OO
o
* — t
fee-



CO
g-



oo
00
CO



•^J.
p
fee-







9









o\
1— t
2
o
CM
ฃ~^
ON
CO
CO
fee-





ON
06
CO
CO





VO
VO
ON



in
ON
CO



-H
CO
fee-


co
ฃ
00
0
•€/5-






•— i
OO









2

T-H
C1*!
co
•^
^/5-





vq
r-H
VO





O
VO
i — i



OO
oo
vq.



co
CO
sf


in
ON
00
r— <
•&e






Q









OS
ON
2
o



















-


•*
•g
-3
ซJ
CO
8
*
•s
J-J
5
ฃ
in
*— i
vo
i— i
T— i
CO
03
1






















VO
CO

fee







OO
o





in
CO



•^j-
8
ป— ^
fee



•^j-
i



vo
ON
CO
fee-







$









S

r-H
"^J*
•G^-







VO
in





CO
S



T— 1
I



CO
CO
fee-


•^j-
n





S3
ts
B
5
S
o
0
in
co


S
3
'S
C3

. Q.| C
ft aS S c
Qj ^ —
!D CO rrri
<*> x 2 c
•S2 "" 'S. c
(O B i—5 ta *;
s s "a ฃ *
o 2. & o F
I-H O ^  , ,^j -c
S ซ 3- .s ง* |
ซS-&5 "g J
o 2 o <: g ^
s- 1 2 & ง g
ซ 1 1 * I !
U fli U W3 1-H '
D _o C O ฃ
^ "S r—t *^ S^ (
^ Ojl "13 4-ป 4-1
1 o^l S -| ]
i-t'st li
•งซซ!' ซs 2
g*— e ซ ซ t
a3 'fe S 2 "S'tE
ja .ซ ง -a M |

? S •ง "3 'g -j
co " ~; -ฐ s ซ
fll C "l r~* " C -^ *
.2 ^ t, b 
 P c t
" CC ^3 K (
CO •• (SO C ^> '
S m •ง 'M , i2 i
S co "a 3 &. f
| | J, cj | j
S •ซ ^ 13 -a .J
Sl^.J ง | }

C3 CJ ^ * >J^ ^
S"^ " *"^ '"^ (
^ ---—__ ^ •
^ป 5r* S •*"* ^ 'c
o o *^ c i - M k
g-^^g g ll
ง | | 1 ^ 8 t
•|''l ^ |f| 'j
& ^ " >, ซ g t
•^ *^ (ซ ri p^ c
'& '3 S -3 "ro io ฐ
-, >-i <" ft O ^J .S
sS t2 'S ซ 'cS a3 ^
2 2 ^ M ง *S (

co to > ^ w ra ^
F3 S co p co _H j






M^ "
3 'co'
? 1
3 to
2 "ra
1
I 8-
3 ซ.
3 S
1 Q^
> C
\ 'S
) O
3 ^

ture (assumed
'r*i
ซ c
^ CO
-U d,
><
1 w
s 1
&
J (J
J -s
3 - &

-t -^
3 g
3 o3
i ฐ-
J .1ฐ
_, ^^
"I ^^
:^ g co
if ST C/3
-> ^^4 CQ
Lf |
f 2 ^
i| o O
^ o "a.

•j *3 o
ij* p* _
^  02 "-M
5 fc-i .52
•' 

  • -------
    where O;2 and ay represent the variance and covariance of revenues, payroll, and material costs,
    respectively (Mendenhall et al., 1990). Although; payroll and material cost do not comprise all operating
    expenses included in EBIT, they do comprise the vast majority of EBIT. Hence, excluding the variance
    for the remaining components should not cause a significant error in the variance estimate.
           B.2.3  Facility Net Income
    
           B.2.3.1 Average Facility Net Income by Employment Class
    
           EPA calculated net income for each employment class model facility in each industry from
    EBIT, using additional assumptions to estimate tax and interest payments. Data for these two additional
    components of net income were derived from two Census Bureau publications, Annual Survey of
    Manufactures (ASM) and Economic Census, along with the Internal Revenue Service code. Because one
    must use an additional layer of assumptions to estimate net income from EBIT, the uncertainty associated
    with the net income estimate is greater than that for EBIT.
    
           Estimating tax payments is relatively straightforward. EPA assumed that establishment EBIT is
    equal to business entity EBIT as the basis for calculating taxes. To estimate facility tax payments, EPA
    multiplied the model facility's EBIT by the sum of the relevant federal corporate income tax rate and the
    average state corporate income tax. To estimate net income, EPA subtracted the estimated tax payment
    from EBIT for each model facility.
    
           EPA estimated interest payments using a combination of ASM data on past investment by
    industry, Census data on relative investment in buildings and equipment, and assumptions about
    investment behavior. EPA first scaled ASM time series data on industry investment, which is based on
    Standard Industrial Classification (SIC) codes, to represent the current NAICS meat product industries.
    EPA then used the average percentages of meat product industry investment in equipment and structures,
    as presented in the Economic Census, to divide the ASM investment time series into those two
    components.
                                                 B-15
    

    -------
           In estimating interest payments from the time series of past investment in equipment and .
    structures, EPA made a series of assumptions concerning industry borrowing behavior. EPA assumed
    that:
    
           •      All investment in each year was funded through bank loans.
           •      The interest rate on those loans was equal to the nominal prime rate for that year plus 1
                  percent. (Since ASM investment time series data is in nominal terms; a nominal interest
                  rate is appropriate.)
           •      The average loan period was 7 years for equipment and 25 years for structures.
    
    Using these assumptions, EPA developed a time series estimate of loan payments made by the industry,
    and the portion of each year's loan payments accounted for by interest (e.g., using the Lotus @EPAYMT
    function). Total interest payments in the baseline year equals the sum of this year's interest payments on
    the stream of past years' investment.4  Interest payments were then attributed to each employment class
    based on the percentage of industry investment accounted for by that employment class in the 1997
    Census. Table B-5 presents a sample calculation of establishment net income by industry and
    employment class using the methods described above for attributing tax and interest payments to
    employment classes.
           B.2.3.2 Distribution of Net Income Within Employment Class
    
           EPA also estimated the variance of net income for each model facility from its estimated
    variance forEBIT. If the mean of a distribution is multiplied by some scalar a, then the variance of that
    distribution will change by the square of a. That is, if the mean net income for a model facility is some
    percentage of facility EDIT (XNI = crxE), then the variance of facility net income is equal to the square of
    that percentage multiplied by the variance of EBIT (o2NI = a2a2^). EPA used the ratio of facility net
    income to EBIT to determine the scalar for estimating the variance of net income (adjustments to
    variance are discussed in more detail in Section B.4.3). The estimated mean  and variance for net income
    in each employment class by NAICS code is presented in Table B-2.
            4 For example, interest payments on equipment investment for the year 1997 would equal the sum of interest
    paid in year 25 of loans from 1973 plus the interest paid in year 24 of loans from 1974, and so on.
                                                 B-16
    

    -------
    
    
    
    
    
    •*•*
    ซ
    r9
    1
    1
    S
    ft
    ฃ
    fa
    1
    O
    "S
    03
    :i
    j
    ^
    l
    13
    M
    w
    •a
    -2
    
    .3
    H
    
    -
    •s
    .]•
    •ซ.
    a
    ฃ
    |
    1
    
    
    *1
    tg S
    &•< ง0
    11
    u ^.
    v-'
    
    g-I
    O 03
    5-g
    5 ^
    v-'
    
    *i
    S ซ
    ฃ "^
    QJ 9
    C3 ฐ
    ซ9-
    
    
    T3
    *tn od
    a> BO
    H J
    
    
    
    %
    &s
    2 i
    [V] C
    i
    "s -2
    n
    s
    C^ c/
    ง*^^ fll
    >,
    *•&
    i
    T^ปซ n..
    
    
    
    
    
    
    
    
    
    
    
    
    
    hf)
    .a
    s
    I
    Js
    C/J
    ฃ
    ^u
    •*-ป
    1
    &
    I
    ซ!
    r!4
    v.n
    T-H
    n
    ^
    
    
    
    
    
    
    
    
    
    o
    CO
    fee
    
    
    
    
    
    r-
    ie^
    
    
    
    
    (-ป
    CO
    
    
    
    
    1 — 1
    00
    
    
    
    
    VO
    fee
    
    
    in
    
    
    
    
    
    |
    '~H
    
    T-H
    
    
    
    
    
    o
    CO
    
    
    
    
    00
    VO
    fee
    
    
    
    
    oo
    fee-
    
    
    
    
    vo
    fee-
    
    
    S
    
    
    
    
    
    ON
    
    vo
    ฃe
    
    
    
    
    
    in
    ON
    
    
    
    
    m
    VO
    T-H
    
    
    
    
    CM
    8
    
    
    
    
    CO
    o
    **
    
    oi
    
    
    
    
    ON
    2
    f-1
    
    
    m
    in
    CO
    fee
    
    
    
    
    •^j-
    CO
    w-
    
    
    
    CO
    CO
    
    
    
    
    vo
    CO
    
    
    
    
    CO
    GN
    ON
    VO
    fee
    
    T-H
    
    
    
    
    
    Q
    
    
    m
    CO
    CO
    ,_H
    fee
    
    
    
    •ซ3-
    o
    T— H
    CM
    i — l
    
    
    
    CM
    8
    
    
    
    
    r-
    o
    S
    
    
    
    CO
    CM
    CO
    CM
    (ft
    
    ON
    r-
    
    
    
    
    C5N
    ON
    2
    (— )
    in
    CO
    0
    i — i
    0
    cO
    (ft
    
    
    
    in
    o
    CM
    
    
    
    CO
    ง
    (ft
    
    
    
    
    O
    s
    fee^
    
    
    
    00
    i
    m
    (ft
    
    S
    
    
    
    
    CM
    Q
    1 — '
    
    OO
    IT--
    CO
    •^f.
    (ft
    
    
    
    T-H
    CM
    CO
    (ft
    
    
    
    •*$•
    s
    
    
    
    
    o
    00
    s
    CO
    fee
    
    
    
    in
    g
    fe
    
    CO
    CO
    
    
    
    
    ON
    ON
    ^1-
    Q
    in
    CM
    ON
    3
    VO
    fee
    
    
    
    in
    oo
    fee-
    
    
    
    in
    00
    oo
    iff
    
    
    
    •c^-
    in
    S
    
    
    
    T}-
    VO
    m
    T-H
    •Ge-
    i — i
    CM
    
    
    
    
    ON
    ON
    ON
    .S
    R
    in
    i — i
    oo
    o
    i — i
    T^-
    CO
    fee
    
    
    r-
    VO
    CM
    fee-
    
    
    in
    S
    fee
    
    
    
    r-
    t~-
    i — i
    CM
    fee
    
    
    
    ON
    2
    CM
    in
    
    ON
    CO
    
    
    ON
    O\
    c-f
    O
    o
    T-H
    oo
    in
    oo
    T-H
    t--
    
    •&e
    
    
    CO
    CM
    ON
    -fee-
    
    
    
    o
    ON
    vo
    n
    VO
    fee-
    
    
    
    CM
    CM
    T-H
    T-H
    fee-
    
    
    
    -3-
    oo
    s
    CM
    -&e
    ON
    
    a3
    *-j
    S
    a
    o
    CM
    
    
    
    
    
    
    
    
    
    
    • -.'
    
    
    
    !i
    M
    ;~~
    -1
    '"?
    
    
    *
    J
    CM
    -T-H
    VO
    T-H
    
    5^
    '•%
    
    
    . ,-„
    
    
    
    T;
    
    
    CO
    fee
    
    
    
    
    
    •<*
    -ฃซ-
    
    
    
    
    O
    OO'
    .fee-
    
    
    
    
    oo
    ON
    fee-
    
    
    
    
    n
    (ft
    
    
    
    vo
    ON
    oo
    feS
    
    
    
    CO
    fee-
    
    
    
    
    vo
    ON
    .59
    CO
    •&e
    
    
    
    Tt-
    in
    CO
    00
    fee
    
    3
    
    
    
    
    S
    Q
    *""*
    OO
    r-
    T-H
    o
    [--
    fee-
    
    
    
    ON
    ON
    m
    00
    
    -------
    PQ
    

    -------
           Note that the link between impacts measured by comparing net income with compliance costs is
    much stronger than the link between revenues and compliance costs, although not stronger than the link
    between cash flow and compliance costs. However, because the estimate of net income is dependent
    upon a series  of assumptions, the uncertainty concerning the accuracy of the net income measure is
    greater than for revenues. Thus, this analytic approach represents a tradeoff between the accuracy of the
    income measure and the certainty of the impacts based on that measure.
           B.2.4  FacUity Cash Flow
    
           Cash flow is calculated as net income plus depreciation^ Depreciation was estimated for the
    calculation of model establishment EBIT as described in section B.2.2.1 above. Estimated model facility
    cash flow is presented in Table B-5 along with net income estimates.
    
           The distribution for estimated cash' flow has an identical variance to net income, but a larger
    mean because depreciation is added to the mean of net income. The probability that cash flow is less
    than zero tends to be about 3 percent to 5 percent smaller than the probability that net income is less than
    zero.
    
           Cash flow is the preferred method in financial management to evaluate investments (FASB,
    1996; Brealey and Meyers, 1996; Brigham and Gapenski,  1997). When post-compliance cash flow is
    negative, the facility can be reasonably projected to close. This is the basis of the closure model (see
    Section 3.1.2 for more detail). Once again, however, given the additional assumption required to
    estimate cash flow .from net income, there is a tradeoff between the level of certainty regrading impacts
    and the precision of the income measure.
    
           EPA uses cash flow to estimate the number of potential facility closures and related employment
    impacts from the effluent guidelines by comparing posttax annualized compliance costs and cash flow.
    Cash flow is also used to calculate the number of facilities with compliance costs greater than 3 percent,
    5 percent, or 10 percent of revenues.
                                                 B-19
    

    -------
    B.3    SUBCATEGORIZATION, DISCHARGE TYPE, AND FACILITY SIZE
    
           B.3.1   Basis for Subcategorization
    
           To develop the engineering models used for estimating compliance costs and pollutant load
    reductions, EPA classified meat products industry based on the type of meat produced at the facility:
    
           •       Red meat (primarily beef and pork)
           •       Poultry (primarily chicken and turkey)
           •       Mixed (both red meat and poultry)
           •       Rendering products or meat byproducts (either red meat or poultry)
    
    and the type of processes performed at the facility:
    
           •       First processing (slaughter)
           •       Further processing
           •       Rendering (the process resulting in meat byproducts)
    
    The meat type and process classes resulting from this classification consist of combinations of the
    processes for each meat type. For example, a poultry facility may perform any of the following six
    combinations of processes, each one of which will place it in a different subcategory: (1) first
    processing, (2) further processing, (3) first and further processing, (4) first processing and rendering, (5)
    further processing and rendering, or (6) first processing, further processing, and rendering. Facilities that
    only perform rendering are subcategorized as Tenderers; facilities that perform rendering in combination
    with the other two processes are subcategorized with the appropriate meat type (red meat or poultry). As
    an empirical matter, EPA found that all affected facilities that process both red meat and poultry
    ("mixed" facilities) were found to perform only further processing or further processing and rendering
    activities.
           EPA also classified facilities by discharge type and facility size. Discharge type distinguishes
    those facilities that discharge process wastewater directly into U.S. surface waters (direct dischargers)
                                                 B-20
    

    -------
    from those that discharge wastewater to treatment works (indirect dischargers). Under the Clean Water
    Act, EPA may apply different standards to direct and indirect dischargers (see Section 1.1). Size, as
    determined by facility production and wastewater flow, was used to cost the appropriate treatment
    capacity for the facility. For the purposes of costing, EPA divided facilities in each subcategory into
    small, medium, large, and very large. Detailed information on subcategorization can be found in the
    Development Document (EPA, 2002).
            B.3.2  Matching Economic Model Facilities With Engineering Model Facilities
    
            In order to perform the economic impact analysis, EPA matched its economic model facilities to
    the engineering model facilities used to estimate costs. This matching was performed on the basis of two
    characteristics: (1) the relationship between production process and NAICS industry and (2) the
    relationship between production and revenues.
    
            The Census Bureau classifies the meat product industry into four groups. All red meat facilities
    that perform animal slaughter (first processing), whether alone or in combination with other processes,
    fall into NAICS 311611. All red meat facilities that perform further processing (with or without
    rendering), but no slaughtering activities, are classified as belonging to NAICS 311612. Facilities
    performing poultry slaughter, poultry further processing, or both (with or without rendering), are
    contained in NAICS 311615. Finally, facilities that perform rendering, but no other processing activities,
    are classified in NAICS  311613.
    
            Thus, model economic facilities were matched to the model engineering facilities, based on
    production, as follows:
                   Red meat facilities — whether beef or pork — that perform first processing alone or in
                   combination with further processing and/or rendering were assigned an economic model
                   facility from NAICS 311611.
                   Red meat facilities — whether beef or. pork — that perform further processing alone or
                   in combination with rendering, but no first processing, were assigned an economic
                   model facility from NAICS 311612.
                                                  B-21
    

    -------
           •       Poultry facilities — whether chicken or turkey — that perform either first processing or
                   further processing, alone or in combination with other processes, were assigned an
                   economic model facility from NAICS 311615.
          • •       Facilities that perform rendering — whether red meat or poultry — but no other
                   processes were assigned an economic model facility from NAICS 311613.
           •       Mixed facilities — both red meat and poultry — perform further processing only and
                   were assigned an economic model facility from NAICS 311612.
    
    All model engineering facilities were assigned an economic model from one NAICS code only.
    
           The economic model facilities were developed from data classified by employment size, while
    engineering cost models were sized by production and flow (for details see Development Document,
    U.S. EPA, 2002). EPA classified engineering models into small, medium, large, or very large based' on
    examination of production and flow characteristics of facilities contained in the screener survey
    database. EPA then determined the appropriate size for each engineering cost model facility, and
    assigned each facility to a size class within a meat type and process class. To match the economic model
    facilities with the engineering model facilities, EPA calculated the median production for all facilities in
    that class. EPA then combined median production data for the engineering model facilities with meat
    product indicator prices to estimate revenues for each engineering model facility. These estimated
    revenues were then compared with each economic  model facility's average revenues, and the model
    facility with the closest match was selected to represent the economic characteristics of that engineering
    facility.
           EPA used the baseline prices from the market model as the indicator prices for the meat products
    (for more detail on the market model see Section 3.1.4.2). The baseline prices are estimated for the four
    meat types: beef, pork, chicken, and turkey.  The engineering model facilities are categorized on the
    basis of: red meat, poultry, and rendering. To account for this, EPA calculated revenues twice for each
    engineering model facility using the prices of two meat types. For example, EPA estimated revenues for
    red meat facilities first using the price of beef, then using the price of pork. Similarly, EPA calculated
    revenues for poultry using the price of chicken as well as the price of turkey. This resulted in a range of
    revenues for each model class to be compared with economic model facility revenues. For mixed meats,
    EPA used production for each of the four meat types as a percentage of total model class production as
    calculated from screener survey data. These percentages were multiplied by the price for each meat type
    
                                                  B-22
    

    -------
    in order to calculate model facility revenues as a weighted average. There were a few instances where the
    range of revenues complicated the assignment of facilities. In such cases, EPA assigned the engineering
    model facility to the economic model facility whose revenues were closest to both measures of estimated
    revenues.
    
           Table B-6 presents each subcategory and facility size for which engineering models were
    developed, as well as the economic model EPA assigned to each size for the purpose of projecting
    impacts. For example, based on its examination of the screener survey database, EPA estimated that
    median production for the 28 indirect discharging facilities that perform a combination of first and
    further processing of red meat was 196 million pounds. After examining these facilities' production'and
    flow characteristics, EPA determined that they were medium-sized producers for the purposes of
    costing. The production data was multiplied by  the price indicators and this resulted in a range of
    estimated revenues from $197,000 to $218,000.  Based on this, EPA assigned these 28 facilities an
    economic model facility from the 500 to 999 employee class in NAICS 311611 which has model facility
    revenues of $262,700, the closest match.
    B.4    NEGATIVE BASELINE FACILITY INCOME
    
           Estimating the means and variances for the distribution of each model facility's income results in
    some probability greater than zero that facilities in each employment class earn negative income. Table
    B-7 presents the model facility mean and standard deviation for each income measure by employment
    class and NAICS code, as well as the probability that income is less than zero (based on that mean and
    standard deviation, and assuming income is normally distributed). This section discusses the reasons
    why model facilities might have negative income,  as well as those reasons' implications for the model.
            B.4.1   Actual Establishment Income Is Less Than Zero
    
            Two possible reasons for negative establishment baseline income are attributable to the actual
    establishment financial data (collected by the Census Bureau) on which the estimated distribution is
    based:
                                                 B-23
    

    -------
    
    
    
    
    
    
    
    
    
    5
    a
    3
    (S
    •3
    i
    u
    'ง
    0
    a
    
    TJ
    S
    *9 '8
    M <
    ซ ซ
    S a
    •3 ซ
    T
    ubcategories
    oo
    TS
    O
    S
    6UD
    B
    
    U
    o>
    n
    f
    W
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    'ฃป
    3
    MM
    a>
    o
    US
    
    
    
    
    i'- *
    
    
    
    
    t
    e
    1
    s
    5
    
    
    
    
    
    
    :H
    i
    S
    J
    
    
    
    
    
    
    
    
    S
    r-T
    **
    ^
    O
    a
    •a
    i5
    1
    
    •3
    
    
    
    1
    •ce
    " Hi
    ^
    
    ^
    S - B
    s-n
    3-9
    ZTJ
    - g
    
    o-
    u
    a
    4=
    S
    2:
    JO-^
    •4 -
    1,
    3 *
    Si
    •H
    a
    
    "5L
    1
    1
    bk
    
    B
    1
    a
    i— i
    
    A
    ft
    S
    
    
    
    Revenues
    
    •i
    
    1
    •u
    
    S~ S
    w
    ~ ••••
    >^
    Q
    O
    T1
    Ci
    .8
    ••a
    
    ซ
    
    
    rS
    rir>
    
    1
    • fe
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    'irst Processing
    
    Red Mea
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    01
    
    
    
    
    en
    en
    
    
    
    oo
    04
    
    
    
    
    I
    
    
    
    
    1
    
    2
    ,
    VO
    en
    
    o
    
    in
    
    
    
    ง
    
    
    
    
    
    
    
    o:
    E
    
    in
    "*
    
    
    oo
    3
    CO
    
    
    VO
    of
    
    
    
    
    ON
    8
    
    
    
    ON
    VO
    •be-
    
    1
    t
    VC
    CO
    
    .^j.
    Vi?
    vo
    V,
    
    
    VO
    
    
    
    
    
    
    S
    1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    hh
    ซ
    irther Processil
    rr.
    Red Meat!
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    01
    
    
    
    
    I
    
    
    
    S
    
    
    
    
    •be
    
    
    
    
    en
    i — i
    
    2
    04
    VC
    en
    
    
    
    en
    
    
    
    in
    ol
    
    
    
    
    
    
    1
    en
    in
    en
    
    
    in
    in
    in
    i —
    
    
    8
    en,
    •be-
    
    
    
    g
    oo
    •be-
    
    
    
    VO
    S
    i
    be-
    
    ON
    1
    in
    O4
    04
    VO
    en
    
    
    
    
    rs
    i-H
    
    O
    
    
    
    
    
    
    g
    1
    on
    en
    f-
    
    
    oo
    vn
    
    
    vo
    ^ — i
    
    
    
    ง
    oo
    04
    •be
    
    
    
    ON
    
    ON
    ON
    ON
    O
    O
    in.
    04
    VC
    en
    
    
    -------
       I
       "3
       -a
        o
       13
    S  OB
    ง -a
    VO
        fl
        08
        BD
    
        ซ
        03
        O
        s
    
    
    
        {
        OK
        a
    
    
    
    
    
    
    > ~
    
    
    
    
    ฃ?
    3
    !i
    
    
    
    
    *
    
    
    *
    *r
    
    i
    %
    
    *
    &
    
    
    
    -
    
    o"
    s
    
    >ฃ<
    a>
    a
    1
    A
    1
    TV*
    •3
    o
    
    ,1
    |
    1*
    V
    1"
    *s
    2
    
    
    2 s
    ci.
    ~ {
    c
    i
    1
    S-
    3>1
    i <
    1
    "3.
    |j
    ^
    OS
    U
    1
    
    3
    M"(
    •s
    S5
    H
    1
    
    •
    0!
    j
    >
    tJ
    )
    ^
    > -0
    1 ^
    
    2*
    5 <•!
    1*
    i!
    "
    H
    ~~"yt
    •* P*
    l|
    *" r^i
    -
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Processing
    ^S
    !
    OH
    
    
    
    
    
    
    
    
    
    
    VO
    
    ง
    
    
    
    8
    
    
    
    
    
    ^-
    "
    
    
    ir
    
    ON
    in
    
    
    vn
    cn
    cn
    
    
    ON
    
    V
    n
    n
    
    S
    vo
    
    
    
    vo
    cn
    •se-
    
    
    
    en
    oc
    
    
    ฃ
    
    ON
    ON
    3
    in
    r-
    
    
    vn
    en
    vo
    cn
    oo
    ^j.
    ON
    ON
    i
    i
    <
    O
    cn^
    
    ฃ
    fir
    
    
    g
    
    
    
    
    
    
    
    
    •.-"
    
    
    
    
    
    
    
    'fif)
    • -v.
    urtherProci
    w
    -;C
    1
    fe
    nฐ
    
    
    
    
    
    
    
    
    •':,
    CN
    
    00
    
    
    
    ฃ
    
    
    
    
    
    "3-
    &
    
    
    oo
    a
    
    i
    
    
    VO
    en
    S
    
    
    a
    
    ' a
    C/
    n
    
    i
    s
    
    
    1
    en
    •be-
    
    
    
    
    oo^
    
    
    o
    ฃ
    
    250-499
    
    
    vo
    cn
    ^
    en
    
    
    S
    . (
    S
    cn
    
    ฃ
    a
    
    
    s
    1 —
    -&0-
    
    
    
    •-I
    ON
    CN
    en
    •fee
    
    
    ON
    tr-
    cs
    oc
    
    ON
    CN
    
    
    vo
    en
    ON
    VO
    C-
    vo
    
    g
    s
    cn
    
    3N
    ie-
    
    
    i
    ~-ซ
    i— <
    
    
    
    --H
    ON
    CN
    
    
    V
    OC
    
    1000-2499
    
    
    vo
    cn
    t~
    >— i
    
    0
    <
    c
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    bfi
    C
    1
    •'•W
    "bi
    V.
    1
    -:S
    'O<
    
    
    
    
    
    
    
    
    CN
    
    OO
    •&e-
    
    
    
    ฃ
    
    
    
    
    
    •*
    &e-
    
    
    OO
    cs
    
    3
    
    
    VO
    cn
    o
    
    
    o
    
    c
    c/:
    VD
    
    g
    S
    
    
    •n
    X
    •fee
    
    
    
    ^
    Iฐr
    
    
    ON
    ON
    ON
    ON
    S
    
    1
    
    
    vo
    cn
    en
    C"
    CN
    
    ON
    G
    HE
    t
    &
    cn_
    
    S:
    •sj
    
    
    ^j*
    * —
    i— i
    
    
    
    i— (
    ON
    CN
    en
    
    
    ON
    r-
    in
    
    ON
    CN
    
    
    VD
    cn
    o
    V
    •—I
    CN
    cn
    O
    
    c
    cn^
    
    7s
    e
    
    
    S
    1 —
    Se-
    
    
    
    T-H
    1
    
    
    ON
    u-
    cs
    oc
    
    ON
    CN
    i
    
    
    VO
    cn
    c-
    o
    oo
    w
    cn
    |
    t
    
    
    
    
    
    
    
    
    
    
    -
    
    -
    M
    a
    • }_{
    -"i
    . cc
    --.'W
    ":.ฃ
    V
    :--t
    . o
    ryEurther;Pr
    •-iฑ2
    nฐ
    
    '.. >
    .';TJ-
    
    "'•"•'
    
    
    t;'.:
    cn
    
    OO
    
    
    
    n
    
    
    
    
    
    cs
    r^J
    i
    
    
    CN
    cn
    V-t
    
    ON
    6
    
    m
    vo
    cn
    T— I
    in
    
    
    
    "ca
    cr
    ฃ
    
    CN
    S
    
    
    vo
    vo
    Tt-
    m
    ฑH-
    
    
    
    CN
    OO
    c:
    
    \ 	
    O
    5
    
    250-499
    
    in
    vo
    cn
    r-
    VD
    r~-
    ON
    fi
    C
    en^
    
    3N
    ^
    
    
    in
    S
    r*-
    5
    
    
    
    i— (
    ON
    CN
    cn
    
    
    ON
    in
    
    ON
    
    m
    vo
    cn
    oo
    ON
    in
    ON
    
    vo
    
    t.
    a
                                                                                                                                   m
                                                                                                                                   CN
    

    -------
     VO
    • CN
    

    -------
       -Q
        C3
       JS
        O
        Ui
        a
        UD
    CO -g
     ซ  g
    
        I
       i
       ON
       "3
       &
    
       •1
        o
    
    - o
    * 1
    i
    i
    ซ
    •t
    .g
    - i
    &
    -
    '
    
    - s-
    •56-
    - '.J
    4>
    > Q
    ;-2
    •J
    r:fCC
    •_
    
    ..".,'*'"
    , s
    *•ง
    "*:ฃ,
    2
    i
    i
    ซ
    '1
    i— i
    :'l
    &
    JS fe
    "1 _ง
    o fa
    a
    •*i
    h
    
    CO
    . , - 4,
    ^ A
    • i
    . ฃ
    -a
    i
    n
    * a
    ,, ^ g
    - 1
    
    1
    ซ
    _ w
    ง
    / I
    5
    tt
    . 1
    ~ ~" ~" ฃฃ<
    .>- ;-g
    ซ
    -•'1
    - ฃ
    i 4
    :„ .jzj
    ' , ee
    EC
    2
    _!
    — ~ci
    ซ•ซ-
    3* c3
    "J v| i
    Ll|5
    -2 & S
    diJ
    
    
    
    
    
    
    
    
    
    
    in
    _c
    •J3
    •61
    3
    1
    /is
    1
    A<
    I
    fl
    ^cS
    g
    ^
    T— t
    T— 1
    VO
    i-H
    T— I
    m
    E73
    1
    
    
    
    
    Eฃ
    i— i
    oo
    04
    *
    r— t
    T— (
    cn
    *
    1—1
    i— i
    CO
    *
    vo
    VO
    
    VO
    in
    
    VO
    in
    
    CN
    oo
    
    CN
    a\
    04
    
    cn
    m
    ee
    
    00
    ol
    ee-
    •*
    ee
    •*
    •*
    ee
    rt
    o
    5ฃ
    00
    VO
    01
    *
    01
    O
    en
    *
    l
    
    CN
    r~
    cn
    
    VO
    VO
    r-
    T— i
    VO
    00
    fft
    
    vS
    &e-
    CN
    VC
    ^
    o
    c^l
    oo
    cn
    6O-
    VO
    cn
    cn
    •&"3-
    cn
    >n
    r-
    V*
    cn
    5
    00
    (ft
    ON
    •*
    O
    o
    CN
    ^
    CN
    VO
    CN
    ^
    CN
    00
    CN
    {ฃ
    CN
    OO
    CN
    #
    VO
    VO
    
    O
    VO
    CN
    CN
    O
    VO
    CN
    CN
    '
    5:
    cn
    ^J-
    TT
    vo
    c\
    •<*
    T— (
    00
    cn
    •*
    1—1
    &t
    s
    cn^
    i — i
    •se-
    en
    o
    >n
    of
    ffr
    O
    ON
    •*
    CN
    CN
    ae-
    o\
    ON
    2
    o
    in
    ^
    r-
    VO
    CN
    ^
    CN
    O
    cn
    ^
    •CN
    O
    cn
    *
    vo
    vo
    
    T— t
    1 	 1
    CN
    v>
    * — t
    i— i
    CN
    >O
    ,
    SN
    ON
    T}-
    0
    t-
    CN
    (N
    vo
    ^r
    00
    •*
    cs
    cn
    ซซ•
    VO
    ON
    VC
    CS
    •ae-
    ON
    9
    m
    ff,
    ^
    ON
    VO
    v$
    ON
    ^f
    CN
    O
    c
    *
    oo
    t—
    CN
    *
    ON
    O
    cn
    ^
    ON
    O
    cn
    #
    VO
    VO
    
    "^-
    CN
    O
    oo
    •*
    g
    00
    c^
    r-
    m
    i — i
    ON
    VD
    O
    S
    2;
    t—
    -*
    6O-
    V
    • c
    •*
    fft
    ir
    oc
    r~
    69-
    ^r
    CT
    VC
    V.
    ON
    ON
    Tf
    O
    V
    CN
    ^
    cn
    t~,
    CN
    *
    VO
    T-H
    cn
    *
    VO
    i — i
    cn
    *
    VO
    VO
    
    cn
    o
    •*
    o
    i — i
    cn
    ?
    0
    T 	 (
    VO
    T— H
    in
    cn
    ON
    1—4
    OO
    •*
    r-
    i — i
    •3-
    g
    VO
    •69-
    cn
    oo
    ON
    •*
    •ee-
    o
    CN
    OO
    vo"
    1 — 1
    •&e-
    ^~
    cn
    ^
    CN
    VO
    CN
    &$
    ON
    ON
    ON
    O
    i
    w-i
    ^
    VD
    VO
    CN
    fl*
    ?CN
    ON •
    CN
    * .
    CN
    -ON
    CN
    #
    VO
    VO
    
    CN
    vo
    VO
    cn
    >n
    CN
    vo
    VO
    .cn
    >n
    •
    oo
    VO
    VO
    CN
    O
    i— i
    in
    ON
    O
    >n
    ^t-
    ON
    oo
    •*
    cn
    cn
    fee-
    l—I
    CN
    **!
    ON
    CN
    ffi
    OO
    ON
    O
    vo"
    •n
    60-
    oo
    •*
    ON
    t~
    5o-
    *e-
    2
    II
    ~* c<
    ^
    CN
    OO
    CN
    *
    OO
    r-
    cn
    *
    OO
    p-
    cn
    ^
    VO
    VO
    
    oo
    oo
    ON
    1 — 1
    cn
    oo
    '00
    ON
    i— I
    cn
    O
    r-
    t—
    cn
    ON
    CN
    r-
    oo
    00
    S
    \r
    •oo
    &e
    •*
    cn
    ON
    ON
    &e
    ON
    ON
    CN
    •ae-
    "*
    >n
    VC
    CN
    **
    5
    if
    C-
    Al
    
    
    
    
    
    
    
    
    
    
    
    W
    fl
    &
    i
    ts
    &
    fซ5
    CN
    i — i
    •EVO
    ~T— H
    i^H
    V
    <,
    z
    -
    
    ---
    
    ^
    1— 1
    1 — 1
    cn
    ^
    oo
    >n
    cn
    ^
    oo
    n
    ^
    ON
    00
    CN
    jฃ
    T — i
    cn
    cn
    *
    i — i
    cn
    cn
    *
    ON
    cn
    
    r-
    VO
    cn
    
    t-
    vo
    cn
    
    f-
    cn
    0
    T— 4
    VO
    S
    01
    ^r
    o:
    
    VC
    &
    CN
    n
    ee
    O
    IT]
    rl-"
    ee-
    S
    ON
    OC
    ee-
    
    -------
    I
    N
    e)
    S
    o
    
    s
    &
    5
    1
    |
    |
    p
    ฃ
    ป .
    
    i
    1
    :/ 1
    CO
    
    
    
    i
    ' 	 OJ
    OJ
    ncome IV
    
    —:•--'-
    i
    & *
    fl|
    \J fc*
    O
    w
    
    1
    1
    1
    Cash Flow
    I
    u
    a
    r
    o
    w
    1
    i
    - i
    • - - i
    ""-I
    41
    • •-- ' ฃ
    •W
    
    1
    tt
    S
    ~~"~~""pi
    1 Establishment
    Employment
    Size Class
    
    
    
    
    
    
    
    
    
    
    Ml
    'S
    C
    cc
    vo
    
    OT
    
    
    
    
    
    
    
    
    
    ON
    5
    CS
    oo
    CS
    oo
    oo
    CS
    cs
    en
    en
    
    vo
    
    in
    u-i
    ป— i
    o
    
    —
    
    ON
    
    VC
    00
    
    0
    VO
    en
    CS
    O
    vo
    CS
    *
    O
    VO
    cs
    oo
    cs
    CS
    ON
    r-
    S
    
    r-
    ON
    in
    
    oo
    CN
    in
    cs
    r-
    m
    
    O
    
    vc
    c
    if.
    oo
    oc
    cr
    ON
    o
    in
    en
    cs
    S
    oo
    cs
    s
    oo
    cs
    oo
    s
    i
    T— 1
    i
    T— I
    VO
    CS
    T— <
    CS
    VO
    oo
    o
    en
    r—
    
    00
    VC
    
    in
    er
    CS
    ฃ•
    VO
    VC
    ON
    o
    e
    S
    ?
    oo
    55
    oo
    cs"
    oo
    cs
    cs
    ON
    en"
    CN
    ON
    en
    CS
    S
    VO
    oo
    oo
    VO
    ป— '
    cs
    c<
    ON
    r—
    oo
    5=
    \r
    f
    oc
    vo
    -5
    20 to 49
    VO
    ?i
    in
    ON
    cs
    in
    ON
    CSI
    oo
    CS
    CN
    VO
    5*.
    VO
    
    in
    vo
    oo
    VO
    cs
    ON
    VO
    o
    en
    vo
    CN
    CS
    m
    -6O-
    OO
    o
    r-
    cr\
    ON
    0
    C
    ซj
    
    n
    
    g
    
    <;
    
    *
    g
    
    ฃ
    
    ฃ
    
    ฃ
    
    &
    
    n
    
    1
    
    CS
    o
    C
    ซj
    
    n
    
    ll
    
    n
    oo
    cs
    oo
    cs
    
    00
    in
    
    oo
    T-H
    
    oo
    
    r-
    
    S
    
    00
    cs
    
    0
    en
    oo
    cs
    o
    C--
    en
    O
    r-
    en
    ^
    m
    o
    o
    r-
    
    en
    en
    i — i
    
    m
    vo
    
    O
    
    en
    CS
    •60-
    
    &>
    
    in
    WJ-
    0
    •n
    r— t
    CS
    r-
    en
    cs
    r-
    en
    cs
    ^
    in
    en
    vo
    i — i
    en
    vo
    
    in
    oo
    CS
    i-H
    T— t
    0
    CS
    oo
    
    en
    in
    •ae-
    c--;
    0-
    
    cs
    o-
    en
    ON
    O
    C
    VO
    1 — 1
    CS
    ON
    cs
    cs
    ON
    cs
    cs
    sฃ
    in
    VO
    vo
    en
    vo
    VO
    cs
    en
    VO
    oo
    in
    oo
    T 	 1
    r~
    Tf
    vo
    in
    cs
    00
    5?
    CS
    ON
    VO
    en
    40-
    r— 4
    cs
    T 	 1
    T-H
    || 20 to 49
    *
    en
    en
    cs
    in
    in
    in
    m
    cs
    ^,
    in
    in
    cs
    cs
    cs"
    in
    cs
    cs
    cs
    ง
    cs
    
    o
    cs
    ON
    oo
    VC
    &
    en
    VC
    **
    r-
    VO
    r—
    
                                                                                      3 !>
                                                                                          0
                                                                                     •
                                                                                       .SJ .'S3
    os
                                                                                       ^•s
                                                                                        O C
                                                                                     ฃ & S
                                                                                     US g
                                                                                     T3  t-i O
    
                                                                                     ?l!
                                                                                     a  to S
                                                                                     QPQ
    

    -------
           •      The parent company that owns the establishment does not assign costs and revenues that
                  reflect the true financial health of the establishment. Two important examples are cost
                  centers and captive sites, which exist primarily to serve other facilities under the same
                  ownership.5
           •      The establishment is in financial trouble; that is, true costs exceed revenues.
    
    To the extent that these types of establishments are contained in an employment class, the projection of
    negative baseline income is accurate. In either case, EPA would be unable, even with the use of facility-
    specific survey data, to evaluate impacts to these establishments as a result of the rule.
           B.4.2   Skewed Distributions
    
           Two additional possible reasons for projected negative baseline establishment income are
    attributable to the methodology used to estimate the distributions:
    
           •       EPA assumed that the distribution of income around the model facility mean is normally
                   distributed when, it fact, it is positively skewed.
           •       EPA could not directly measure the variance of the income distributions, but instead had
                   to estimate it from incomplete data.
    
    In these two cases, EPA's methodology would project that more establishments have negative baseline
    income than would be expected in the industry.
    
           The effects of a positively skewed income distribution can be most apparent when one considers
    the distribution of establishment revenues. For the reasons listed above, it is possible — even probable
    — that some establishments earn negative income, whether measured by net income, or cash flow.
    However, an establishment cannot earn negative revenues, though establishments can earn zero
    revenues; the distribution of establishment revenues for an employment class should show zero facilities
            5 Captive sites may show revenues, but the revenues are set approximately equal to the costs of the operation.
     Cost centers have no revenues assigned to them.
                                                  B-29
    

    -------
    earning negative revenues.6 If, however, some facilities earn atypically large revenues, then the
    distribution may be positively skewed (e.g., the probability of the mean cash flow of $100,000 in Figure
    B-l would be significantly higher than 0.5; more than half of facilities in the model class would earn less
    than the mean cash flow).  In such a case, using a normal, symmetric distribution to approximate the
    skewed distribution would likely result in an overestimate of the percentage of establishments earning
    negative income. The Census Bureau has confirmed that in general, the distribution of facilities in  an
    employment size class tends to be positively skewed (Quash, 2001). However, even if the distribution of
    a variable such as revenues, payroll, or material costs is positively skewed, the distribution of a function
    of those variables (e.g., revenues minus payroll and material costs) will not necessarily be skewed.7
            B.4.3  -Adjustments to Variance
    
            EPA used the Census special tabulation to directly calculate the variance for [value of shipments
    - (payroll + material costs)] in each NAICS code and employment class. However, the actual measures of
    facility income used in the facility-level economic impact model are:
    
            •      EBIT = value of shipments - (payroll + material costs + benefits + all other costs)
            •      Net income = [value of shipments - (payroll + material costs + benefits + all other
                   costs)] x (1 - tax rate) - estimated interest payments
            •      Cash flow = net income + depreciation
    
    Because the actual income measures differed from the approximate income measure on which variance
    was estimated, EPA needed to adjust the variance of [value of shipments - (payroll + material costs)]
    associated with each of the actual income measures used in the model.
            6 Table B-7 presents the model facility mean and standard deviation for each income measure by employment
    class and NAICS code, as well as the probability that income is less than zero (based on that mean and standard
    deviation, and assuming income is normally distributed).
            7 The results of sensitivity analyses based on the assumption that the distributions of revenues and cash flow
    are skewed may be found in Appendix E.
                                                   B-30
    

    -------
           To adjust income variance, EPA used the following rules concerning the expected value of mean
    and variance:
    
                                             E[kx]=kE[x]
    
                                            V[kx] = k2V[x]
    
                                          E[a ฑ kx] = a ฑ kE[x]
    
                                           V[a ฑ kx] = k2V[x]
    
    where k and a are scalars, E[x] is the expected value of the variable x (i.e., the mean), and V[x] is the
    variance of x (Hamett, 1982). Intuitively, if one multiplies the mean of a distribution by some scalar k,
    the variance of that distribution expands or shrinks by the square of that scalar value. However, if
    instead of scaling the mean, one changes its value by adding or subtracting some constant, then the
    distribution shifts to the right or left on its x-axis, but its variance does not change.
    
            In the context of the mean and variance for the model facilities, to estimate the adjustment of the
    variance for net income, EPA had to first do the same for EBIT. EPA applied these rules in the following
    manner:
    
                   EPA first decreased the mean value of EBIT relative to the mean of [value of shipments
                   - (payroll + material costs)] by subtracting from it all other costs; however, the variance
                   for EBIT is unchanged and equals the variance for [value of shipments - (payroll +
                   material costs)].
    
     Conceptually/because it has a smaller mean but an identical variance, the distribution of EBIT will result
     in a larger probability of negative income relative to the distribution for the [value of shipments -
     (payroll + material costs)]. In practice, the probability that the [value of shipments - (payroll + material
     costs)] is less than zero in the four meat products NAICS codes ranges from 22 percent to 26 percent,
                                                  B-31
    

    -------
    while the probability that EBIT is less than zero generally ranges from 26 percent to 30 percent (in some
    
    isolated instances, it may be as high as 40 percent).8
    
    
           To estimate net income adjusted variance, EPA then did the following:
    
    
           •       The primary—but not the only—difference between net income and EBIT is tax
                   payments, which are calculated by multiplying EBIT by (1 - tax rate). Therefore, the
                   variance of net income is adjusted by multiplying the variance EBIT by the square of (1 -
                   tax rate).
    
    The probability that model facility net income is less than zero is thus identical to the probability that
    
    EBIT is less than zero.
    
    
            The distribution for estimated cash flow has an identical variance to net income, but a larger
    
    mean because depreciation is added to the mean of net income. The probability that cash flow is less
    
    than zero tends to be about 3 percent to 5 percent lower than the probability that net income is less than .
    
    zero.
    
    
            Had EPA simply scaled  the variance for net income and cash flow from the variance of the
    
    [value of shipments - (payroll +  material costs)], the probability that income was less than zero would be
    
    identical for each employment class within each NAICS code regardless of what income measure was
    
    used. That probability would also equal the probability that the [value of shipments - (payroll + material
    
    costs)] was less than zero, and would range from 22 percent to 26 percent according to NAICS code.
            8 EPA "smoothed" the estimated variance of the [value of shipments - (payroll + material costs)] by applying
    the median coefficient of variation (i.e., standard deviation divided by mean) within a NAICS code to all employment
    classes in that code. This results in an identical probability that income is less than zero for all employment classes
    within a NAICS code, though that probability differs between NAICS codes. EPA felt smoothing was appropriate
    because of: (1) relatively small populations in some employment classes, (2) relatively large differences in the
    coefficient of variation between employment classes within a NAICS code, and (3) the fact that only 12 different
    model facilities were selected from the 35 total model facilities, potentially increasing the effect of an outlier on the
    impact analysis.
    
                                                   B-32
    

    -------
           B.4.5   Effect on Modeling Impacts
    
           There.are many reasons why EPA's model results in a high probability of negative baseline
    income for facilities. First, true facility income may be negative in the baseline, due either to how
    multifacility companies choose to allocate costs and revenues among facilities or to financial distress.
    Second, EPA found it necessary to make certain assumptions when modeling a distribution of income
    for each class rather than single facility. The available data do not make it possible to determine what
    proportion of facilities will be projected to have negative baseline income results due to each reason.
    
           As one might expect, the percentage of facilities with negative baseline income will increase if:
    (1) the mean of a distribution decreases while the variance remains constant, or (2) the variance of a
    distribution increases  while the mean remains constant. In both cases, the percentage of facilities with
    negative baseline income increases because the portion of the distribution's tail lying below zero (i.e., to
    th: left of the $0 value in Figure B-l) is larger.   ,
    
            The effect of this issue on EPA's projection of economic impacts is not straightforward. The
    interaction'between the mean income and variance of a distribution on the one hand, and the range of
    estimated compliance costs on the other can be quite complex. Intuitively, one can observe on Figure B-
    1 that the incremental probability of closure will depend on the slope of the cumulative distribution
    function between $0 and the estimated compliance costs. Changes hi mean or variance will change the
    slope of the distribution function where it crosses the $0 value. However, the net effect on incremental
    probability will also vary according to the size of the compliance costs. The key point here is that an
    overestimate of "baseline closures" (i.e., facilities with income less than zero) does not necessarily lead
    to an underestimate of incremental closures.9
            9 Appendix E contains a sensitivity analysis where EPA used an alternate data source to estimate variance
     that resulted in a smaller probability of baseline closures.
                                                   B-33
    

    -------
    B.5    LIMITATIONS OF THE MODEL FACILITY APPROACH
    
           EPA based its model economic facilities on Census data, the only high-quality source of both
    revenue and cost data at a relatively disaggregated level. The limitation this places on the data is that the
    Census Bureau does not provide data distinguishing different production processes performed within
                                                                                                 •
    each employment class. All facilities in the 50 to 99 employee class for NAICS 311611, for example, are
    known to perform red meat slaughtering. However, an unknown percentage of those facilities also
    perform further processing, rendering, or both. Other things being equal, facilities that perform
    additional processes will also incur additional production costs and earn additional revenues. The
    financial data presented by the Bureau will be a weighted average of all those facilities.
    
           The effect of this is as follows. Consider two model engineering facilities with roughly equal
    full-time equivalent employment. One facility performs cattle slaughtering, the second performs cattle
    slaughtering, further processing, and rendering. Because both facilities slaughter cattle and have equal
    •.mployment, both facilities would be assigned  identical economic model facilities with identical income
    measures. The economic model facility would  probably overstate operating costs and revenues for the
    slaughtering facility but understate them for the slaughtering, further processing, and rendering facility
    (although the net effect on facility income cannot be determined). Other things equal, the second facility
    (slaughtering, further processing, and rendering) would incur larger compliance costs; measured against
    the same model facility income, it would also incur larger impacts.
    B.6     REFERENCES
    Brealey, R. A., and S. C. Meyers. 1996. Principles of Corporate Finance, 5th edition. New York: The
            McGraw-Hill Companies, Inc.
    Brigham, E. F., and L. C. Gapenski. 1997. Financial Management: Theory and Practice, 8th edition. Fort
            Worth: The Dryden Press.
    Financial Accounting Standards Board. 1996. Financial Accounting Standards: Explanation and
            Analysis. SFAS No. 105 (Disclosure of information about financial instruments with off-balance
            sheet risk and financial instruments with concentrations of credit risk), No. 107 (Disclosures
            about fair value of financial instruments), and No. 119 (Disclosure about derivative financial
            instruments and lair value of financial instruments). Bill D. Jamagin, ed.  18th edition. Chicago:
            CCH Incorporated, pp. 564-586.
                                                 B-34
    

    -------
    Harriett, Donald L. 1982. Statistical Methods (3rd ed.)- Reading, Massachusetts: Addison-Wesley
           Publishing.
    
    Mendenhall, W., D. D. Wackerly, and R. L. Scheaffer. 1990. Mathematical Statistics with Applications
           (4th ed.). Boston: PWS-Kent Publishing Co.
    
    Quash, 2001. Personal communication from Nishea Quash, U.S. Census Bureau, to Calvin Franz, ERG,
           September 10, 2001.
    
    U.S. Census Bureau.  1999a. Animal (Except Poultry) Slaughtering. EC97M-3116A. 1997 Economic
           Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
           November.
    
    U.S. Census Bureau.  1999b. Meat Processed From Carcasses. EC97M-3116B. 1997 Economic Census:
           Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.
    
    U.S. Census Bureau.  1999c. Poultry Processing. EC97M-3116D. 1997 Economic Census:
           Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce. November.
    
    U.S. Census Bureau. 1999d. Rendering and Meat Byproduct Processing. EC97M-3116C. 1997
           Economic Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of
           Commerce. December.
    
    U.S. Census Bureau. 2001. Special Tabulation of Census Data for NAICS 311611, 3.11612, 311613,
           311615. Washington, D.C.: U.S. Department of Commerce. May.
    
    U.S. EPA. 2002. Development Document for the Proposed Revisions to the Effluent Limitations
           Guidelines for the Meat Products Industry. EPA-821-B-01-007. Washington, D.C.: U.S.
           Environmental Protection Agency, Office of Water.
                                                B-35
    

    -------
    

    -------
                                           APPENDIX C
                            MARKET MODEL METHODOLOGY
    C.I    INTRODUCTION
    
           EPA developed a market model to examine the impacts of the meat products industry effluent
    guideline on the price and output of various meat products. The distinguishing feature of EPA's market
    model is that it explicitly incorporates cross-market impacts among meat types into the analysis. The
    demand for meat products such as beef, pork, broilers, and turkey is closely related; a 1 percent increase in
    the price of pork, for example, may cause'a 0.7 percent fall in the quantity of pork demanded and a 0.2
    percent increase in demand for beef.
    
           In the context of EPA's proposed ELG for the meat products industry, this increases the
    complexity of the market analysis. Because EPA's proposed ELG may simultaneously affect the price of
    beef, pork, chicken, and turkey, the market analysis for each product depends not only on the compliance
    costs for that product but on the impact of compliance on the prices of the other three meat products.
    
           For example, if the ELG imposes compliance costs on the producers of beef products, then the
    supply of beef products will tend to decrease (i.e., the supply curve for beef will shift to the left; a smaller
    quantity of beef will be offered for sale at the current price). If all other things remained constant, this
    would tend to increase the price of beef products while decreasing the quantity sold. However, EPA's ELG
    may also impose compliance costs on pork producers, tending to increase the price of pork. All other things
    being constant, the increase in the price of pork would increase the demand for beef products; the demand
    curve for beef will shift to the right. This would tend to increase the price of beef as well as increase the
    quantity of beef sold. The final impact on the price and output of beef products will depend on the relative
    magnitude of supply and demand shifts. Figure C-l illustrates the general rule behind this example.
    
            If all meat products incur relatively similar per-unit compliance costs, cross-market impacts would
    tend to be roughly offsetting. However, if per-unit compliance costs are asymmetric (e.g., per-unit
    compliance costs are significantly larger for some subcategories than for others), then potentially
    
                                                   C-l
    

    -------
                                                      Decrease in supply of
                                                      Meat Product i caused by
                                                      ELG on Meat Product i
    ppost
    ppr
                                                                                     s2
                                                                                     s1
                                                                Increase in demand for
                                                                Meat Product i caused by
                                                                ELG on Meat Product j
    
    
                                                                  D2
    
    
                                                                  D1
                                       Qpost   Qpre
               D1, S1 = preregulatory market supply and demand conditions
               D2, S2 = postregulatory market supply and demand conditions
               ppra> Qpra = prerequlatory equilibrium price and quantity
                   t, Qpost _ postrequlatory equilibrium price and quantity
                                            Figure C-l
    
                      Impact of the Effluent Guideline on Market for Meat Product i
                                               C-2
    

    -------
    significant shifts could occur between meat product markets. EPA's model was developed with the
    flexibility to analyze the latter situation as well as the former.
    
           In order to incorporate both cross-market effects and international trade into the model, EPA
    specified linear supply and demand equations in each market to make the model tractable. The slopes of the
    equations were derived from estimated price elasticities of supply and demand found in existing research.
    These elasticities were then converted to slopes at the baseline equilibrium price and quantity. Because
    domestic supply, domestic demand, import supply, and export demand are all specified as linear functions,
    the model components are additive, and simultaneous equilibrium can be solved for in multiple markets
    using linear algebra.
    
            Of major concern to observers of the meat product industry is the issue of potential market power.
    EPA selected a perfectly competitive structure for the meat products market model after performing an
    extensive literatec search. EPA found that most researchers were unable to reject the existence of
    perfectly competitive markets in the beef and pork markets; in the poultry market, market power was found
    to exist for meat processors vis-a-vis livestock suppliers, but not against customers in the output market.
    The results of this literature search are presented in the industry profile.
    
            Section C.2 presents the basic market model specification and solution. Section C.3 discusses data
    sources for the model.
     C.2    MARKET MODEL APPROACH
    
            First, standard domestic supply, domestic demand, import supply, and export demand equations
     are developed for each meat product. These equations express quantity as a linear function of a product's
     domestic price. The linear function's slope is expressed by a price parameter, derived from elasticities in
     the literature. Domestic demand for each meat product is specified as a function of the price of the other
     three meat products in addition to its own price. For the market for each meat product to be in equilibrium,
     U.S. domestic demand for a meat product and foreign demand for U.S. production of that meat product
     (exports) must be equal to U.S. domestic supply of the product and foreign sales of that product to the U.S.
                                                    C-3
    

    -------
    (imports) at its current market price. This equilibrium condition is used to derive an excess demand
    function for each meat product.
    
           Second, the excess demand equations are solved. Because the excess demand function for each
    meat product is linear, expressing the equations for the four meat products in matrix form results in a
    convenient way to solve the equations simultaneously. Given pre-regulatory prices, quantities, and price
    parameters, linear algebra is used to solve for the pre-regulatory intercept for all four excess demand
    equations.
    
            Third, the supply curve shift for each meat product is calculated. (Imposing ELGs on the industry
    causes the supply curve for each meat product to shift.) The supply curve shift for a meat product is
    estimated as a function of average per-unit compliance costs for that product. Once the post-regulatory
    (i.e., post-shift) supply  curve is estimated, the excess demand equation for each meat product is re-written.
    
            Fourth, the post-regulatory excess demand equations for all four meat products—like the pre-
    regulatory equations—are expressed in matrix form. The post-regulatory intercept for each excess demand
    equation, however, is'already known: it is a function of the pre-regulatory intercept, per-unit compliance
     costs, and the supply equation price parameter. By using linear algebra to invert the matrix containing the
     price parameters, then  multiplying the post-regulatory intercept vector by that inverted matrix, EPA can
     evaluate the set of meat prices that results in simultaneous equilibrium for all four meat products.
    
            Finally, the individual component equations for each meat product's domestic supply, domestic
     demand, import supply, and export demand are evaluated using the post-regulatory prices to solve for post-
     regulatory quantities. Changes in these four quantities for each meat product, as well as changes in the
     price of each meat product, measure the market-level impacts of a meat products effluent guideline.
    
             Each of the steps used to model market-level impacts is described in detail below.
                                                    C-4
    

    -------
           C.2.1.  Development of Excess Demand Functions for Individual Meat Products
    
           EPA modeled the market for each of the four meat products: beef (B), pork (P), chicken (C), and
    turkey (T) using four linear equations:
                                       Qis = ซst
                                                         + ฃ
    where the U.S. domestic quantity demanded of meat product i, QiD, is a function of both the U.S. domestic
    price of meat product i, Pj, and the U.S. domestic price of other meat products j, Pj. U.S. domestic supply
    of meat product i, QiS, is modeled as a function of domestic price, Pi; only, as are "rest-of-the-world"
    (ROW) demand for U.S. meat product i, Qsx (exports), and U.S. demand for ROW meat product i, Q;M
    (imports). Clearly, each meat product's supply and demand (both domestic and foreign) depend on the price
    of many other factors as well as its own price (and the price of other meat products in the case of domestic
    demand).  However, because EPA is holding the prices of these other factors constant for the purposes of
    this analysis, it is not necessary to explicitly represent them in the relevant equation.
    
            The parameters d;i, s;i, xh and nij represent the slopes of their respective functions (i.e., the change
    in quantity of product i for a given change in the price of product i). The dy parameters shift the demand
    curve (the change in demand for product i for a given change in the price of product j — holding P;
    constant). The parameters aDi, axi, asi, and aMi are the intercepts of their respective equations.
    
            The values for the domestic demand equation slope and shift parameters are estimated from
    published estimates of own- and cross-price demand elasticities. One linearizes these elasticities by
    multiplying the elasticity by baseline quantity and dividing by baseline price. Thus, if:
                                                   C-5
    

    -------
    then:
                                             _
                                                3Qฐ
    where s, is the elasticity of demand for product i with respect to the price of product j, and both quantity
    demanded (Qj0) and price (P;) are set equal to their baseline values.
    
            Similarly, the slopes of domestic supply, S,, import supply, tn,, and export demand, xi; functions
    can be defined as:
                                                        Qis     '
                                                              I
                                                3P
                                                 dp,      P,
                                           Xi  " IP
    where Yii, nซ.
                             elasticities with respect to U.S. domestic price.
             In equilibrium, U.S. demand for meafrproduct i (Q;D) and foreign demand for U.S. meat product i
      (Qix) must be equal to U.S. supply of meat product i (QiS) and foreign sales of meat product i to the U.S.
      (QjM) at the current market price for meat product i:
                                                    C-6
    

    -------
                                        QSD + Q*  =  Q;s
    This can then be expressed as an excess demand equation for meat product i:
                                             Q*  -  Qis - QiM = 0
    or:
                      diiPi  + E
            Simplifying the excess demand function for each meat product, and making a notational
    
     substitution for convenience, results in:
                                                     *, - sa - m,)P,  * ^  dyPj
                                                     ' '                  i*J
                                                                = 0
                                                      I*J
     The solution for the intercept of the individual meat product excess demand function is:
                                           iPi + E  dyPj  =
                                                    C-7
    

    -------
            C.2.2  Simultaneous Solution of Pre-Regulatory Excess Demand Equations
    
            To solve the excess demand equations for all four meat products simultaneously, one writes the
     equations in matrix form:
                                      dBP   dBC
                                                    BT
                                PB
                                             pc     pT
                                CB     CP
                                TB
                                                    CT
                                             TC
    
    
    
    Pp
    PC
    PT
    
    
    
    -Up
    -*c
    -TCT_
    If this is expressed in vector notation as A*P = n, the intercept for each excess demand equation, TCJ, can be
    solved for using known prices and values for the price parameter elements of the A matrix.
            C.2.3  Post-Regulatory Excess Demand Functions
    
            The imposition of regulatory costs causes a decrease in the supply of each meat product for which
    an effluent guideline is developed. If di represents the per unit compliance costs for meat product i, the
    post-regulatory supply curve is:
    
    
    
    
    Substituting the post-regulatory supply curve into the excess demand function and rearranging it (using the
    notation-simplifying substitutions), the excess demand for each product i is:
    
    
                                     Vi+E  dflPj =  - *„ซ,-*,'
                                                  C-8
    

    -------
            C.2.4  Simultaneous Solution of Post-Regulatory Excess Demand Functions
    
            The post-regulatory excess demand functions for each meat product are again placed in matrix
    form to solve the system of equations for the set of post-regulatory prices that generate equilibrium in all
    four markets simultaneously. The system of simultaneous equations is:
                            PB
                           dTB
                                   BP
                           dCB   dCP
                                   TP
    dBC    dBT
    dpc    dPT
                                                CT
    -Spp6p  - Tip
    ~STT5T ~
    In this set of simultaneous equations, the elements of matrix A are known (e.g., X{, dy), as are the elements
    of the new vector IE* (e.g., s^, 8f, %). The set of meat product prices that will .esult in equilibrium in all
    four meat product markets can be solved for by multiplying the vector II* by the inverse of the A matrix
    (i.e., P' = A'1!!*).
           C.2.5   Post-Regulatory Price and Quantities
    
           The new equilibrium price for each meat product, P;', is substituted back into the component
    equations to solve for the post-regulatory domestic demand, Q;D', domestic supply, Qjs', export demand,
    QiX/, and import supply, QjM/, for each meat product:
                                                 ซxi + xipi'  = Qi
                                                                 x/
                                                  C-9
    

    -------
    The changes in market price (P; - P,'), domestic demand, (QjD - QiD'), domestic supply, (QjS - QjS'X export
    demand, (Q* - QjX/), and import supply, (QjM - QiM/) for each meat product are the projected market-level
    impacts of the effluent guideline.
    C3     DATA SOURCES FOR MARKET MODEL ANALYSIS
    
    
            Following is an evaluation of potential publicly available data sources for baseline values and key
    parameters.1
    
    
    
            C3.1  Baseline Market Quantities and Prices
    
    
            EPA examined a number of possible sources for baseline quantity and price data. Of these, the
    three most important are:
                   Economic Census of Manufacturers, which provides both value and quantity data for a
                   fraction of 1997 industry shipments at the 10-digit product level. The transactions price
                   can be calculated for those products with both value and quantity data. Use of Census data
                   limits the baseline to 1997, because the Annual Survey of Manufactures provides only on
                   value of shipments, and there is no Current Industrial Report for meat products. For these
                   products, data are available on both value and quantity of shipments as a percent of value
                   of industry shipments:1
    
                   —      Beef: 27.4 percent of combined Animal Slaughtering and Processing Industries
                           (NAICS 311611 and 311612; Census, 1999a and 1999b), including boxed beef.
    
                   —      Pork: 11.4 percent of the combined Animal Slaughtering and Processing Industries
                           (NAICS 311611 and 311612; Census, 1999a and 1999b).
    
                   —      Chicken: 39.9 percent of Poultry (NAICS 311615; Census, 1999c).
       1 Dividing value data by quantity results in the transactions price of the product, thus both are necessary to
    determine baseline price and output. In the combined Animal Slaughtering and Processing industries (NAICS
    311611 and 311612), 20.8 percent of products had both value and quantity data, but could not be classified by meat
    type; 25.3 percent of products with price and quantity data in the Poultry industry could not be classified by meat
    type. For Animal Slaughtering and Processing, 40.4 percent of products had value data only, while 22.6 percent of
    Poultry products had only value data. No products in Rendering (NAICS 311613; Census, 1999d) had both value
    and quantity data.
    
                                                   C-10
    

    -------
                  —     Turkey: 12.2 percent of Poultry (NAICS 311615; Census, 1999c).
    
                  USDA Livestock, Dairy and Poultry Situation and Outlook (Outlook), which provides
                  quantity and price data for relatively aggregated meat products: carcass weight of beef and
                  pork, ready-to-cook (RTC) weight for broilers and, turkeys.2 Prices are for selected
                  wholesale and retail products. Outlook also provides the carcass and RTC weight for both
                  imports and exports of meat products at the same level of aggregation through USDA's
                  Foreign Agricultural Trade of the United States (FATUS) database.3 Data for 1995
                  through 2000 were obtained from the USDA Web  site.
    
           •      USDA Food Consumption, Prices, and Expenditures, 1970-97 (Putnam and Allshouse,
                   1999), which provides quantity of meat products by carcass weight (RTC weight for
                  poultry), retail weight, and boneless weight.4 Carcass, RTC, and trade weights reported
                   are generally within 1 percent of those reported in  Outlook. Interestingly, this source cites
                   small quantities of broiler and turkey imports (e.g., 5 million pounds, RTC weight for
                   broilers, less than 0.02 percent of domestic production), while both Outlook and the
                   FATUS database report no imports for these two meat products. This report also provides
                   the Bureau of Labor Statistics' Consumer Price Index and average annual retail price at a
                   more detailed level than does Outlook.
    
    Table C-l presents baseline output data by meat type for 1997 from all three sources; it also presents
    
    estimated transactions prices from Census data and selected average wholesale and retail prices from
    
    Outlook and Putnam. Although the Census production data differ significantly from the carcass weight
    values reported in Outlook and Putnam, with the exception of pork, the Census data is reasonably similar
    
    to Putnam's retail and boneless weight figures.
    
    
           EPA selected Outlook data for the baseline price and quantity. Although EPA's first choice would
    
    have been to use Census data where the price could be calculated as each product's transactions price
       2 Carcass weight of beef is defined as the chilled, hanging carcass, including the kidney and attached internal
    fat (kidney, pelvic, and heart fat), but not the skin, head, feet, and unattached internal organs. Carcass weight of
    pork is defined as the chilled, hanging carcass, including the skin and feet, but excluding the kidney and attached
    internal fat. RTC weight of poultry consists of the entire dressed bird, including bones, skin, fat, liver, heart,
    gizzard, and neck (Putnam, 1999).
    
       3 The trade data for beef include veal; domestic production of veal is recorded separately.
    
       4 Retail and boneless weights adjust for those parts of the carcass not generally bought by consumers. These are
    not directly calculated, but instead are estimated using conversion factors. For beef, retail weight is 70 percent, and
    boneless weight is 67 percent, of carcass weight. For pork, retail weight is 78 percent, and boneless weight is 73
    percent, of carcass weight. For broilers, retail weight is 87 percent, and boneless weight is 61 percent, of RTC
    weight. For turkeys, boneless weight is 79 percent of RTC weight (Putnam, 1999).
    
                                                   C-ll
    

    -------
                         Table C-l
    1997 Baseline Quantity and Price Data for Market Model
    Data Source
    Meat Product
    Beef
    .Pork
    Chicken
    Turkey
    U.S. Domestic Production (millions of pounds) '
    1997 U.S. Census
    USDA Outlook: Carcass/RTC Weight
    15,133
    25,384
    5,720
    17,244
    21,180
    27,271
    4,119
    5,478
    USDAFCPA:
    Carcass/RTC Weight
    Retail Weight
    Boneless Weight
    25,490
    17,843
    17,053
    17,242
    13,380
    12,569
    U.S. Jtoports(rnillions of pounds) :
    1997 U.S. Census
    USDA Outlook: Carcass/RTC Weight
    USDA FCPA: Carcass/RTC Weight
    NA
    2,343
    2,343
    NA
    633
    633
    27,041
    23,499
    16,441
    
    
    NA
    NA
    5
    U.S. Exports (millions of pounds) .';.
    1997 U.S. Census
    USDA Outlook: Carcass/RTC Weight
    USDA FCPA: Carcass/RTC Weight
    NA
    2,136
    . 2,136
    NA
    1,044
    1,044
    NA
    4,664
    4,664
    5,412
    NA
    4,275
    
    NA
    NA
    1
    
    NA
    606
    598
    Representative U.S. Domestic Prices ; !; ^ ; ;
    1997 U.S. Census: Transactions Price
    USDA Outlook: Average Wholesale Price
    Beef, Central, Boxed, Choice, 550-700 Ib.
    Beef, Central, Boneless, 90% Fresh
    Pork, Central, Cutout, Composite
    Pork, Central, Loins, 14-19 Ib., Bl 1/4"
    trim
    Broilers, 12 City Average
    Broilers, Northeast, Boneless Breast
    $1.323
    $1.454
    $0.584
    
    $1.033
    $0.908
    
    
    
    
    
    
    $0.709
    $1.081
    
    
    
    
    
    
    $0.588
    $1.720
    $0.915
    
    
    
    
    
    
    
                           C-12
    

    -------
                      Table C-l (cont.)
    1997 Baseline Quantity and Price Data for Market Model
    
    Data Source
    Turkey, Eastern, Hens, 8-16 Ib.
    Turkey, Eastern, Drumsticks
    MeatProduct
    Beef
    
    
    Pork
    
    
    Chicken
    
    
    Turkey
    $0.649
    $0.311
    USDA FCPA: Average Retail Price
    Ground Beef, 100% Beef
    Chuck Roast, Choice, Boneless
    Sirloin Steak, Choice, Boneless
    Bacon, Sliced
    Chops, Center Cut, Bone-in
    Ham, Boneless, Excluding Canned
    Sausage, Fresh, Loose
    Chicken, Fresh, Whole
    Chicken, Breast, Bone-in
    Turkey, Frozen, Whole
    $1.40
    $2.43
    $4.21
    
    
    
    
    
    
    
    
    
    
    $2.68
    $3.48
    $2.79
    $2.15
    
    
    
    
    
    
    
    
    
    
    $1.00
    $2.04
    
    
    
    
    
    
    
    
    
    
    $1.05
                             C-13
    

    -------
     weighted by output share, too many observations were missing in the Census data. Outlook's primary
     advantage over Putnam's data is that it is more up to date.5 Given the highly aggregated nature of Outlook
     data, and given that the Outlook data are tracked at the carcass weight level, EPA selected Outlook's
     wholesale price measures to use as baseline price; these are best interpreted as indicator prices rather than
     the explicit price of all output. EPA determined that Putnam's retail price measures were not linked closely
     enough to the carcass weight output to be suitable for use as the baseline prices.
            C.3.2   Compliance Costs
    
            In order to estimate the supply curve shift for each meat type, EPA calculated average compliance
     costs per unit of output. Conceptually, per-unit compliance costs for each meat type are simply the sum of
     annualized compliance costs divided by meat output.
    
            EPA initially estimated compliance costs by process (first, further, and rendering) within general
     meat type categories (e.g., red meat and poultry). This meant that EPA had to attribute (1) estimated
     compliance costs for red meat to beef and pork and (2) estimated compliance costs for poultry to chicken
     and turkey. To do this, EPA first estimated total annualized compliance costs for each subcategory and size
     class (e.g., red meat, further processors, medium size). Then, for each subcategory size class, EPA
     calculated the quantity and percent of total meat production accounted for by each meat type (beef, pork,
     chicken, and turkey). Costs were attributed by the percent each meat type made up of total meat production
     for that subcategory size class (e.g., if red meat, further processors, medium sized facilities produced 70
     percent beef, 70 percent of annualized compliance costs for that subcategory size class would be attributed
     to beef). Per-unit costs were estimated by dividing the attributed compliance costs for each meat type by the
     quantity of that meat type produced.
            To determine the average per-unit compliance costs for each meat type over all subcategories and
    size classes, EPA took a weighted average of the per-unit costs for each subcategory and size class by meat
       * Putnam cites small quantities of broiler and turkey imports (e.g., 5 million pounds, RTC weight for broilers,
    less than 0.02 percent of domestic production), while both Outlook and the FAT US database report no imports for
    these two meat products. EPA used Putnam's import quantity data for chicken and turkey rather than Outlook's
    data.
                                                   C-14
    

    -------
    type. The weights were calculated as the meat type output within each subcategory and size class expressed
    as a percent of total output of that meat type over all subcategofies and size classes. (Note that, to an
    estimation of market-level compliance costs per unit, the distinction between direct and indirect dischargers
    is irrelevant.) Finally, to estimate market-level impacts, EPA entered average per-unit compliance costs by
    meat type directly into the market model.
            C.3.3  Price Elasticities
    
            C.3.3.1 Price Elasticities of Demand
    
            Domestic price elasticities of demand are widely available from a variety of sources, including
    USD A and academic research. The results of the literature search for demand elasticities is documented in
    the record. Fr.r use in its market model, EPA selected K. S. Huang's A Complete System of U.S. Demand
    for Food (1993).
    
            The advantage of Huang's estimates is that they were generated in a single, coherent, consistent
    framework that satisfies theoretical constraints of symmetry, homogeneity, and Engel aggregation. This
    should make using them better than selecting individual elasticities from among several sources with
    varying methodologies, degrees of aggregation, and time horizons. The internal consistency of Huang's
    work is of particular importance because EPA is modeling cross-product impacts in the market model. The
    own- and cross-price elasticities of demand are presented in Table C-2.
            C.3.3.2 Price Elasticities of Supply
    
            EPA undertook a literature search for estimates of the price elasticities of meat supply for both the
    feedlots and meat products effluent limitations guideline (ELG). This search resulted in a wide range of
    estimated elasticities with little apparent consensus.
                                                   C-15
    

    -------
    
    
    
    
    
    S-l
    8
    CO
    1
    ^*
    2
    ot
    •"S
    ^
    S
    _o
    •3
    Ol
    JU
    fe
    .s
    TJ
    Table C-2
    mand Identifie
    OJ
    Q
    13
    e
    CJ
    3?
    04
    cu
    s
    frfc ,
    o
    2
    ••a
    '3
    H
    
    To "^
    "So
    o
    W)
    a
    W
    1
    ^
    ,12
    OJ
    a
    s
    1
    PH
    
    
    CM
    a>
    w
    'PQ
    
    
    
    ^.1'
    J3 >>
    
    
    11
    OJ J>
    CO ^
    
    & ^
    >-
    
    
    
    
    •||
    S >
    
    
    •a ^
    CO
    
    g
    O ^jjjj*
    t-5 ^>-
    
    
    "
    1
    OJ
    CO
    s
    o
    0
    00
    T-H
    o
    CD
    t— i
    <— I
    CD
    
    
    ^
    ปy
    
    
    
    
    O
    to
    , j
    o
    
    
    vo
    O
    
    0
    ฐ?
    
    
    
    
    o
    CO
    
    
    1
    -*
    
    R
    ^~*
    9
    
    
    
    ซ*-<
    u
    PQ
    CO
    O
    CD
    CO
    T-H
    O
    CD
    S
    
    
    CO
    •— <
    CD
    
    
    
    1.
    CD
    
    
    OO
    CD
    
    8
    T1
    
    
    
    
    oo
    04
    vo
    CD
    
    
    00
    01
    ,VO
    ฐ
    
    ti.
    ฃ5
    CD
    
    
    
    ^
    ^,
    O
    PH
    i
    CD
    
    Si
    
    ^
    o
    CD
    
    
    S
    T-H
    c5
    
    
    
    s
    CD
    
    
    CO
    CD
    
    O
    04
    ~?
    
    
    
    
    OO
    CD
    
    
    O
    O
    
    S
    ^3
    CD
    
    
    ฃ2
    J-
    1
    ra
    <.
    
    ^
    o
    o
    r— 1
    CD
    
    
    OO
    o
    CD
    
    
    
    04
    C3
    
    
    CO
    m
    CD
    
    O
    oo
    VO
    9
    
    
    
    
    oo
    CD
    
    
    0
    ,o
    <=>
    
    o
    I— 1
    04
    C3
    
    
    
    1
    3
    H
    
    
    
    
    *T-1
    market moi
    0
    3
    0
    ฃ5
    4—(
    J5
    o
    is
    &
    S .
    &
    o
    '(4-1
    : a consensus o
    ง
    CO
    i,
    ,
    t
    &!
    w
    1 •
    o
    ^o
    (D
    J"ฐ
    CO
    (D rn
    T-* Q\
    
    3 *~*
    1 1
    jy cc
    ^H •"*
    ง ••
    '^ O
    Kj O
    m co
     8
    ฃ
     I
    o
     o
    1
             If
                     oo
                     10
                     IT)
    
                     o
                     00
                     10
                     O
                     00
                                     ON
                                <  5^ 5!
                o
                oj
                oi
                           VC3
    
                           CD
                                04
                                i—i
    
                                10
                     vq
                     9
                           00
                           CO
                          04
                          r-
                          co
                          CD
       •g .-.
       e  w
       s-'-S
       A -a
       ^
                     •ON
                     0
                     O
                           CO
                           o
    o
    o
    0
    c5
    
                     O
                     00
                     O
                     00
                                      c\
                                O
    42
          I
             I?
             PH
                r-
                i—i
                o\
                           o
    
                           CO
                                s
                     
    -------
            EPA undertook a literature search for estimates of the price elasticities of meat supply for both the
     feedlots and meat products ELGs. This search resulted in a wide range of estimated elasticities with little
     apparent consensus.
    
            Because of this lack of consensus, EPA decided to use the elasticities from the ELG for
     concentrated animal feeding operations (CAFOs). These elasticities were selected for the CAFOs model
     with the concurrence of EPA's expert consultants (U.S. EPA, 2001). It is reasonable to use these
     elasticities for the meat products market model, because meat (in the form of both live animals for
     slaughter and meat products) makes up the majority of material costs in the meat products industry (79
     percent in animal slaughtering, 63 percent in meat processing, and 76 percent in poultry (U.S. Census
     Bureau, 1999a through 1999d). In addition, the other major cost component of meat production is unskilled
     labor, and the price elasticity of primarily unskilled supply tends to be  large. Thus, the CAFOs supply
     elasticities should represent a reasonable lower-bound estimate for the price elasticity of meat supply. The
     supply elasticities selected for use in the model are presented in Table C-2.
            C.3.3.3 Import and Export Elasticities With Respect to U.S. Domestic Price
    
            EPA used an Armington-type specification to model the effects of international trade on U.S. meat
    products markets. If foreign-produced and domestically produced goods are perceived as perfect substitutes
    for each other—that is, if consumers do not differentiate between foreign- and domestically produced
    goods—then one would expect a country to either import those good or export them, but not to both import
    and export them simultaneously. However, if consumers perceive Foreign and domestically produced goods
    in a particular class as close but not perfect substitutes, then their country may import and export that class
    of products simultaneously. The U.S. both imports and exports meat products; the Armington specification
    that EPA selected incorporates product differentiation in the meat products industry market model.
    
            Econometrically, the Armington model measures the degree of substitutability between traded
    products. This is expressed as  the percentage change in market share of the imported product relative to the
    domestically produced good caused by a change in the relative prices of the imported and domestic goods.
    An elasticity of zero implies that consumers will not substitute imported meat products for domestic meat
                                                  C-17
    

    -------
     products; the higher the elasticity, the more willing consumers are to make this substitution. This means
     that if the elasticity of substitution is equal to one, then market shares remain constant; if this elasticity is
     greater than one, then an increase in U.S. price means that U.S. market share will decrease (Armington,
     1969a).
    
            The Armington elasticity of substitution cannot be directly used in EPA's market model. However,
     Armington demonstrated that own price and cross price trade elasticities are a function of domestic demand
     elasticities, market shares of domestic and foreign products, and the value of the elasticity of substitution
     (Armington, 1969a, 1969b). This means that EPA could use Armington's results to derive formulae for the
     uade elasticities used in its market model.6
    
            The U.S. elasticity of demand for imports of meat product i with respect to the U.S. product price
     dri) is a function of its domestic elasticity of demand (EH), the ratio of "rest of world" (ROW) and U.S.
     market shares (0% and 0UR; EPA assumed for simplicity that there are only two countries, the U.S., and
     the ROW, thus 6% =  1 - 8UR), and the elasticity of substitution parameter for the U.S. (ฃu):
                                               n—  -   f P   -t- Sf \
                                            mi     „ \<3     *",;)
    The expected value of TJ^ is positive. That is, an increase in the U.S. domestic price of meat products is
    expected to increase U.S. demand for ROW meat products. The elasticity specified above meets this
    expectation as long as the elasticity of substitution between U.S. and ROW meat products, ฃu, is greater
    than the U.S. domestic price elasticity of demand for U.S. meat products, eH.
    
           Similarly, EPA estimated the elasticity of ROW demand for U.S. meat products (ry, e.g., U.S.
    exports) with respect to U.S. price as:
       6 Further details of this derivation may be found in the rulemaking record.
                                                  C-18
    

    -------
    which specifies that the elasticity of ROW demand for U.S. meat products is a function of the ROW
    demand for ROW meat products (eRri), relative market shares (9RR and 0Ru), and ROW consumers'
    elasticity of substitution between ROW and U.S. meat products (ฃR). Because own price elasticity of
    demand is small, the value of r|xi is negative: an increase in U.S. price will decrease U.S. exports. •
            Due to a lack of data availability, EPA calculated a numerical value for this elasticity assuming
    that:
            •       The ROW elasticity of substitution for U,". meat products is identical to the U.S.
                   elasticity of substitution for ROW meat products (i.e., ฃR = ฃu).
            •       The elasticity of ROW demand for meat products with respect to ROW price, SRU, equals
                   the elasticity of U.S. demand for meat products with respect to U.S. price, eu.
    
    Note that because the U.S. share of ROW expenditures on meat products is small, the value of the ROW
    trade elasticity approaches the value for the elasticity of substitution (i.e.,  T|xi - -ฃR). Thus, the assumption
    that the overall elasticity of ROW meat product demand equals the overall elasticity of U.S. meat product
    demand (i.e., 8Ri; = eH) is not crucial to the results of the analysis.
            Sources for domestic demand elasticities are discussed above. Market shares of meat production
    were estimated at a consistent level of aggregation using quantity data from the United Nations Food and
    Agriculture Organization.
    
            Long-run Armington elasticities were obtained from Gallaway et al. (2000). Note that Gallaway
    estimated elasticities at the 4-digit SIC level for Meat Packing (SIC 2011) and Poultry and Egg Processing
    (SIC 2015). Because these SIC codes contain more than one product, but do not distinguish between beef
    and pork (SIC 2011) or chicken and turkey (SIC 2015), EPA used the same elasticity of substitution (ฃ) for
    each product described by a code. EPA did use the own price elasticity and market shares specific to each
    
                                                   C-19
    

    -------
    meat type in calculating that meat type's trade elasticities. Table C-3 presents a summary of the trade
    parameters and elasticities with respect to changes in domestic price that were used in the model.
    C.4    REFERENCES
    
    
    Armington, Paul S. 1969a. A theory of demand for products distinguished by place of production.
           International Monetary Fund Staff Papers. 16(1): 159-177.
    
    Armington, Paul S. 1969b. The geographic pattern of trade and the effects of price changes. International
           Monetary Fund Staff Papers. 16(2): 179-199.
    
    Gallaway, Michael P., Christine A. McDaniel, and Sandra A. Rivera. 2000. Industry-Level Estimates of
            U.S. Armington Elasticities. Office of Economics Working Paper. Washington, D.C.: U.S.
            International Trade Commission. September.
    
    Huang, K. S. 1993. A Complete System of U.S. Demand for Food. Technical Bulletin Number 1821.
            Washington, D.C.: U.S. Department of Agriculture, Economic Research Service.
    
    Outlook. Various dates. Livestock, Dairy and Poultry Situation and Outlook. Washington, D.C.: U.S.
            Department of Agriculture, Economic Research Service.
    
    Putnam, Judith J., and Jane E. Allshouse.  1999. Food Consumption, Prices, and Expenditures, 1970-97.
            Statistical Bulletin Number 965. Washington, D.C.: U.S. Department of Agriculture, Food and
            Rural Economics Division, Economic Research Service.
    
     Unnevehr, Laurian J., Miguel I. Gomez, and Philip Garcia. 1998. The Incidence of Producer Welfare
            Losses from Food Safety Regulation in the Meat Industry. Review of Agricultural Economics.
            20:186-201.
    
     UN FAO data: downloaded 2/20/01 from: http://apps.fao.org/page/collections?subset=agriculture
            Agricultural Production/Livestock Primary & Processed/World+, United States of America/.   •
    
     U.S. Census Bureau. 1999a. Animal (Except Poultry) Slaughtering. EC97M-3116A. 1997 Economic
            Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
            November.
    
     U.S. Census Bureau. 1999b. Meat Processed From Carcasses. EC97M-3116B. 1997 Economic Census:
            Manufacturing Industry Series. Washington, D.C.:  U.S. Department of Commerce.  November.
    
     U.S. Census Bureau. 1999c. Poultry Processing. EC97M-3116D. 1997 Economic Census: Manufacturing
            Industry Series Washington, D.C.: U.S. Department o? Commerce. November,
                                                  C-20
    

    -------
    U.S. Census Bureau. 1999d. Rendering and.Meat Byproduct Processing. EC97M-3116C. 1997 Economic
           Census: Manufacturing Industry Series. Washington, D.C.: U.S. Department of Commerce.
           December.
    
    U.S. EPA. 2001. Economic Analysis of the Proposed Revisions to the National Pollutant Discharge
           Elimination System Regulation and the Effluent Guidelines for Concentrated Animal Feeding
           Operations. EPA-821-R-01-001. Washington, D.C.: U.S. Environmental Protection Agency,
           Office of Water. January.
                                              C-21
    

    -------
    

    -------
                                        APPENDIX D
                       SUMMARY OF DEMAND AND SUPPLY
                               ELASTICITY LITERATURE
    D.I    SUMMARY OF PRICE ELASTICITY ESTIMATES
           This appendix presents the results of EPA's literature review and the magnitudes of published
    demand and supply elasticities for the beef, pork, and poultry sectors.
    
           EPA has reviewed the available literature on the demand and supply characteristics of the beef,
    pork, and poultry markets. These expanded reviews include an annotated summary of each study and are
    contained in the record (Section 8.3.2). The majority  of the models in the literature are based on
    econometric estimations of various demand and supply system specifications, such as .the Almost Ideai
    Oemand System (AIDS) and the Rotterdam model. However, given the prevalence of non-theoretical
    approaches to estimating demand and supply responses in the literature using such techniques as vector
    autoregression (VAR), EPA also includes those studies in the tables where applicable.
                                               D-l
    

    -------
                                          Table D-l
    Demand Elasticities for Beef Products Ranked from the Lowest Estimate to the Highest Estimate
    Source
    Bales and Unnevehr (1988)
    Capps (1989)
    Brester and Wohlgenant (1991)
    Heien and Pompelli (1988) l
    Moschini and Meilke (1989)
    Huang and Hahn (1995) l
    Gao and Shonkwiler (1993) l
    Kesavan et. al. (1993) !
    Brester and Wohlgenant (1991)
    Ospina and Shumway (1979)
    Alston and Chalfant (1993)
    Choi and Sosin (1990)
    Brester (1996)
    Chavas (1983)
    Hahn (1994) l
    Eales and Unnevehr (1993)
    Heien and Pompelli (1988) :
    Moschini, Moro, and Green (1994)
    Ospina and Shumway (1979)
    Brester and Wohlgenant (1991)
    Brester (1996)
    Wohlgenant (1989)
    Marsh (1992)
    Heien and Pompelli (1988) '
    Capps (1989)
    Elasticity Estimate |
    -2.59 (hamburger) . |
    -1.27 (roast beef)
    -1.155 (fed beef)
    -1.11 (roast)
    -1.05 (beef)
    -1.036 (high quality beef)
    -1.03 (beef)
    -1.02 (long-run, beef)
    -1.015 (ground beef)
    -0.98 (fed beef; Langemeier and Thompson, 1967) . jj
    -0.98 (beef) ||
    -0.971 (red meat)
    -0.96 (ground beef)
    -0.916 (beef)
    -0.869 (beef)
    -0.850 (beef)
    -0.85 (ground beef) jj
    -0.84 (beef) ||
    -0.83 (fed beef; Freebaim and Rausser, 1975)
    -0.811 (table-cut beef)
    -0.80 (table-cut beef)
    -0.76 (beef and veal)
    -0.742 (retail beef)
    -0.73 (steaks)
    -0.72 (steak)
                                           D-2
    

    -------
                                        Table D-l (coat.)
    Demand Elasticities for Beef Products Ranked from the Lowest Estimate to the Highest Estimate
    Source
    Brester (1996)
    Bales and Unnevehr (1988)
    Marsh (1991)
    Huang (1993)
    Huang (1986)
    Hahn(1988)
    Bales and Unnevehr (1988)
    Ospina and Shumway (1979)
    Marsh (1992)
    Marsh (1992)
    Arzac and Wilkinson (1979)
    Brester and Wohlgenant (1993) *
    Huang and Hahn (1995) l
    Capps (1989)
    Elasticity Estimate ::]y^: ..:/':.:~ . :"•.,,- 'v. ,
    -0.70 (beef)
    -0.68 (table-cut beef)
    -0.66 (choice slaughter beef)
    -0.6212 (beef and veal)
    -0.6166 (beef and veal)
    -0.58 (beef)
    -0.570 (beef)
    -0.57 (wholesale beef)
    -0.536 (farm beef)
    -0.495 (wholesale beef)
    -0.49 (fed beef)
    -0.45 (beef)
    -0.401 (manufacturing grade beef)
    -0.15 (ground beef)
    As cited in Hahn (1996a).
                                              D-3
    

    -------
                                               Table D-2
       Supply Elasticities for Beef Products Ranked from the Lowest Estimate to the Highest Estimate
    Source
    Marsh (1994)
    Ospina and Shumway (1979)
    Ospina and Shumway (1979)
    Marsh (1994)
    Marsh (1994)
    Marsh (1994)
    Marsh (1994)
    Marsh (1994)
    Marsh (1994)
    Marsh (1994)
    Buhr (1993)
    Elasticity Estimate
    -0.17 (short-run, fed cattle)
    0.06 (steer-heifer fed beef; Folwell and Shapouri, 1977)
    0.14 (slaughter beef)
    0. 14 (all beef; Freebairn and Rausser, 1975)
    0.14 (fed beef; Shuib and Menkhaus, 1977)
    0.200 (wholesale fed beef; Bedinger and Bobst, 1988) . '
    0.23 (fed beef; Langemeier and Thompson, 1967)
    0.606 (intermediate run, fed cattle)
    0.993 (beef; Tvedt, et. al., 1991)
    3.24 (long-run, fed cattle)
    9.505 (beef, long-run - 5 years) '
    1 The estimate is not comparable to the other elasticity estimates. The reported figure is the impact of a 10
    percent change in farm price rather than the standard 1 percent. Given the nonlinear nature of the system,
    the figure cannot be translated into a standard elasticity estimate via division by 10.
                                                   D-4
    

    -------
                                      Table D-3
    Demand Elasticities for Pork Ranked from the Lowest Estimate to the Highest Estimate
    Source :
    Eales and Unnevehr (1993)
    Kesavan et. al. (1993) l
    Gao and Shpnkwiler (1993) ]
    Arzac and Wilkinson (1979)
    Moschini and Meilke (1989)
    Huang and Hahn (1995) '
    Huang (1994)
    Capps (1989)
    Eales and Unnevehr (1993)
    Lemieux and Wohlgenant (1989)
    Hahn (1988)
    Brester and Wohlgenant (1991)
    Brester and Wohlgenant (1991)
    Eales and Unnevehr (1988)
    Huang (1986)
    Huang (1993) :
    Chavas (1983)
    Chavas (1983)
    Capps (1989)
    Moschini, Moro, and Green (1994)
    Hahn (1994) '
    Brester and Schroeder (1995)
    Bales and Unnevehr (1988)
    Eales et. al. (1998)
    Wohlgenant (1989)
    Capps and Schmitz (1991)
    Elasticity Estimate
    -1 .234 (pork - AIDS with SI)
    -0.99 (pork - long-run)
    -0.95 (pork)
    -0.87 (pork) •
    -0.839 (pork)
    -0.838 (pork)
    -0.8379 (pdrk)
    -0.8279 (pork loin)
    -0.801 (pork - AIDS without SI)
    -0.80 (pork)
    -0.784 (pork)
    -0.779 (pork - ground beef model)
    -0.775 (pork - nonfed model)
    -0.762 (pork - aggregate system)
    -0.7297 (pork)
    -0.7281 (pork)
    -0.723 (pork - SC)
    -0.7 14 (pork -WSC)
    -0.7005 (pork chops)
    -0.68 to -0.72 (pork)
    -0.699 (pork)
    -0.69 (pork)
    -0.565 (pork - disaggregated system)
    -0.52 (pork)
    -0.51 (pork - unrestricted)
    -0.45 10 (pork)
                                         D-5
    

    -------
                                         Table D-3 (cont.)
          Demand Elasticities for Pork Ranked from the Lowest Estimate to the Highest Estimate
    Source
    Wohlgenant(1989)
    Capps (1989)
    Capps(1989) . .
    Alston and Chalfant (1993)
    Alston and Chalfant (1993)
    Elasticity Estimate
    -0.36 (pork - restricted)
    -0.3596 (ham)
    -0.2639 (composite pork commodity)
    -0.17 (pork -Rotterdam) ,
    -0.07 (pork - AIDS)
    1 As cited in Hann (1996a).
                                               D-6
    

    -------
                                                 Table D-4
            Supply Elasticities for Pork Ranked from the Lowest Estimate to the Highest Estimate
    •Source-.;.1.";" "" --;::- .;••;.'; -.:•-_-: •'. - .-•: .;•: W^L.
    Elasticity Estimate .' •'• .":' .:-'•'.;-; ;-;:''-.. . .". ' /;., ..."• .. .
    Short-Run
    Holt and Johnson (1988)
    Heien (1975)
    Meilke et. al. (1974)
    Meilke et. al. (1974)
    Lemieux and Wohlgenant (1989)
    Buhr (1993)
    0.007 (pork, short-run - 3 quarters)
    0.09 (pork)1
    0.16 (hog, short-run - GDL)
    0.17 (hog, short-run - PDL)
    0.4 (pork, short-run)
    2.63 (pork, short-run - 1 quarter) 2
    Intermediate-Run
    Meilke et. al. (1974)
    Holt and Johnson (1988)
    Lemieux and Wohlgenant (1989)
    0.24 (hog, intermediate-run - PDL)
    0 338 (pork, intermediate-run - 10 quarters)
    1.8 (pork, intermediate-run
    . Long-Run
    Meilke et.' al. (1974)
    Meilke et. al. (1974)
    Holt and Johnson (1988)
    Buhr (1993)
    0.43 (hog, long-run - GDL)
    0.48 (hog, long-run - PDL)
    0.628 (pork, long-run - 40 quarters)
    7.35 (pork, long-run - 5 years) 2
    1 The reported figure is the elasticity of total number of pigs slaughtered with respect to the ratio of farm to
    retail price of pork.
    2 The estimate is not comparable to the other elasticity estimates. The reported figure is the impact of a 10
    percent change in farm price rather than the standard 1 percent. Given the nonlinear nature of the system,
    the figure cannot be translated into a standard elasticity estimate via division by 10.
                                                   D-7
    

    -------
                                             Table D-5
         Demand Elasticities for Broilers/Chickens Ranked from the Lowest to the Highest Estimate
    Source
    Kesavan et. al. (1993) !
    Aizac and Wilkinson (1979)
    Alston and Chalfant (1993)
    Bales and Unnevehr (1988)
    Capps (1989)
    Eales and Unnevehr (1988)
    Huang (1986)
    Gao and Shonkwiler (1993) '
    Huang (1993)
    Hahn (1994) '
    Bales and Unnevehr (1988)
    Eales and Unnevehr (1993)
    Huang and Hahn (1995) l
    Huang (1994)
    Eales and Unnevehr (1993)
    Eales et. aU (1998)
    Bales et. al. (1998)
    Hahn (1988)
    Eales et. al. (1998)
    Moschini and Meilke (1989)
    Elasticity Estimate
    -1.25 (chicken - long-run)
    -0.98 (chicken)
    -0.94 (chicken - AIDS and Rotterdam)
    -0.677 (chicken - whole bird)
    -0.6557 (chicken)
    -0.610 (chicken - parts/processed)
    -0.5308 (chicken)
    -0.47 (chicken)
    -0.3723 (chicken)
    -0.299 (chicken)
    -0.276 (chicken)
    -0.233 (chicken - AIDS with SI)
    -0.197 (broiler)
    -0.1969 (broiler)
    -0.162 (chicken - AIDS without SI)
    -0.15 (chicken -Model 3)
    -0. 14 (chicken - Model 1)
    -0.140 (chicken)
    -0. 13 (chicken - Model 2)
    -0.104 (chicken)
    1 As cited in Hahn (1996a).
                                               D-8
    

    -------
                                                Table D-6
         Supply Elasticities for Broilers/Chickens Ranked from the Lowest to the Highest Estimate
    Source
    Elasticity Estimate
    Short-Run
    Chavas and Johnson (1982)
    Chavas (1982)
    Holt and Aradhyula (1990)
    Holt and Aradhyula (1990)
    Aradhyula and Holt (1989)
    Holt and Aradhyula (1990)
    Buhr (1993)
    0.064 (broiler, short-run)
    0.072 (broiler, short-run) 3
    0.216 (broiler, short-run-adaptive expectations) '
    0.232 (broiler, short-run - GARCH) l .
    0.305 (broiler, short-run) l
    0.399 (broiler, long-run - adaptive expectations) 'l
    0.49 (chicken, short-run - 1 quarter) 2 .
    Long-Run
    Holt and Aradhyula (1990)
    Holt and Aradhyula (1990)
    Buhr (1993)
    0.399 (broiler, long-run - adaptive expectations) l
    0.587 (broiler, long-run .- GARCH) '
    0.68 (chicken, long-run - 5 years) 2.
    1 The reported elasticity figure is based on the expected rather than the actual mean price of broilers.
    2 The estimate is not comparable to the other elasticity estimates. The reported figure is the impact of a 10
    percent change in farm price rather than the standard 1 percent. Given the nonlinear nature of the system,
    the figure cannot be translated into a standard elasticity estimate via division by 10.
    3 The reported figure is the elasticity of supply with respect to the one-quarter lagged product price.
                                                   D-9
    

    -------
                                            Table D-7
         Demand Elasticities for Turkey Ranked from the Lowest Estimate to the Highest Estimate
    Source
    Huang (1986)
    Bales et. al. (1998)
    Huang (1993)
    Hahn (1994) '
    Soliman (1971)
    Soliman (1971)
    Soliman (1971)
    Soliman (1971)
    Elasticity Estimate
    -0.6797 (turkey)
    -0.63 (turkey - Model 1)
    -0.5345 (turkey)
    -0.459 (turkey)
    -0.412 (turkey - 3SLS) .
    -0.411 (turkey -LISE)
    -0.394 (turkey - 2SLS)
    -0.372 (turkey - OLS)
    1 As cited in Hahn (1996a).
                                              D-10
    

    -------
                                               Table D-8
          Supply Elasticities for Turkey Ranked from the Lowest Estimate to the Highest Estimate
    Source •• . ' ../;•; ••, • ,X'.ป 'v." -.••;.; u; •:.--'. ./.-:/
    Elasticity Estimate; 7 • . '•'--••" >^: •• '- -••v-->? • --;':'•!'•"
    Short-Run
    Chavas and Johnson (1982)
    Chavas (1982)
    Soliman (1971)
    0.210 (turkey, short-run)
    0.222 (turkey, short-run) '
    0.353 (turkey, short-run) 2
    Long-Run
    Soliman (1971)
    0.518 (turkey, long-run) 2
    1 The reported figure is the elasticity of supply with respect to the one-quarter lagged product price.
    2 The reported figure is the elasticity of turkey production with respect to the lagged turkey-feed price ratio.
                                                  D-ll
    

    -------
                  D.2    REFERENCES
    
                  Alston, J.M. and J.A. Chalfant.  1993. The Silence of the Lambdas: A Test of the Almost Ideal and
                         Rotterdam Models. American Journal of Agricultural Economics. May.
    
                  Aradhyula, S.V. and M.T. Holt.  1989. Risk Behavior and Rational Expectations in the U.S. Broiler
                         Market. American Journal of Agricultural Economics:  November.
    
                  Arzac, E.R. and M. Wilkinson.  1979. A Quarterly Econometric Model of United States Livestock and
                         Feed Grain Markets and Some of Its Policy Implications. American Journal of Agricultural
                         Economics. May.
    
                  Brester, G.W. 1996. Estimation of the U.S. Import Demand Elasticity for Beef: The Importance of
                         Disaggregation. Review of Agricultural Economics. 18(l):31-42. January.
    
                  Brester, G.W. and T.C. Schroeder. 1995. The Impacts of Brand and Generic Advertising on Meat
                         Demand. American Journal of Agricultural Economics. 77(4):969-979.  November.
    
                  Brester, G.W. and M.K. Wohlgenant. 1991. Estimating Interrelated Demands for Meats Using New
                         Measures for Ground and Table Cut Beef. American Journal of Agricultural Economics.
                         73(4):1182-1194. November.
    
                  Buhr, B.  1993.  A Quarterly Econometric Simulation Model of the U.S. Livestock and Meat Sector.
                         University of Minnesota, Department of Agricultural and Applied Economics. Staff Paper P93-
                         12. May. http://agecon.lib.umn.edu/mn/p93-12.pdf
    
                  Capps, O. and J.D. Schmitz. 1991. A Recognition of Health and Nutrition Factors in Food Demand
                         Analysis. Western Journal of Agricultural Economics. 16(l):21-35. July.
    
                  Capps, O. 1989. Utilizing Scanner Data to Estimate Retail Demand Functions for Meat Products.
                         American Journal of Agricultural Economics. 71(3):750-760. August.
    
                  Chavas, J-P.  1983.  Structural Change in the Demand for Meat. American Journal of Agricultural
                         Economics  65(1):148-153. February.
    
                  Chavas, J-P.  1982.  On the Use of Price Ratio in Aggregate Supply Response: Some Evidence From the
                         Poultry Industry.  Canadian Journal of Agricultural Economics. 64(4):345-358. November.
    
                  Chavas, J-P., and S.R. Johnson.  1982. Supply Dynamics: The Case of U.S. Broilers and Turkeys.
                         American Journal of Agricultural Economics. 64(3):558-564. August.
    
                  Choi, S. and K. Sosin. 1990. Testing for Structural Change: The Demand for Meat.  American Journal of
                         Agricultural Economics. 72(l):227-236.  February.
    
                  Bales, J.S., J. Hyde, and L.F. Schrader.  1998. A Note on Dealing with Poultry in Demand Analysis.
                         Journal of Agricultural and Resource Economics. 23(2):558-567. December.
    
                  Bales, J. S. and L. J. Unnevehr.  1993. Simultaneity and Structural Change in U.S. Meat Demand.
                         American Journal of Agricultural Economics. 75(2):259-268. May.
    .
                                                              D-12
    

    -------
     Bales, J. S. and L. J. Unnevehr. 1988.  Demand for Beef-and Chicken Products: Separability and
            Structural Change. American Journal of Agricultural Economics. 70(3):521-532. August.
    
     Hahn, W.F. 1996a. An Annotated Bibliography of Recent Elasticity and Flexibility Estimates for Meat
            and Livestock.  Washington, DC: U.S. Department of Agriculture, Economic Research Service
            Staff Paper 9611. July.
    
     Hahn, W.F. 1988. Effects of Income Distribution on Meat Demand. The Journal of Agricultural
            Economics Research. 40(2): 19-24.  Spring.
    
     Heien,D.M.  1975. An Econometric Model of the U.S. Pork Economy. The Review of Economics and
            Statistics. 57(3):370-375. August.
    
     Holt, M.T. and S.V. Aradhyula.  1990.  Price Risk in Supply Equations: An Application of GARCH Time-
            Series Models to the U.S. Broiler Market.  Southern Economic Journal. 57(l):230-242. July.
    
     Holt, M.T. and S.R. Johnson. 1988. Supply Dynamics in the U.S. Hog Industry.  Canadian Journal of
            Agricultural Economics. 36(2):313-335. July.
    
     Huang, K.S. 1994. A Further Look at Flexibilities and Elasticities. American Journal of Agricultural
          .  Economics. 76(2):313-317. May.                                       .
    
     Huang, K.S. 1993. A Complete System of U. S. Demand for Food. Technical Bulletin Number 1821.
            Washington, DC: U.S. Department of Agriculture, Economic Research Service.
    
     Huang, K.S. 1986. U:S. Demand for Food: A Complete System of Price and Income Effects. Technical
            Bulletin Number 1714. Washington, DC: U.S. Department of Agriculture, Economic Research
            Service.
    
     Lemieux, CM. and M.K. Wohlgenant.  1989. Ex Ante Evaluation of the Economic Impact of Agricultural
            Biotechnology: The Case of Porcine Somatotropin. American Journal of Agricultural Economics
            71(4):903-914.  November.
    
     Marsh, J.M.  1994.  Estimating Intertemporal Supply Response in the Fed Beef Market. American Journal
            of Agricultural Economics. 76(3):444-453. August.
    
     Marsh, J.M.  1992.  USDA Data Revisions of Choice Beef Prices and Price Spreads: Implications for
            Estimating Demand Responses. Journal of Agricultural and Resource Economics.  17(2):323-334.
    
     Marsh, J.M.  1991.  Derived Demand Elasticities: Marketing Margin Methods versus an Inverse Demand
            Model  for Choice Beef. Western  Journal of Agricultural Economics. 16(2):382-391.  December.
    
     Meilke, K.D., A.C. Zwart, and -L. J.Martin.  1974. North American Hog Supply: A Comparison of
            Geometric and Polynomial Distributed Lag Models. Canadian Journal of Agricultural Economics.
    
     Moschini, G., D. Moro, and R.D. Green. 1994.  Maintaining and Testing Separability in Demand
           Systems.  American Journal of Agricultural Economics. 76(l):61-73: February.
    
    Moschini, G. and K.D. Meilke. 1989. Modeling the Pattern of Structural Change in U.S. Meat Demand.
           American Journal of Agricultural Economics. 71(2):253-261. May.
                                               D-13
    

    -------
    Ospina, E. and C.R. Shumway. 1979. Disaggregated Analysis of Short-run Beef Supply Response.
           Western Journal of Agricultural Economics. 4(2):43-59. December.
    
    Soliman, M.A. 1971. Econometric Model of the Turkey Industry in the United States. Canadian Journal
           of Agricultural Economics. 19:47-60. October.
    
    Wohlgenant, M.K.  1989. Demand for Farm Output in a Complete System of Demand Functions.
           American Journal of Agricultural Economics. 71(2):241-252. May.
                                                D-14
    

    -------
                                             APPENDIX E
                                     SENSITIVITY ANALYSES
             EPA performed several analyses of the projected impacts reported in Chapter 5 and 6 to determine
      how sensitive the results are to changes in key assumptions.  Section E.I examines impacts under the
      alternative assumption that facilities are able to pass through to their customers some percentage of
      compliance costs in the form of higher prices. Section E.2 looks at the question of baseline closures, and
      how a potential overestimate of baseline closures may affect results.  Finally, Section E.3 determines how
      projected impacts would differ under the assumption that the distribution of income is not normally
      distributed, but rather is skewed.
     E.1     COST PASS THROUGH
    
            EPA's proposed rule will cause meat processing facilities to incur compliance costs.  These
     increased costs of production will cause a decrease in market supply.  Processors will need to realize a "
     higher price per unit in order to sell the same quantity of output after promulgation of the rule that they sold
     prior to promulgation of the rule.
    
            Figure E-l illustrates how the proposed rule would affect the market for meat products and how
     costs are passed through to customers. Compliance costs shift the supply curve upward by an amount
     equal to the average compliance cost per unit of the proposed rule; this represents the increase in per unit
     revenues meat processors would have to realize in order to be willing to sell the same quantity of meat
     products as they sold prior to regulation.  Consumers, however, are unwilling to pay that much more to
    purchase this meat product and the market moves to a new equilibrium at P-<, QPซ. Price per unit sold is
    higher than the original market price.d*-) _ although not as high as the per unit increase in costs - but
                                                 E-l
    

    -------
                                   Market for Meat Product i
      Pi
    ppost
    ppre
    Shift in
    Market
    Supply
                                          Qpost    Qpre
              D1, S1 = preregulatory market conditions
              D1, S2 = postregulatory market conditions
                   QPTC = pre-requlatory equilibrium price and quantity
                  ^ Qpost _ post-requlatory equilibrium price and quantity
                                                 Figure E-l
    
                         Impact of the Compliance Costs on Market for Meat Product i
                                                    E-2
    

    -------
    fewer unit are sold.  Thus, at least some of the costs of the proposed rule incurred by meat processors are •
    partially offset by an increase in price per unit sold.  That is cost pass through (CPT).1
    
            EPA projected facility level impacts in Chapters 5 and 6 under the conservative assumption that
    CPT is zero.  In this sensitivity analysis EPA will project facility level impacts assuming some percentage
    of compliance costs are passed through to customers in the form of higher prices. EPA will us its market
    model to determine the percentage of costs that are passed through, multiply compliance costs per facility
    by one minus that percentage, then project the ratio of compliance costs to net income, the incremental
    probability of closure, and the number of closures under that scenario.
    
            Conceptually, CPT is measured as described above and as illustrated in Figure E-l:
                                                         /•p post _ p pre\
                              cost pass through  =
                                                   per unit compliance costs
    The price elasticities of supply and demand determine how much price increases relative to per unit
    compliance costs. CPT is the percentage of compliance costs paid by consumers in the form of higher  .
    prices, therefore the percentage of compliance costs incurred by facilities is equal to one minus the CPT
    percentage. For example, if GPT is 40 percent, and compliance costs increase per unit costs by $1, then
    consumers pay $0.40 per unit in higher prices, and producers incur $0.60 per unit in higher costs.
    
            One complication to the calculation of CPT as outlined above occurs in EPA's analysis of the meat
    products industry. EPA's engineering model facilities do not distinguish beef processors from pork
    processors, or broiler processors from turkey processors. Rather the models  distinguish only between red
    meat and poultry. Therefore, EPA first used its market model to calculate CPT individually for the beef,
    pork, broiler, and turkey meat types.  EPA then constructed a CPT estimate for red meat as an average of
    beef CPT and pork CPT weighted by relative market quantities, and a similar weighted average for poultry
       1 Zero CPT is can occur if market price does not increase at all in response to a decrease in supply.  This could
    occur if demand is perfectly elastic (i.e., the demand curve in Figure E-l is horizontal), or if supply is perfectly
    inelastic (i.e., the supply Curve in Figure E-l is vertical).  Empirical studies show that neither is the case in markets
    for meat products (e.g., the price elasticity measures cited in Appendix D).
                                                    E-3
    

    -------
    from the individual CPT measures for broilers and turkey. The CPT estimates used for this sensitivity
    
    analysis are:
    
    
            •       red meat — 43.5 percent
    
            •       poultry — 25.6 percent
    
    
    Thus, EPA assumes for the purpose of this analysis that red meat processors will incur 56.5 percent and
    poultry processors will incur 74.4 percent of compliance costs.
    
    
            Table E-l presents the results of the CPT sensitivity analysis, and includes the results of the zero
    CPT analysis from Table 5-6.2 EPA used upper-bound costs as the basis for this comparison.  As would
    be expected, when compliance costs incurred by the facility are decreased by 25 to 45 percent, impacts are
    smaller. Under the proposed options (BAT 3 for all subcategories except J, for which BAT 2 has been
    proposed), trio ratio of posttax annualized compliance costs to net income is:
                   Subcategory A through D:
    
    
                   Subcategory E through I:
    
    
                   Subcategory J:
    
    
                   Subcategory K:
    
    
                   Subcategory L:
       1.07 percent with CPT
    1.90 percent with no CPT
    
       0.27 percent with CPT
    0.40'percent with no CPT
    
       0.51 percent with CPT
    0.68 percent with no CPT
    
       2.96 percent with CPT
    3.98 percent with no CPT
    
       3.14 percent with CPT
    4.23 percent with no CPT
    The incremental probability of closure is also lower, resulting in smaller potential closure impacts. EPA
    projects that 0.5 (out of 209) facilities may close under the CPT scenario, compared to 0.8 facilities
    projected closures assuming zero CPT.
       2 EPA applied the smaller of the two CPT figures (poultry) to rendering as a more conservative asumption. To
    determine the CPT for mixed processors, EPA weighted the CPT of red meat and poultry by the relative production
    of each meat type by mixed processors (61 percent red meat, 39 percent poultry).
                                                   E-4
    

    -------
    I
     ซS il
     tg 53
    
    
    
     **
     I1
     II
    OH
    et_,  01
    
    ll
    .W PN
    *W 4^
     1
                     II
                     QJ ^5 •
                    >-> s
                    O 01
    
                   Is
       •ง
    H 3 ,ฃ>
    
    III
                     u
                     -•PH. S
                                   <ฃ
                                   2
                                   P-
                    t"
                    Ch
    :-
                                    v
                                   ,<
                                                            CO
                                                            CM
                                         0
                                         o
                                                         O
    
                                                         O
                                             ง
                                              CO
                                              1—
                   SO
                   O
                                                                         0
                                                                         o
                                                                                  Os
    
                                                                                  c-i
                                                                            o
                                                                            oo
                          o
                          0
                                                                                          o
                                                                                          o
                                                                                                           CN
                                                            s;
                                                            r—
                                                      OS
                                                      Os
                                      o\
                                      o
                             o
                             •ee-
                             c o
                                                O
                                                s
    t~-
    co
    CO
                                                                SO
    
                                                                o>
    
                                                                fel
    o
    Os
                             CO
    
                             ^f
                             t—
                             fee-
                                                                                               CN
                                                                                               t-^
    S
    cT
    CO
                                                                                      o
                                                                                      t—
                             CO
    
    
                             CN
                                                     i
                                          o
                                          p
                                          o
                                                                   g
                                                       O
                                                       SC
                                                                      oo
                                                                      cs
                                                                                           o
                                                                                           o
                                                           o
                                                           C3
    
                                                           O
                                             O
    
                                             K
                                                                oc
    
                                                                so"
                             a
    
                             o
                                                                           "O.
                                                                           oc
                                                                           CM
                                    CM
    
    
                                    f>
                    O!
    
                    r~
                    t-~
                 t-~
                 o
                                          8
                                          SO
                                          SO
                                       Os
                                       in
                                       t-;
    
                                       in
                                       CO
                                       OS
                                                  so
                                                  t-~
                                                  t-~
    
                                                  OS
                                                                                   so
                                                                                   SO
                                                          m
                                                                   CO
                                                                   p-
                                                                s
                                                                co
                                                                           CO
                                                                           m
                                                                                                w
                                                                                                   CO
                                                                                                   a.
    

    -------
    r
                i
    
               I
    
               •s
                C3
                bja
                M S
               ฃ! O
    
               If
    
               3 I
               "ซ i
                w 5
    
             fSfl
    
             l^i
             * ai
             •S B <
             S S js
             ~ ซ M
               ซ- 3
               3 P
    
               |l
               •s I
               J3 CM
    
               I
               a
               <
               i>
    *ซ
    aU
               a
               s
    
    
    
    
    
    
    VI
    C
    2
    r5
    H
    S
    04
    t
    u
    p
    ^
    eg
    to
    "c
    C
    
    
    
    cc
    
    
    
    
    
    
    
    
    
    
    .a
    
    cd
    B
    •a
    o
    S
    -e
    H
    C/)
    P-(
    •*•*
    a
    s
    
    *
    
    
    
    
    
    
    
    
    
    
    
    42
    *-, a
    
    
    
    *ซ
    A
    2
    
    
    a * M ซ
    S 8S
    ""^ to S
    2*130
    0 ^ &
    y P*
    a
    •a ,ง "^
    N rt> 'MM
    I M '3
    a B ca
    1 ~
     rt (JD C
    
    is •*•* P
    auli
    cn
    "S <3*ฃ>
    *- S s3
    w a ซ
    ง IS* ซ-
    
    
    
    
    
    
    
    2? 55
    
    A S
    
    
    w
    -8
    .ง
    0
    
    ฃ S 8 "ป
    e w g jg
    12 """ O
    ;ซ S s
    (3
    
    5iir
    O ~ o
    J*; CS C
    to ป••)
    "S
    2
    i
    •d
    
    
    1
    ฃ
    
    .ง• "c
    "7^ S*
    S •
    
    
    CO
    i
    jฃ
    U
    ^ H 
    -2 A S "cr
    fa H .ง ฃ
    s3 ซ S
    l^J ^^ 9
    o
    •33 ^> gซ
    "o S 1
    IS 03 c3
    •4-*
    1
    I
    
    
    X
    •3
    ,0,
    i- S
    S '-S
    1 '=
    
    a
    ^
    n.
    0
    
    
    
    
    
    
    c
    
    
    
    ^
    ^
    — <
    
    
    
    Cf
    CO
    
    
    
    VO
    SN
    oc
    E^O
    
    cc
    V
    c
    V5-
    
    2;
    
    
    
    
    m
    o
    
    
    <^5
    p
    
    
    
    fR
    oo
    to
    
    cn
    S
    c
    
    
    
    o
    
    oo
    00
    
    
    
    
    H
    CQ
    
    
    
    
    
    
    O
    O
    
    
    
    I*
    0
    o
    
    
    
    cn
    o
    
    
    
    g
    ON
    
    
    t~-
    C
    {/
    
    O
    
    
    
    
    o
    o
    
    
    f8
    c
    C3
    
    
    
    f^
    a
    o
    
    cn
    vo
    CM
    ft
    
    
    OO
    VO
    cn
    
    
    
    
    
    CQ
    vo
    o
    
    
    
    
    V
    o
    
    
    
    pฃ>
    r-
    o
    
    
    
    CO
    ON
    cn
    
    
    
    r~-
    cn
    CM
    e'-
    en
    ซy
    
    ON
    O
    1
    V
    
    o
    *—(
    
    
    
    cn
    o
    
    
    fS
    en
    f
    0
    
    
    
    H?
    vo
    ON
    CM
    
    ง
    ฐt
    ^^
    
    S
    oo
    f-
    
    
    
    
    
    g
    CQ
    t —
    en
    to
    
    
    
    
    t—
    O
    
    
    
    f9
    en
    ON
    0
    
    
    
    •f
    to
    
    
    
    t~-
    to
    VO
    VO
    c-
    t^
    
    s
    VI
    iff
    
    vr
    CN
    
    
    
    to
    O
    
    
    fS
    ON
    VO
    c!
    
    
    
    pg
    en
    oo
    cn
    
    ON
    r-
    cn
    ^-
    
    1
    o\
    5
    ifr
    
    
    
    
    
    1
    CQ
    ON
    to
    
    
    
    
    ON
    O
    
    
    
    f9
    c
    ~H
    
    
    
    2
    Vฃ
    to
    
    
    
    f-
    oo
    CM
    2
    tf
    
    CM
    f
    to
    ON
    VC
    
    to
    vn
    c-
    
    
    
    ปo
    o
    
    
    fs
    to
    r-
    o
    
    
    
    i~ฐ
    oo
    *5j-
    
    CJN
    en
    tf*
    
    CM
    r-
    to
    
    
    
    
    
    CQ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    oo
    cn
    
    
    
    
    
    o
    
    
    
    ^o
    O
    o
    
    
    
    to
    to
    o
    
    
    
    c
    r-
    
    
    03
    cn
    S
    
    OO
    cn
    
    
    
    
    j—*
    o
    
    
    fฃ
    CJ
    
    
    
    fs
    g:
    o
    
    |
    S
    
    
    oo
    I
    oo
    cn
    
    
    
    
    1
    cn
    to
    vo
    ,_<
    
    
    
    CN
    
    
    
    t?
    ON
    to
    -H
    
    
    
    2
    oo
    
    
    
    S
    CM
    OO
    V
    
    oc
    C"
    V*
    ON
    (N
    
    
    
    vo
    — '
    
    
    fS
    oo
    
    
    
    fS1
    ~
    vo
    
    o
    en
    vo
    
    
    s
    CM
    cn
    I
    
    
    
    
    
    CS
    *o
    o
    ^H
    
    
    
    ซ
    
    
    
    tp
    to
    r-H
    
    
    
    cn
    to
    vo
    
    
    
    vS
    to
    ff
    
    Vฃ
    Tl-
    CM
    ON
    oc
    
    vo
    CN
    
    
    
    j
    •— <
    
    
    ^
    4^H
    
    
    
    fฎ
    to
    oo
    1;J.
    
    CM
    ON
    iff
    
    |
    
    
    V
    OO
    t--
    Vฃ
    
    VO
    
    
    
    
    
    O
    
    
    T*
    oc
    o
    
    
    
    tS
    5f
    
    
    ^o
    5?
    o
    er
    tf,
    
    vo
    ON
    ON
    vo
    iff
    
    
    
    
    
    ro
    so
    
    
    
    
    
    o
    
    
    
    
    CM
    
    
    
    
    S
    r-
    VO
    
    
    
    ฃ
    O
    oc
    ff\
    
    5
    r-
    o
    
    VO
    
    
    
    
    
    cs
    
    
    t*
    CM
    ON
    O
    
    
    
    0
    ON
    ON
    
    
    $
    in
    cr
    &=*
    
    p^
    
    \o
    f— 1
    
    
    
    <
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
                                                                                              vo
    

    -------
    lป
         -d ..
         U CLi
    
         •"3 fi
         ^ hi*!
    
         '5s
           3 E H
           J^
                1
      S =8 S)
    
    
      a ซ "S
      as S
         ฃ
         o —.
    
       "2 u i
    
       i 8S
           ซ
                  o
                  *
                  ฃ
                  PH
    2 "S fa
    
    
       j
               OS
    
               R
               •be
                        ง
    $2
                                S
                         s
                                  Osj
                                  r- —<
                                  S
                                    CQ
                                      S
                                          a
                                            bf
                                            •a
                                                CO
                                                          
    -------
    E.2    BASELINE CLOSURES
    
           As discussed in Appendix B, EPA used a Census special tabulation to calculate the variance of its
    model facility income measures. Combined with model facility mean income, and the assumption that
    income is normally distributed, these estimated variances result in a relatively high percentage of facilities
    earning negative income (about 25 to 35 percent based on cash flow, see Table B-7). Because negative
    cash flow implies that a facility is a baseline closure, EPA believes its methodology may result in an
    overestimate of variance. Therefore, EPA used an alternative method to estimate variance that would result
    in a smaller percentage of baseline closures, and compared projected impacts under the different estimates
    of variauce. This sensitivity analysis is presented below.
    
           EPA used the U.S. Small Business Administration's "births and deaths" database (U.S. SBA,
    1998) to determine that over the 1995 to 1998 time frame firms have exited the meat products industry
    ("deaths") at a rate of 6.8 percent per year.  Assuming the rats of firms exiting the market is equivalent to
    the percentage of baseline closures, EPA calculated the variance for the mean cash flow of each model
    facility class that would result in a 6.8 percent probability of negative cash flow (maintaining the
    assumption that cash flow is normally distributed).
    
           Figure E-2 illustrates the method used to perform this sensitivity analysis. The curve marked
    "Census Variance" represents the cumulative distribution function of cash flow (with mean cash flow equal
    to $100,000), where the variance is calculated from the Census special tabulation as described in Appendix
    B. This curve intercepts the vertical axis at about 28 percent; thus 28 percent of facilities in this group
    earn negative cash flow. The curve marked "SBA Variance" has identical mean and is also normally
    distributed, but the variance is estimated so that about 7 percent of facilities earn negative cash flow.  EPA
    compared facility level impacts under the alternative estimates of variance using identical estimated average
    compliance costs.
    
           Table E-2 presents the results of this sensitivity analysis. Per facility compliance costs and costs
    as a percent of model facility income are identical; the difference between the two methods occurs in the
    incremental probability of closure and the projected number of closures.  The results display only minor
    variation in projected impacts between the alternative estimates; in some cases impacts are slightly higher,
                                                   E-8
    

    -------
                                             o
                                             a
                                             a
                                                     o
                                                     o
                                                     c
                                                     03
                                             w
                                             a
                                            U  co
     s
     o
     3   V
    
    .**> .-
        -ฃ3
         ซj
         C8
        ,2
         a
         o
    2   a
    "3   3
    ^3  fa
                                                                               o
                                                                               o
                                                                               o
    
                                                                               o
                                                                               o
                                                                                   o
                                                                                   o
                                                                                   o
    
                                                                                   o"
                                                                                   o
                                                                                   en
                                                                                   o   ฃ
                                                                                   o   ฎ
                                                                                   ฐ.-E
    
                                                                                   ง   5
                                                                               o
                                                                               o
                                                                               o
    
    
                                                                               o
                                                                               ,—<
                                                                               •Efl-
                  o
                  o
                                t-~
    
                                o
    in
    
    
    o
                                                                                O
                                                                                o
    

    -------
     1
     o^
     "o u
     ea "a
     II
     f?
     1|
     sf
     •a I
     w a)
     o o
     P. C
     p ca
    ซ*ฃ!
    •ts
    ill
    H |S
     ซs
    ง5
    o
    f
       o
       O
     *i
     s-a
     
    i
    B
    %
    ft
    •a
    a
    "2
    a
    
    3
    CO
    
    
    
    ta
    
    
    
    <
    O!
    
    
    
    
    
    
    
    „
    43
    ซ S
    ซ 5
    2 I"
    ฃ
    
    
    
    ฃ.|
    Hi!
    s •ซ
    •s
    • cu
    o
    "E
    J
    
    g
    a
    8
    O
    
    
    !|1
    /•••i.o
    
    o 3 S S
    *w a
    
    •<*
    •o M
    ฃ> B
    2 -^
    o
    S3
    fa
    
    
    >> &
    c
    a
    e
    EB
    1
    
    
    e
    D
    
    B V •>
    3 j= _ง g 42
    3 E H. .2 s
    list0
    
    PM y
    e> _ •ซ >,ซ
    o w o ;& aj
    a ei M>P3 |
    o.'S •ง ฃ a
    Ou g -3 2
    u S "g •ฃ
    
    el-
    's cj ฃ>
    il?
    <|S
    
    
    Posttax
    
    |
    PH
    
    •s
    tj .g
    ll
    
    c
    <
    1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Q
    o
    L
    ฃ
    cv
    c?
    o
    
    
    
    
    
    o
    CD
    
    
    
    
    1
    CD
    
    
    
    O
    
    
    O
    0
    
    
    
    
    
    f55
    CD
    O
    0
    
    
    0
    o
    
    c
    
    s
    
    
    Vฃ
    
    
    
    CO
    o
    
    
    
    
    
    o
    o
    
    
    
    
    1
    
    
    
    o
    
    
    o
    CD
    
    
    
    
    
    f9
    C
    
    
    ex
    o
    o
    
    sc
    c
    f
    ex
    
    CO
    OS
    c<
    
    
    
    
    
    g
    co
    oo
    CO
    
    
    
    
    0)
    o
    
    
    
    
    CO
    CD
    
    
    
    Os
    
    
    CD
    
    
    
    
    
    fฃ>
    c<-
    o
    
    
    0
    0
    
    550,223
    
    0
    o
    c^
    o
    
    
    
    
    
    CO
    
    t—
    
    
    
    
    in
    o
    
    
    
    
    •*
    CD
    
    
    
    P^
    
    
    C
    
    
    
    
    
    r*
    t-
    O
    
    
    2
    
    st
    C
    O
    s
    c
    ซ
    
    
    
    
    
    t
    CQ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    
    
    
    
    
    o
    CD
    
    
    
    
    OS
    O
    0
    
    
    
    o
    
    
    o
    CD
    
    
    
    
    
    jg
    S
    CD
    
    
    C
    
    a
    
    ot
    oo
    c
    V
    
    o
    vo
    
    
    
    B.
    CO
    0-
    o
    CO
    
    
    
    
    
    ^
    
    
    
    
    r-
    
    
    
    o
    ^
    
    
    T:
    
    
    
    
    
    jg>
    VO
    t-;
    
    
    >n
    c<
    c
    
    Os
    r-
    f*
    c
    oc
    r>
    c-
    
    
    
    
    
    CN
    CO
    EL
    co
    cu
    
    SO
    
    
    
    
    SO
    CD
    
    
    
    
    Os
    
    
    
    OS
    O
    M2
    
    
    S
    
    
    
    
    
    fS
    00
    
    
    2
    o
    t-
    
    oe
    c-
    Ol
    
    t~-
    co
    CO
    oc
    5
    
    
    
    
    
    CO
    CO
    EC
    CO
    cu
    oo
    f-
    
    
    
    
    c—
    CD
    
    
    
    
    CO
    
    
    
    g
    
    
    CD
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    
    
    
    
    
    o
    0
    
    
    
    
    8
    o
    
    
    
    o
    
    
    o
    CD
    F 1
    
    
    
    
    )=9
    m
    CO
    
    
    S
    oo
    
    OS
    cs
    oo
    CO
    v>
    CO
    CN
    oc
    
    
    
    
    
    Tf
    V.
    cu
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    s
    1
    ฃ
    c^
    c?
    
    
    
    c^
    o
    
    
    0
    o
    
    o
    ifr
    
    O
    
    
    Os
    
    
    
    H
    co
    o
    
    
    
    
    
    o
    CD
    
    
    
    
    O
    O
    
    
    
    o
    
    
    o
    CD
    
    
    
    
    
    fS
    s
    CD
    
    
    CD
    
    SO
    s
    
    ^
    a
    
    
    
    
    
    CQ
    O
    
    
    
    
    
    o
    o
    
    
    
    
    VO
    o
    CD
    
    
    
    O
    
    
    CD-
    CD
    
    
    
    
    
    fS
    S
    o
    
    
    o
    
    t-
    s
    
    oo
    CO
    CO
    V)
    
    
    
    
    
    CO
    CQ
    O
    
    
    
    
    
    o
    CD
    
    
    
    
    O
    
    
    
    O
    
    
    o
    0
    
    
    
    
    
    fR
    OS
    CD
    
    
    2
    o
    CN
    
    CS
    c*.
    
    O
    Os
    f—
    O
    c<
    
    
    
    
    
    CQ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    GS
    
    
    
    
    
    CO
    CD
    
    
    
    
    CO
    CD
    
    
    
    OS
    
    
    CO
    CD
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    f*R
    c<-
    CD
    
    
    0
    OC
    0
    
    OS
    V
    r-
    
    so
    0
    CO
    p-
    V)
    
    CO
    CN
    
    
    V.
    tt
    Os
    "*
    
    
    
    
    oo
    
    
    
    
    o
    
    
    
    1
    
    
    oo
    
    
    
    
    
    f^
    oo
    t-
    CD
    
    
    CO
    V
    
    CO
    s
    s
    c-
    
    Os
    t-~
    vo
    CO
    c
    
    
    
    
    
    ex
    EL
    CO
    Qu
    
    CO
    
    
    
    
    CO
    
    
    
    
    Os
    
    -------
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    e
    _0
    -*j
    & w
    O <ซ
    •go
    CS 73
    •&ง
    0 0
    BJOJQ
    ซJ i
    83 CU
    u o.
    •2 ง'
    "^
    T3 I
    cn Q,)
    O O
    3&S
    •s > -c
    Table E-2 (co
    e Impacts by I
    to Census Va
    U T3
    S 0>
    ป fa
    •2 S.
    w S
    "a o
    0 
    13
    •s ป
    g
    oa
    
    
    
    
    
    
    
    
    ^
    
    ' , a
    -.-83
    'C
    8S
    
    
    ซ
    S
    :• V
    - u
    
    - -.a
    ,Cfl
    "" ซ
    •'•e
    
    -' '2
    ป
    
    
    
    
    
    
    . -2
    • ' _8
    81
    
    
    •<
    ' CO
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    '. -S3
    " "O 83
    ซ a
    .M
    
    p_, a
    
    03
    fe
    
    |1 j
    A rf3 5
    
    O oi e
    ฃ *"
    
    
    
    a
    |l
    .ฃ, Jl
    : '^l
    rv
    
    
    
    
    Probability
    Cashflow
    O
    |i
    0 ^
    ^*
    
    
    
    
    a -
    Q r- :
    :i<8-l
    
    ^' fl ^
    
    " ::*-^
    
    
    
    •5
    t
    o
    •-•"S
    a
    W
    
    	 ^
    2
    
    
    
    u
    i Sซ
    :^u
    
    
    r ^
    .. -
    ; ง
    ^•a
    -'"a
    .. E
    
    H
    o
    
    
    
    
    Less loan
    Compliance
    Costs3
    }!|
    b fO **
    
    IS
    
    
    
    r:. . - '•'<
    
    
    
    
    - 1
    
    Number of
    TTncilitipc
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    O
    
    
    
    
    
    
    o
    o
    
    
    
    
    Js
    O
    
    
    
    
    ^-*
    
    
    
    
    ง
    
    
    
    
    
    OO
    o
    0
    o
    >n
    o
    
    
    
    
    
    S
    c
    5
    
    
    0
    VD
    
    ^=
    I
    
    
    3
    
    
    
    
    
    
    ~C^"
    _J
    
    
    
    
    g
    —
    
    
    
    
    S
    
    
    
    
    ^
    
    
    
    
    
    V
    0
    r-
    oe
    
    
    
    
    
    _
    c
    *fl
    
    
    oc
    oc
    c
    
    
    (N
    •e/=
    
    
    
    oo
    
    
    
    
    
    
    
    
    
    
    
    ^
    
    
    
    
    o
    
    
    
    
    
    c
    o
    1
    o
    
    
    
    
    
    o
    
    
    
    o
    V*
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    o
    
    
    
    
    If
    <->
    
    
    
    
    ^*^
    
    
    
    
    o
    CD
    
    
    
    
    
    tn
    c
    0
    CO
    o
    
    
    
    
    
    s
    o
    V
    
    
    e-
    t~
    o
    
    Vs
    
    
    
    
    
    
    
    
    
    
    lf~)
    0
    
    
    
    
    c^
    ซ-ป
    
    
    
    
    ^
    
    
    
    
    C3
    
    
    
    
    
    VO
    o
    o
    
    
    
    
    
    t"-
    c<
    c*"
    */5
    
    
    ON
    i/-j
    ON
    oo
    
    ir^
    •&E
    
    
    
    CO'
    
    
    
    
    
    
    f^.
    o
    
    
    
    
    s
    
    
    
    
    
    ฐ
    
    
    
    
    vq
    
    
    
    
    
    VO
    oc
    o
    ฃ
    V")
    
    
    
    
    
    r-
    xn
    f,
    
    
    \
    ^
    
    ^
    
    i
    
    3s
    
    
    
    
    
    
    Os
    O
    
    
    
    
    S
    •^
    
    
    
    
    s
    
    
    
    
    oo
    
    
    
    
    
    c
    •o
    in
    
    
    
    
    
    t—
    oc
    5
    
    
    g
    v^
    
    VC
    v=
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    
    
    
    
    
    
    ^^
    0
    
    
    
    
    0
    •^
    
    
    
    
    oo
    
    
    
    
    
    
    
    
    
    
    I
    o
    2
    
    
    
    
    
    E
    s
    
    
    oo
    to
    t—
    fS
    
    
    oc
    c7
    p-
    
    >n
    VO
    -1
    
    
    
    
    
    ^H
    cS'
    
    
    
    
    OS
    >n
    •— '
    
    
    
    
    VC
    
    
    
    
    0
    CS
    
    
    
    
    
    2
    s
    oo
    
    
    
    
    
    s
    oc
    
    
    oc
    I'-
    ve
    CM
    
    O
    V
    U.
    V
    Q.
    O
    '-'
    
    
    
    
    
    m
    ^-4
    
    
    
    
    Vs
    '-
    
    
    
    
    r—
    OS
    
    
    
    
    L
    
    
    
    
    
    •^t
    •<:
    CO
    V
    VO
    
    
    
    
    
    p
    o
    c
    V
    
    
    v
    
    c:
    oc
    
    
    
    o
    
    
    
    
    
    
    
    
    
    
    
    O
    o
    
    
    
    
    
    0
    o
    I
    
    
    
    
    
    8
    
    
    
    s
    
    
    
    IT-
    p
    cc
    
    
    
    
    
    
    
    o
    o
    
    
    
    
    1
    o
    
    
    
    
    0
    
    
    
    
    o
    o
    
    
    
    
    
    VO
    o
    0
    o
    
    
    
    
    
    0
    5
    
    
    oo
    5
    OC
    
    
    
    p
    cc
    ^
    
    
    
    
    
    
    ^-<
    0
    
    
    
    
    F:
    o
    
    
    
    
    VO
    
    
    
    
    o
    
    
    
    
    
    o
    c-
    
    
    
    
    
    r^
    Os
    o\
    Os
    
    
    oo
    I
    o
    
    &i
    
    i
    a
    -
    
    
    
    
    
    
    ^H
    o
    
    
    
    
    0
    
    
    
    
    
    VO
    
    
    
    
    o
    
    
    
    
    
    c
    c
    VO
    
    
    
    
    
    VO
    r-
    
    
    c
    r-
    
    ff.
    
    ?
    cc
    i
    
    
    
    
    
    
    1— 1
    o
    
    
    
    
    1
    
    
    
    
    
    VD
    
    
    
    
    0
    
    
    
    
    
    oo
    r—
    vd
    
    
    
    
    
    IT
    ฃ
    
    
    5
    •^
    
    
    
    h
    a
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    OS
    
    
    
    
    
    
    VO
    o
    
    
    
    
    I?
    o
    
    
    
    
    f-
    r-
    
    
    
    
    o
    
    
    
    
    
    o
    d
    c
    
    
    
    
    
    oo
    c*~
    
    
    r-
    c
    t^
    
    
    <
    i
    W
    
    

    -------
                  •a  ..
                  •2  S
                  u  s
                  •S*
                     CS
                    &ซ
                  ซ J3
                  •81
            a
    
            I
    •s I
    .2i,tS
    
    
    
      1
    0
                          '"3
                          a
                          o
                         ••s
                         o
    
                                               CQ
    
                                                    CQ
                                                          CQ
                                                                 8
                                                                          a
                                                                          •a
                                                                    Sฃ
                                                                                   ฃ  o
                                                                                   o  ?p
                                                                                   o •••
                                                                              |
                                                                              •g
                                                                                    I
                                                                                    . 3 •
                                                                                     >,
                                                                                    X>
                                                                                    •a
                                                                                     u
                                                                                                  w
                                                                                   8 -o
                                                                                   ;• rz
    
    
                                                                                   8
                                                                                  •a
                                                                                II!
                                                                                         3
                                                                S g %  S
                                                                o .S ^ ฃ
                                                                    s
                                                                               -
    
    
    
    
    
                                                                              8 g X!  g O1 " C
    
                                                                              t
                                                                  TS ฐ S
                                                                  I I S,
                                                                  ง is
                                                           •"'bs'S.S S — ~ a  o,
    
    
    

    -------
    in others impacts are slightly lower using the "SBA Variance" rather than the "Census Variance." For
    example, under PSES 2 for Subcategory A through D, the incremental probability of closure is slightly
    larger for the model using the SBA variance (1.76 percent) compared to the model using the Census
    variance (1.73 percent). However, under PSES 3, the incremental probability of closure is slightly smaller
    using the SBA variance (1.18 percent compared to 1.19 percent). Intuitively, this suggests that within the
    range of estimated compliance costs per facility relevant to this proposal, the slopes of the two cumulative
    distribution functions are approximately equal. This result cannot be generalized however to different
    ranges of compliance costs or baseline closures.
    E.3     DISTRIBUTIONAL ASSUMPTIONS
    
            As discussed in Appendix B, EPA assumed in its analyses that model facility income measures are
    norma'ly distribuurJ. However, there is reason to suspect, especially for revenues, that the distribtion of
    income Jor each model facility class may be skewed.  That is, more than 50 percent of facilities in a Jass
    earn less than the average class income, and less than 50 percent of facilities earn more than the average
    income. EPA performed two sensitivity analyses, one based on revenues, the other based on cash flow, to
    examine the significance of the distributional assumption for the determination of impacts.
    
            EPA selected the lognormal distribution to use as the alternative to the normal distribution for the
    purpose of this sensitivity analysis. EPA used the same model facility mean income and variance that it
    estimated for the normal distribution in each model class, and applied the following transformation to
    determine mean and variance for the lognormal distribution:
                                          'tax
                                                ^
                                                   ln| 1 + — !
                                                  E-13
    

    -------
    where fe, ax2) are the mean and variance for the normal distribution, and (Mlnx, olnx2) are the transformed
    mean and variance for the lognormal distribution.  Thus, EPA uses equivalent means and variances for the
    two distributions.
    
            Figure E-3 illustrates the alternative distribution assumptions for average model facility revenues
    of SI million using the normal and lognormal cumulative distribution functions. The normal distribution
    shows about 7 percent of facilities earning revenues less than $0, which is consistent with the variance for
    revenues provided by Census to EPA.  The skewness of the lognormal distribution can be observed by the
    fact that about 68 percent of establishments earn less than the mean revenues of $1 million under the   .
     lognormal distribution, compared to 50 percent undei the normal distribution.
    
             Section E.3.1 presents the results of the sensitivity analysis of the projected number of facilities
     incurring compliance costs exceeding specified percentages of revenues (the "sales test") under the
     alternative d^ributional assumptions. Section E.3.2 performs an analysis of closure impacts under the two
     different distributions.
             E.3.1   Sales Test Impacts Under Alternative Distribution Assumptions
    
             Table E-3 presents the results for the sensitivity analysis of sales test impacts under the normal and
      lognormal distribution assumptions (see Section 6.4.3 for further discussion of the sales test).  In general
      _ but not invariably - the.sales test impacts are larger under the assumption that revenues are
      lognormally distributed rather than normally distributed. Under the proposed options (BAT 3 for all
      subcategories except J, for which BAT 2 is proposed), EPA projects  that 19.4 facilities (of 209) would
      incur compliance costs exceeding one percent of revenues based on the lognormal distribution, while 17.9
      facilities would exceed that threshold using the normal distribution.
    
              Note that in Figure E-3, the lognormal distribution shows no facilities earning negative revenues
       (i.e., one cannot take the natural log of a negative number).  While intuitively this .seems an improvement
       over the normal distribution, which suggests 7 percent of facilities earn negative revenues, this result may
       not be entirely reflective of reality either. With the exception of cost centers, it is unlikely a facility would
                                                     E-14
    

    -------
                                                            c
                                                            o
                                                        1  e
                                                        o   *
                                                        c   c
                                                        o   o
         *.   ง
         o  -3
         52  t3
         en   e<
              SPN
              3
         •    S •
    <
    H
    t  ง
        '
     a
    
         w   S3   ซ
    3  •<   is   >
    0*  ->   09   ซ
    fa
          A   O>
          e   .g
    
          V   D
              CQ
                         s
                                                                                     o
                                                                                     0
                                                                                     0
                                                                                      0
                                                                                      o
                                                                                      0
    
                                                                                      ro.
                                                                                      V3-
                                                                                          o
                                                                                          o
                                                                                          o
                                                                                           ,
                                      f-
                                                         o
                                                         vt
    O
    o
    

    -------
                                   Table E-3
     Sensitivity Analysis of Nonclosure Impacts by Proposal Subcategory and Option
    Lognormal Distribution Compared to Normal Distribution — Upper-Bound Costs
    Option
    Sumber of
    Facilities
    Compliance Cost
    as a Percentage i
    of Model
    Facility
    Revenues 1
    Lognormal Distribution:
    Facilities Incurring Compliance Costs
    Greater Than % of Revenues 2
    "1 Percent
    3 Percent
    5 Percent
    Normal Distribution
    Facilities Incurring Compliance Costs
    Greater Than % of Revenues2
    1 Percent
    3 Percent
    5 Percent
    Suheatepnni A thrnueh D 	
    BAT1
    BAT2
    BATS
    BAT4
    
    PSES1
    PSES2
    PSES3
    PSES4
    66
    
    
    
    
    60
    
    
    
    0.00%
    0.02%
    0.12%
    0.27%
    
    0.02%
    0.46%
    0.30%
    0.36%
    0.0
    0.0
    1.5
    3.3
    
    0.0
    9.6
    3.9
    6.0
    0.0
    0.0
    0.0
    0.0
    
    0.0
    0.7
    0.1
    0.1
    0.0
    0.0
    0.0
    0.0
    
    0.0
    0.1
    0.0
    0.0
    0.0
    0.2
    2.1
    4.8
    
    0.1
    9.1
    5.0
    6.3
    0.0
    0.0
    0.6
    1.3
    
    0.0
    2.1
    1.4
    1.7
    0.0
    0.0
    0.3
    0.7
    
    0.0
    1.3
    0.8
    0.9
    Subcategoty E through I 	
    BAT1
    BAT2
    BATS
    BAT4
    
    PSES1
    PSES2
    PSES3
    PSES4
    19
    
    
    
    
    234
    
    
    
    0.00%
    0.02%
    0.05%
    0.33%
    0.0
    0.0
    0.3
    3.0
    0.0
    0.0
    0.0
    0.3
    0.0
    0.0
    0.0
    0.1
    0.0
    0.1
    0.2
    1.7
    0.0
    0.0
    0.1
    0.5
    
    0.09%
    0.52%
    0.41%
    0.55%
    1.3
    64.3
    50.8
    71.8
    0.0
    14.3
    7.1
    14.6
    0.0
    4.8
    1.8
    4.8
    5.1
    40.8
    30.1
    43.4
    1.6
    11.2
    8.5
    11.9
    0.0
    0.0
    0.1
    0.3
    
    0.9
    6.4
    4.9
    6.8
    Subcategorv J 	
    BAT1
    BAT2
    BATS
    BAT4
    
    PSES1
    PSES2
    PSES3
    PSES4
    21
    
    
    
    0.00%
    0.17%
    1.85%
    2.02%
    o.a
    2.3
    18.2
    18.5
    L 0.0
    0.2
    10.8
    11.5
    0.0
    0.0
    • 6.7
    7.3
    0.0
    0.9
    10.7
    11.4
    
    75
    
    
    
    0.12%
    2.04%
    2.47%
    2.60%
    4.3
    65.8
    68.5
    69.1
    0.3
    40.6
    46.3
    47.7
    0.0
    26.3
    31.6
    33.2
    2.2
    40.7
    46.9
    48.4
    0.0
    0.3
    3.3
    3.7
    
    0.6
    13.4
    16.5
    17.4
    0.0
    0.2
    1.8
    2.1
    
    0.3
    7.6
    9.4
    9.9
    Subcategorv K 	 	 	
    BATI
    BAT2
    BATS
    BAT4
    BATS
    
    88
    
    
    
    
    
    0.00%
    0.04%
    0.43%
    0.54%
    0.59%
    0.0
    0.0
    12.2
    19.5
    ' 22.5
    0.0
    0.0
    0.4
    1.0
    1.4
    0.0
    0.0
    0.0
    0.1
    0.2
    0.0
    0.6
    12.2
    16.9
    19.2
    0.0
    0.0
    2.8
    3.6
    4.2
    0.0
    0.0
    1.4
    1.8
    2.2
    
                                      E-16
    

    -------
                                                Tabte E-3 (cont.)
                 Sensitivity Analysis of Nonclosure Impacts by Proposal Subcategory and Option
                Lognormal Distribution Compared to Normal Distribution — Upper-Bound Costs
    Option
    PSES1
    PSES2
    PSES3
    PSES4
    dumber of
    Facilities
    138
    
    
    
    Compliance Cost
    as a Percentage
    : of Model
    Facility
    •^Revenues1 ;
    0.06%
    0.94%
    0.67%
    0.70%
    •' . . .".,•.'..-'.',.•
    Lognormal Distribution
    Facilities Incurring Compliance Costs
    Greater Than % of Revenues2 •'
    1: Percent
    0.0
    61.2
    43.5
    45.9
    : 3 Percent
    0.0
    10.5
    3.3
    3.4
    . 5 Percent
    0.0
    '• 3.2
    0.5
    0.6
    - • Normal Distribution . .;,
    Facilities Incurring Compliance Costs
    Greater Than % of Revenues 2
    IPercent
    1.3
    .. 50.0
    35.6
    ' 37.3
    --• 3 Percent
    0.4
    12.7
    7.5
    7.8
    -5 Percent
    0.2
    6.5
    3.9
    4.1
    Subcategory-L - 	 	 	 — 1ฑ-^ 	
    BAT1
    BAT2
    BAT3
    BAT4
    BATS
    
    PSES1
    PSES2
    PSES3
    PSES4
    15
    
    
    
    13 3
    0.00%
    0.05%
    0.48%
    0.69%
    0.75%
    0.0
    0.0
    3.1
    6.0
    5.6
    0.0
    0.0
    0.1
    0.4
    0.3
    0.0
    0.0
    0.0
    0.0
    0.0
    0.0
    0.1
    2.5
    4.0
    4.0
    0.0
    0.0
    0.4
    0.8
    0.8
    
    20S
    
    
    
    0.18%
    1.15%
    0.82%
    1.05%
    2.0
    138.9
    102.7
    128.5
    0.0
    25.5
    9.7
    20.4
    0.0
    5.7
    1.7
    4.6
    8.8
    110.1
    70.9
    97.4
    2.4
    23.2
    14.7
    20.3
    0.0
    0.0
    0.2
    0.4
    0.4
    
    •1.4
    11.7
    7.7
    10.4
    Total Excludins 65 Certainty Facilities
    BAT1
    BAT2
    BATS
    BAT4
    BATS
    
    PSES1
    PSES2
    PSES3
    PSES4
    209
    
    
    
    101 3
    NA
    NA
    NA
    NA
    NA
    0
    2
    35
    50
    28
    0
    0
    11
    13
    2
    0
    0
    7
    8
    0
    0
    2
    28
    39
    23
    0
    0
    7
    10
    5
    
    715
    
    
    
    NA
    NA
    NA
    NA
    8
    340
    269
    321
    0
    92
    66
    86
    0
    40
    36
    43
    18
    251
    188
    233
    5
    63
    49
    59
    0
    0
    L
    t
    •
    
    :
    34
    27
    '32
    Compliance costs as a percent of facility income results are presented as the average for each subcategory, discharge type and
    model facility size combination, weighted by the number of facilities in each combination.
    Number of facilities incurring those impacts is the sum over all facility sizes by subcategory and discharge type.
    1 Ratio of pretax annualized compliance cost to revenues; ratio of posttax annualized compliance costs to cash flow.
    2 Probability compliance costs exceed specified percentage of income measure multiplied by the number of facilities in the
    subcategory size class.                                                             .
    3  Option BAT 5 is only found in Poultry operations.
                                                        E-17
    

    -------
    earn zero revenues, just as it is unlikely that facilities earn negative revenues. Any non-cost center with
    positive production and sales would presumably earn at least some minimal level of revenues, otherwise it
    would not be in business.  However, there is no information available on which to set a benchmark for
    minimum revenues in a model facility class.
           E.3.2   Closure Impacts Under Alternative Distribution Assumptions
    
           EPA performed a similar sensitivity analysis comparing closure impacts under alternative
    distribution assumptions.  One complexity of using the lognormal distribution in the context of the closure
    model is that the lognormal distribution cannot be used with negative values of cash flow. However, unlike
    the revenue model used above (where negative revenues do not make analytic sense), negative cash flow is
    not only logically possible in this context, it is probable.
    
           Figure E-4 illustrates how EPA incorporated negative cash flow into the lognormal model for the
    evaluation of potential closure impacts. EPA used the percentage of baseline closures under the normal
    distribution as a benchmark.  Then EPA calculated the level of cash flow resulting in the same probability
    using the lognormal distribution, and took that as the baseline from which impacts are measured.
    Intuitively, the effect is to shift the lognormal distribution to the left, truncating it at the same probability of
    zero cash flow derived from the normal distribution. This is illustrated in Figure E-4. Note that this
    method probably overestimates the necessary adjustment to the.lognormal distribution.  The reason EPA
    suspects the distribution of cash flow may be skewed in a model class is precisely because of the high
    percentage of baseline closures under the normal distribution. However, for the purpose of this sensitivity
    analysis, this adjustment, is acceptable.
    
            Table E-4 presents projected closure impacts under the alternative assumptions concerning the
    distribution of cash flow. As would be anticipated, given the illustration in Figure E-4, projected
    incremental closures are higher under the lognormal distribution than under the normal distribution. Under
    the proposed options (BAT 3 for all subcategories except J, for which BAT 2 is proposed), EPA projects
    that 4.9 facilities (of 209) would incur compliance costs exceeding cash flow under the lognormal
    distribution, compared to 0.7 facilities exceeding that threshold under the normal distribution.
                                                  E-18
    

    -------
                                                        •^   o
                                                        ฃ   !3
                                                        co   Ml
    
    
                                                        3   S
    
                                                        13   T?
                                                        o
                                                        S5
              ง
          02   gj
          A   s
          O  rฃ
         • |M  P*^
    
    rr   ป  o   ^
    
    
    H   S  '-ง  I
    
    
     2  J  I  S
     S  •<'  ฃ   w
     5P   
    -------
    S
    ^  43  S
        -~ ;-.
    
    
        11
    ฅ
    
    s
    •O
    !i3
    .a
    p
    1
    o
    t5"
    i
    •o
    1*
    el
    •a
    n
    a
    
    •8
    K
    3
    1
    ฃ
    "a
    p
    a
    E
    3
    
    
    „
    .3
    CJ
    •a c3
    4 •*•* I*
    •3^=3 • |
    S 3
    
    JlO n
    a ซ
    ||| 'f 6
    
    
    O Ofi "^
    0) fl "3
    til
    O <" S
    y 
    o
    to
    CO
    00
    
    
    
    
    
    
    
    
    
    
    
    s
    CQ
    OS
    r-
    
    
    
    
    
    tn
    o
    
    
    
    ^
    S
    o
    
    
    
    t—
    CO
    
    
    in
    
    
    
    PR
    CO
    00
    CO
    p^
    00
    in
    CO
    
    
    S
    Tf
    
    s
    OS
    o-
    c
    
    to
    o
    in"
    V,
    V*
    
    
    
    
    
    
    
    
    
    
    CQ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    O
    
    
    
    
    
    
    0
    O
    
    
    
    o
    o
    
    
    
    5?
    
    
    
    0)
    0
    
    
    
    PR
    oo
    
    i
    o
    
    
    r-
    m
    o
    
    &
    5
    
    
    
    
    
    VO
    CO
    ป-l
    
    
    
    1
    ^f
    
    
    o
    
    
    
    ?R
    VO
    VO
    Os
    CO
    vo
    
    
    1
    00
    
    OS
    OS
    ol
    oo
    
    CO
    p!
    ^^
    VO
    oc
    **
    
    
    
    
    
    
    
    
    
    
    K
    a.
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •—
    'Sp
    g
    -s;
    ca
    
    o
    on
    
    
    
    
    p^
    g
    o
    
    
    
    o
    
    
    
    o
    o
    
    
    
    PR
    CO
    0
    PJ5
    01
    o
    
    
    *— 1
    o
    
    vo
    01
    VO
    5
    
    ,_H
    vo
    OS
    
    
    
    
    
    
    
    
    
    
    
    
    CQ
    o
    
    
    
    
    
    
    o
    o
    
    
    
    sR
    VO
    o
    o
    
    
    
    o
    
    
    
    o
    
    
    
    PR
    oo
    CO
    0
    pฃ
    CO
    CO
    o
    
    
    1
    o
    
    01
    oo
    t—
    s
    
    oo
    VO
    m
    CO
    fft
    
    
    
    
    
    
    
    
    
    
    
    CQ
    o
    
    
    
    
    
    
    o
    o
    
    
    
    1
    o
    
    
    
    p
    
    
    
    o
    
    
    
    PR
    o
    r-
    01
    f
    ol
    
    
    2
    Os
    01
    
    oo
    E
    
    0
    OS
    r-
    0
    CO
    
    
    
    
    
    
    
    
    
    
    
    m
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Os
    
    
    
    
    
    
    CO
    C5
    
    
    
    ^
    CO
    T— <
    o
    
    
    
    t—
    00
    
    
    
    .__:
    
    
    
    PR
    vo
    o
    P^
    ฃ
    o
    
    
    SR
    o
    oo
    0
    
    Cs
    E
    
    vo
    en
    13.
    t-~
    
    co
    01
    
    
    
    
    
    
    
    
    I
    OS
    
    
    
    
    
    
    00
    
    
    
    p^
    01
    t—
    o
    
    
    
    01
    o
    vo
    01
    
    
    Os
    
    
    
    PR
    s
    
    pฃ
    t—
    r-
    co
    
    
    CO
    <0
    •*
    
    CO
    s
    s
    
    OS
    
    CO
    I
    
    
    
    
    
    
    
    
    
    
    
    1
    ฃ
    CO
    
    
    
    
    
    CO
    
    
    
    pฃ
    Cs
    o
    
    
    
    ง
    oi
    
    
    Cs
    
    
    
    PR
    OS
    CO
    CO
    Os
    o
    CO
    
    
    01
    f-
    co
    
    r-
    oi
    r-
    cs
    
    Os
    OO
    0
    CO
    en
    
    
    
    
    
    
    
    
    
    
    
    CO
    Ol
    Os
    ^
    
    
    
    
    
    OS
    
    
    
    ซฃ
    T— 1
    OO
    o
    
    
    
    1— 1
    OS
    oi
    
    
    m
    o
    **"*
    
    
    PR
    in
    
    P^
    r— 1
    01
    ••=)•
    
    
    VO
    o
    m
    
    m
    Os
    oo
    s
    
    *n
    I—
    to
    s
    
    
    
    
    
    
    
    
    
    
    
    1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    en
    o>
    1
    eg
    o
    
    
    
    
    
    
    o
    o
    
    
    
    1
    o
    
    
    
    o
    
    
    
    0
    o
    
    
    
    fR
    o
    o
    o
    o
    o
    o
    
    
    0
    0
    
    g
    
    
    en
    vt
    
    
    
    CS
    
    
    
    
    
    
    
    
    
    CQ
    O
    
    
    
    
    
    
    o
    0
    
    
    
    p^
    01
    o
    
    
    
    oo
    
    
    
    o
    
    
    
    PR
    8
    o
    SR
    VO
    >o
    o
    
    
    oo
    vo
    0
    
    oo
    TJ-
    Z
    
    m
    CO
    •et-
    Ol
    -f/5-
    
    
    
    
    
    
    
    
    
    
    
    CO
    T— *
    
    
    
    
    
    
    CO
    o
    
    
    
    ^
    ^^
    
    
    
    00
    
    
    
    T— I
    
    
    
    PR
    r-
    r—
    vo
    Pฃ
    in
    m
    VO
    
    
    P^
    8
    oo
    
    ง
    oo
    VC
    *p.
    
    vn
    t—
    oo
    in
    i/n
    
    
    
    
    
    
    
    
    
    
    
    CQ
    2
    
    
    
    
    
    
    CO
    o
    
    
    
    5R
    Os
    in
    —i
    
    
    
    oo
    
    
    
    in
    
    
    
    ^
    CO
    
    p^
    vo
    r-
    
    
    oo
    t~-
    oo
    
    VO
    oo
    CO
    I
    
    •^jf.
    ON
    1 — 1
    00
    fi
    
    
    
    
    
    
    
    
    
    
    
    1
    CQ
                                                                                                                                                       o
                                                                                                                                                       o)
    
                                                                                                                                                       W
    

    -------
      O
    
      U
    
      •a
      o
    c e?
    2 i
    •C CD
    a, a,
    
    
    
    If
    g -C
    2 •*••ป
    fi* 2
    
    S5
     3 P-
     1 S
    
    
    32
    ^ g
    cc
       I
    T
    •J3 .
    'i2" "'
    
    *s
    'S
    S
    o
    -Vyr^
    C^
    .2
    Ml
    3:.
    c
    1
    43:,
    
    OS
    cn
    C?
    C
    . "^
    •fe
    
    g
    .•g
    EC
    p
    -* •••ซ
    
    e
    c
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    '• - •••.-., •„
    . --,42
    
    .11
    
    S • ^™
    
    •C3
    to
    
    
    ฃป | •:ง
    IB fa H ••
    Ill'
    ฃ ฐ ^ C
    
    
    ฃ
    MM C>
    5 'B 58
    g v p.
    *•* w 9
    ง •••ป ~
    eo 2 •ฃ?
    
    C3
    
    
    
    >-ป fe ฃ3
    is EH
    lie
    
    Cw
    n
    O fifi **
    u a 3
    
    s ง ซ
    3 o fa
    65 a> "3
    •ft, *< •ง
    o ป ง
    U. CS t ^j
    "C
    
    ' *"i-
    I's i
    ' 1 J ฃ
    . 3 3. h
    - ^TS
    u
    
    
    
    V
    "6
    ••• o
    
    :!
    EA
    2
    9
    Q
    Fi
    ,
    
    8ป*>
    a-*
    ง g
    1"
    
    1
    S
    o
    : J
    [s
    w
    g
    ^
    ,- .8
    •' O
    
    
    ai ซ*i i
    11
    g -
    o
    U -
    - 5
    -S
    •g
    - r""
    V.
    
    
    -v
    !
    fi
    c
    1 >
    • c
    (
    c
    
    . , j
    •j
    •' t
    
    
    CM
    O '
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    0
    
    
    
    
    o
    0
    
    
    
    
    
    S
    o
    
    •<*
    
    
    
    
    CO
    o
    
    
    
    
    
    fcS
    c>
    
    
    S
    
    
    
    
    
    • 0
    O
    O
    
    ON
    o
    c
    
    NO
    c
    NC
    •&9
    
    
    in
    r-
    i
    NO
    NO
    
    
    
    
    CM
    ^
    
    
    
    
    
    X)
    m
    
    
    CN
    CO
    
    
    
    
    in
    in
    
    
    
    
    
    ซ
    S
    
    
    c*
    
    
    
    
    
    o
    OO
    r>-
    CO
    
    f
    VO
    oc
    
    OO
    OO
    o
    o
    oc
    tfl
    
    
    c-
    [
    OO
    
    
    
    
    in
    1—3
    
    
    
    
    
    S
    ^^
    
    VO
    
    
    
    
    CO
    vd
    
    
    
    
    
    ?=S
    5
    oc
    
    
    OO
    r-
    oc
    
    
    
    
    
    Tป
    S
    o
    
    CO
    ex
    S
    
    c
    
    &
    
    
    1
    1
    ON
    00
    
    
    
    
    NO
    ^
    
    
    
    
    
    NO
    O
    
    
    85
    
    
    
    
    ,_!
    r-;
    
    
    
    
    
    ^
    ff
    ON
    
    
    V
    
    
    
    
    
    t?
    CO
    — <
    
    o
    ON
    C
    w
    
    t-
    o
    N
    CO
    V
    
    
    c/
    p.
    1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    o
    
    
    
    
    o
    0
    
    
    
    
    
    i
    o
    
    *-*
    
    
    
    
    C3
    o
    
    
    
    
    
    ' t8
    Q
    
    
    1
    o
    
    
    
    
    
    T*
    S
    o
    
    o
    
    
    o
    ffl
    
    
    
    OO
    OO
    p
    pr
    o
    
    
    
    
    o
    0
    
    
    
    
    
    0
    o
    
    OO
    
    
    
    
    ,_<
    o
    
    
    
    
    
    
    ON
    o
    
    
    o
    
    
    
    
    
    t*
    •<*
    CO
    o
    
    CM
    cs
    ON
    *
    
    
    
    
    
    ^
    in
    NC
    
    
    0
    \r
    
    
    
    
    
    t*
    NO
    in
    
    r-
    oo
    CM
    CM
    Vj
    <,
    
    c*"
    in
    ON
    NO
    
    -------
    g
    M
    "3"
    "a
    JS
    •c
    o
    1
    O
    ซฃ*n
    
    I
    <
    i
    i
    in
    
    f
    
    1
    •<
    ง
    istributii
    Q
    €9
    
    0
    a
    c.
    
    ^j
    
    
    
    
    ^r
    42
    ฃ 1
    
    1
    
    
    
    a ^
    5 a
    ^
    PK
    
    •4-1
    a
    o
    a
    "S
    fa*
    1
    o
    O
    
    Piil
    pQ 6
    
    
    *J>
    ^^
    H "5 ^
    c •ง ง
    p •"" *
    1
    
    ^*
    ฃ>
    
    *ซ !
    
    S r
    (ฃ ^
    
    
    *J
    
    
    
    i
    
    3^ 1
    
    g
    |
    _o
    "S
    
    en
    ฃ
    i
    0
    
    
    fc 0 ป "
    5 ^ a -^
    3 ^ '-ง,15
    
    \ T ***
    6
    
    M
    
    OX) "^
    
    ^ "3
    g s
    
    '"H* *i
    1SI
    u <"s
    
    
    
    
    
    •s
    s
    
    
    *
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    a
    S"
    4) t2
    U "H
    ^ CJ
    
    
    |
    fr
    ja
    "5,
    &
    0
    
    tu
    I
    
    X
    a
    
    ?
    Ai
    
    
    X
    s
    ซ
    I
    
    
    *o
    w
    a3 ."
    
    a"
    _ t
    
    iS &
    
    e
    "e
    CM
    $
    
    
    
    VO
    CM
    
    
    
    i— .
    
    
    CO
    10
    CM
    
    
    
    vo
    — <
    
    
    ^
    c
    vd
    
    
    
    t?
    oc
    V:
    
    
    
    
    t&
    c
    
    CM
    C**
    
    o
    s
    
    
    2
    CN
    CO
    CO
    vs-
    
    
    
    
    
    
    
    
    
    
    CO
    c/
    u
    00
    cu
    CO
    00
    >o
    
    
    
    CO
    CO
    
    
    
    CM
    vo
    
    
    OS
    CO
    
    
    
    o
    vd
    
    
    ^
    0
    t-:
    
    
    
    g
    If
    t^-
    
    
    
    
    ^
    VC
    oc
    
    OS
    
    t-
    
    o
    
    
    0
    o
    3
    
    •
    
    
    
    
    
    
    
    
    ^
    d!
    00
    a,
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ,2
    [t
    .s
    (:
    
    Z
    
    
    
    •
    ฃ
    
    
    
    
    
    
    ^
    Z
    
    w
    
    ^
    o
    \r
    
    
    
    vc
    r>
    r
    
    
    
    
    
    
    
    
    
    
    
    •<
    CQ
    CO
    
    
    
    ^
    
    
    
    z
    
    
    VO
    •*"
    
    
    
    CM
    
    
    
    ^
    z
    
    
    
    
    Z
    
    
    
    
    
    
    ^
    Z
    
    ^
    
    ซ•
    VO
    OS
    
    6<
    
    i
    Vฃ
    cx
    
    
    
    
    
    
    
    
    
    
    
    o
    CM
    
    
    IO
    
    
    
    z
    
    
    CO
    CO
    CO
    2
    
    
    
    CM
    CO
    CO
    
    •
    
    Z
    
    
    
    
    Z
    
    
    
    
    
    
    ^ซ
    z
    
    r-
    
    >o
    0
    •U
    
    
    n
    0
    
    >o
    ป— .
    t—
    
    
    
    
    
    
    
    t
    2
    <ฃ
    Z
    
    
    
    
    z
    
    
    
    
    
    
    ^
    z
    
    CO
    o
    >o
    Os
    C)
    CO
    60-
    
    vo
    VD
    t-^
    
    
    
    
    
    
    
    
    
    
    CM
    1
    
    .c
    ^
    •a
    ^
    00
    1
    ง
    'i
    'rl
    
    O
    O
    *co
    
    .a
    1
    1
    "s
    cd
    u
    o.
    &
    &
    ฃ
    u
    e/3
    •3
    o
    bX>
    *-j
    u
    -O
    
    en
    o
    cd
    
    ^
    en
    1
    U-t
    o
    
    
    Ca
    fe
    (U
    •S
    14-4
    o
    a
    CO
    ฃ
    03
    _en
    t— ]
    
    .S
    T3
    U
    C
    
    T3
    !&
    1
    und in Chapter 5, Table 5-6.
    i2
    lysis Results are the same as those :
    g
    <;
    r^
    1
    55
    in
    
    06
    i is only found in Poultry operation
    f]
    <;
    ca
    e
    o
    >o
    

    -------
    E.4    REFERENCES
    U.S. SB A. 1998. Statistics of U.S. Businesses: Firm Size Data: Dynamic Data: Download U.S. industry
           group data, 1990-1998 one year changes and 1990-1995 (U.S. Births, deaths, and job creation by
           U.S. industry group, 1990 - 1998.) U.S. Small Business Administration, Office of Advocacy,
           Available at: http://www.sba.gov/advo/stats/data.html.
                                                  E-23
    

    -------
    

    -------
                                            APPENDIXF
                             COST EFFECTIVENESS ANALYSIS
    F.I    INTRODUCTION
    
           As part of the process of setting effluent limitations guidelines and developing standards, EPA uses
    cost effectiveness calculations to compare the efficiencies of regulatory options for removing priority and
    nonconventional pollutants.1 This cost effec*:veness (CE) analysis presents an evaluation of the technical
    efficiency of pollutant control options for the proposed effluent limitations guidelines and standards for the
    meat products industry based on Best Available Technology Economically Achievable (BAT) and
    Pretreatment Standards for Existing Sources (PSES). BAT standards set effluent limitations on toxic
    pollutants and nnvients for  direct dischargers prior to wastewater discharge directly into a water body such
    as a st- -.. un, river, lake, estuary, or ocean. Indirect dischargers send wastewater to publicly owned
    treatment works (POTW) for further treatment prior to discharge to U.S. surface waters; PSES standards
    set limitations for indirect dischargers on toxic pollutants and nutrients which pass through a POTW.
    
            The analyses presented in this section include a standard cost effectiveness analysis, based on the
    approach EPA has historically used for developing an effluent guideline for toxic pollutants, an analysis of
    the cost reasonableness of nonconventional pollutant removals, and an analysis of the cost effectiveness of
    removing nutrients.  This expanded approach is necessary to evaluate the broad range of pollutants in meat
     slaughtering and processing wastewater, for which nutrients, conventional pollutants, and nonconventional
     pollutants may be more significant than toxic pollutants. EPA's standard CE analysis is used for analyzing
     the removal of toxic pollutants.  EPA's standard CE analysis does not adequately address removals of
     nutrients, total suspended solids, and pathogens.  To account for the estimated removals of nutrients under
     the proposed meat products regulation in the analysis, the Agency has developed an alternative approach to
        1 A list of priority ("toxic") and conventional pollutants are defined in 40 CFR Part 401. There are more than 120
     priority pollutants, including metals, pesticides, and organic and inorganic compounds.  Conventional pollutants
     include bioiogical oxygen demand (3OD), total suspended solids (TSS), pH, fecal coliform, and oil and grease.
     Nonconventional pollutants comprise all other pollutants, including nutrients (i.e., they do not include conventional
     and priority pollutants).
                                                    F-l
    

    -------
    evaluate the pollutant removal effectiveness of nutrients relative to cost. Although pathogens maybe an
    important constituent of meat processing wastewater, EPA has not at this time developed an approach that
    would allow a similar assessment of pathogen removals.
    
           The organization of this chapter is as follows. Section F.2 discusses EPA's standard cost
    effectiveness methodology and presents the results of this analysis; this section also identifies the pollutants
    included in the analysis, presents EPA's toxic weighting factors for each pollutant, and discusses POTW
    removal factors for indirect dischargers. Section F.3 explains the cost reasonableness analysis and presents
    the results of this analysis. Section F.4 discusses EPA's cost effectiveness methodology for nutrients and
    contains the results of the nutrients cost effectivenest, analysis. Section F.5 contains supplementary data
    tables, while Section F.6 lists references.
    F.2     COf-T EFFECTIVENESS METHODOLOGY AND RESULTS: TOXIC POLLUTANTS
    
            F.2.1   Overview
    
            Cost effectiveness is evaluated as the incremental annualized cost of a pollution control option in
    an industry or industry subcategory per incremental pound equivalent of pollutant (i.e., pound of pollutant
    adjusted for toxicity) removed by that control option. EPA uses the cost effectiveness analysis primarily to
    compare the removal efficiencies of regulatory options under consideration for a rule.  A secondary and less
    effective use is to compare the cost effectiveness of the proposed options for the meat products industry to
    those for effluent limitation guidelines and standards for other industries.
    
            To develop a cost effectiveness study, the following steps must be taken to define the analysis or
    generate data used for calculating values:
                   Determine the pollutants effectively removed from the wastewater.
                   For each pollutant, identify the toxic weights and POTW removal factors. (The first
                   adjusts the removals to reflect the relative toxicity of the pollutants while the second
                   reflects the ability of a POTW or sewage treatment plant to remove pollutants prior to
                   discharge to the water. These are described in Sections F.2.2 and F.2.3.)
                                                   F-2
    

    -------
           •       Define the regulatory pollution control options.
           •       Calculate pollutant removals for each pollution control option.
                   Calculate the product of the pollutant removed (in pounds), the toxic weighting factor, and
                   the POTW removal factor. The resultant removal is specified in terms of  "pounds
                   equivalent" removed.
           •       Determine the annualized cost of each pollution control option.
           •       Calculate incremental CE for options.
    
    Table F-l presents the pollutants, their toxic weights, and POTW efficiency and removal factors used in
    the CE calculations for toxic pollutants as well as conventional and nonconventional pollutants.
            F.2.2  Toxic Weighting Factors
    
            Cost effectiveness analyses account for differences in toxicity among the pollutants using toxic
    weighting factors. Accounting for these differences is necessary because the potentially harmful effects on
    human and aquatic life are specific to the pollutant. For example, a pound of zinc in an effluent stream has
    a significantly different, less harmful effect than a pound of PCBs.  Toxic weighting factors for pollutants
    are derived using ambient water quality criteria and toxicity values. For most industries, toxic weighting
    factors are developed from chronic freshwater aquatic criteria.  In cases where a human health criterion has
    also been established for the consumption of fish, the sum of both the human and aquatic criteria are used
    to derive toxic weighting factors. The factors are standardized by relating them to a "benchmark" toxicity
    value, which was based on the toxicity of copper when the methodology was developed.2
    
            Examples of the effects of different aquatic and human health criteria on freshwater toxic
    weighting factors are presented in Table F-2. As shown in this table, the toxic weighting factor is the sum
    of two criteria-weighted ratios: the former benchmark copper criterion divided by the human health
        2 Although the water quality criterion has been revised (to 9.0 /ig/1), all cost effectiveness analyses for effluent
     guideline regulations continue to use the former criterion ot b.b fig/1 as a benchmark so that cost effectiveness values
     can continue to be compared to those for other effluent guidelines. Where copper is present in the effluent, the revised
     higher criterion for copper results in a toxic weighting factor for copper of 0.63 rather than 1.0.
                                                     F-3
    

    -------
                                             Table F-l
                Toxic Weighting Factors and POTW Efficiency and Removal Factors for
                            Meat Products Industry Pollutants of Concern
    POLLUTANT
    TOXICS
    Ajnmonia as Nitrogen
    Jarium
    Carbaryl
    Chromium
    Cis-permethrin
    Copper
    Manganese
    Molybdenum
    Sfickel
    Citrate/Nitrite
    Titanium
    Trans-permethrin
    Vanadium
    Zinc
    NUTRIENTS
    Total Phosphorus
    Total Nitrogen
    Total Kieldahl Nitrogen (TKN) '
    CONVENTIONALS
    5-Day Biochemical Oxygen Demand (BOD)
    rlexane Extractable Material (HEM)
    Total Suspended Solids (TSS)
    NONCONVENTIONALS
    Chemical Oxygen Demand (COD)
    rfexane Extractable Material (HEM)
    Citrate/Nitrite
    Total Nitrogen
    PATHOGENS
    Fecal Coliform (million cfu/day)
    Toxic
    Weighting
    Factor
    
    1.8e-03
    2.0e-03
    2.8e+02
    7.6e-02
    4.5e+00
    6.3e-01
    7.0e-02
    2.0e-01
    l.le-01
    6.2e-05
    2.9e-02
    4.5e+00
    6.2e-01
    4.7e-02
    
    NA
    NA
    NA
    
    NA
    NA
    NA
    
    NA
    . NA
    6.2e-05
    NA
    
    NA
    POTW
    Efficiency
    Factor
    
    • 38.9%
    16.0%
    30.0%
    80.3%
    50.0%
    . 84.2%
    35.5%
    18.9%
    51.4%
    90.0%
    91.8%
    50.0%
    9.5%
    79.1%
    
    57.4%
    UNK
    57.4%
    
    89.1%
    86.1%
    89.6%
    
    81.3%
    86.1%
    90.0%
    UNK
    
    • 99.6%
    POTW
    Remova
    Factor
    
    6.1e-01
    . 8.4e-01
    7.0e-01
    • 2.0e-01
    5.0e-01
    1.6e-01
    6.4e-01
    8.1e-01
    4.9e-01
    l.Oe-01
    8.2e-02
    5.0e-01
    9.0e-01
    2.1e-Ql
    
    4.3e-01
    UNK
    4.3e-01
    
    l.le-01
    1.4e-01
    l.Oe-01
    
    1.9e-01
    1.4e-01
    l.Oe-01
    UNK
    
    4.0e-03
    1 TKN is used to calculate Total Nitrogen for baseline loads.
                                                F-4
    

    -------
                                               Table F-2
                                  Examples of Toxic Weighting Factors
                             Based on Copper Freshwater Chronic Criteria
    Pollutant
    Copper*
    Cadmium
    Naphthalene
    Human Health
    Criteria
    (MfflD
    1,200
    84
    21,000
    Aquatic
    Chronic
    Criteria (/ig/1)
    9.0
    2.2
    370
    Weighting
    Calculation
    5.6/1,200 + 5.6/9.0
    5.6/84 + 5.6/2.2
    5.6/21,000 + 5.6/370
    Toxic
    Weighting
    Factor
    0.63
    2.6
    0.015
    '* The water quality criterion has been revised (to 9.0 /*g/l). Formerly, the weighting factor calculation led
      to a result of 0.47 as a toxic weighting factor for copper.
    
     Notes:     Human'health and aquatic chronic criteria are maximum contamination thresholds. Units for
               criteria are micrograins of pollutant per liter of water.
                                                    F-5
    

    -------
                  criterion for the particular pollutant and the former benchmark copper criterion divided by the aquatic
                  chronic criterion.  For example, using the values reported in Table F-2, four pounds of the benchmark
                  chemical (copper) pose the same relative hazard in freshwater as one pound of cadmium because cadmium
                  ' has a freshwater toxic weight four times greater than the toxic weight of copper (2.6 divided by 0.63 equals
                  4.13).
                          F.2.3   POTW Removal Factors
    
                          Calculating pound or pound equivalent removals for direct dischargers differs from calculating
                  removals for indirect dischargers because of the ability of POTWs to remove certain pollutants.  The
                  POTW removal factors are used as follows: if a facility is discharging 100 pounds of chromium in its
                  effluent stream to a POTW and the POTW has a 80 percent removal efficiency for chromium, then the
                  chromium discharged to surface waters is only 20 pounds (1 minus 0.8 equals 0.2).  If the regulation
                  reduces chromium discharged in the effluent scream to the POTW by 50 pounds, then the amount
                  discharged to surface waters is calculated as 50 pounds multiplied by the POTW removal factor (50
                  pounds times 0.2 equals 10 pounds). The cost effectiveness calculations then reflect the fact that the actual
                  reduction  of pollutant discharged to surface water is not 50 pounds  (the change in the amount discharged to
                  the POTW), but 10 pounds (the change in the amount actually discharged to surface water). A pollutant
                  discharge that is unaffected by the POTW has a removal factor of 1.
                          F.2.4  Pollutant Removals And Pounds Equivalent Calculations
    
                          The pollutant loadings have been calculated for each facility under each regulatory pollution
                  control option for comparison with baseline (i.e., current practice) loadings. Pollutant removals are
                  calculated simply as the difference between current and post-treatment discharges. For toxic pollutants,
                  these removals are converted into pounds equivalent for the cost effectiveness analysis. For direct
                  dischargers, removals in pounds equivalent for toxic pollutants are calculated as:
    
                                        Removals  =  Removals ^^ x Toxic weighting factor
    
                                                                 F-6
    _
    

    -------
    For indirect dischargers, removals in pounds equivalent for toxic pollutants are calculated as:
              Removals e =  Removalspounds x Toxic weighting factor x POTW removal factor
    Total removals for each option are then calculated by adding up the removals of all pollutants included in
    the cost effectiveness analysis for a given subcategory for both toxic pollutants and nutrients.
            F.2.5  Calculation Of Incremental Cost Effectiveness Values
    
            Cost effectiveness ratios are calculated separately for direct and indirect dischargers and by
     subcategory. Within each of these many groupings, the pollution control options are ranked in ascending
     order of pounds equivalent removed.  The incremental cost effectiveness value for a particular control
     option is calculated as the ratio of the incremental annual cost to the incremental pounds equivalent
     removed. The incremental effectiveness may be viewed primarily in comparison to the baseline scenario
     and to other regulatory pollution control options.  Cost effectiveness values are reported .in units of dollars
     per pound equivalent of pollutant removed.
    
            For the purpose of comparing cost effectiveness values of options under review to those of other
     promulgated rules, compliance costs used in the cost effectiveness analysis are adjusted to 1981 dollars
     using Engineering News Records Construction Cost Index (CCI; ENR 2QOO). The adjustment factor is
     calculated as follows:
    
                        Adjustment factor = 1981 CCI / 1999 CCI = 3535 / 6059 = 0.583
    
     The equation used to calculate the incremental cost effectiveness of option k is:
                                          CEk =
                                                 ATCk -
                                                    F-7
    

    -------
    where:
           CEk   =      Cost effectiveness of Option k
           ATCk  =      Total pretax annualized treatment cost under Option k
           PEk   =      Pounds equivalent removed by Option k
    
           Cost effectiveness measures the incremental unit cost of pollutant removal of Option k (in pounds
    equivalent) in comparison to Option k-1. The numerator of the equation, ATCk minus ATCk.i, is simply
    the incremental annualized treatment cost in moving from Option k-1 to Option k. Similarly, the
    denominator is the incremental removals achieved in going from Option k-1 to k. The lower the value of
    the incremental CE calculation, the lower the cost of each additional pound equivalent of pollutants
    removed under that option.
           F.2.6   Cost-Effective Results for Toxic Pollutants
    
           F.2.6.1 Subcategory Cost Effectiveness
    
           Table F-3 shows the average and incremental CE figures for nonsmall direct (BAT) and indirect
    (PSES) dischargers in all subcategories using upper-bound costs (see the introduction to Chapter 5 for the
    distinction between upper-bound and retrofit costs). For direct dischargers, incremental CE ranges from
    $45 per pound under BAT 2 in Subcategory K to a high of $286,000 for BAT 3 in Subcategory A through
    D.3 Cost effectiveness for indirect dischargers ranges from a low of $17 under PSES 1 for Subcategories
    A through D and K to a high of $31,000 for PSES 4 under Subcategory A through D. Note that negative
    CE values can occur if either estimated annualized compliance costs or estimated pollutant removals are
    lower for option k than for option k-1. This can be observed in Subcategory E through I, for example,
    where costs for PSES 3 are lower than for PSES 2, and pollutant removals for PSES 4 are lower than for
    PSES 3.
       3 EPA determined that all nonsmall direct dischargers have sufficient treatment in place to meet BAT 1 standards,
    therefore there are no costs or removals associated with that option.
                                                  F-8
    

    -------
                   Table F-3
        Results of Cost Effective Analysis
    Upper-Bound Costs for Nonsmall Facilities
    1 Regulatory
    Option V
    Pretax Annualized
    Costs
    (Millions of
    $1999)
    Pollutant
    Removals
    (Pounds
    Equivalent)
    Pretax Average
    Cost Effectiveness
    ($1981 Per Pound
    Equivalent
    Removed)
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    Removed)
    Subcategorv A through D
    BAT 2
    BATS .
    BAT 4
    
    PSES 1
    PSES 2 .
    PSES 3
    PSES 4
    $9.93
    $59.52
    $117.98
    93,586
    93,687
    94,195
    
    $7.05
    $151.49
    $96.25
    $120.64
    240,421
    310,768
    309,081
    309,541
    $62
    $371
    $731
    
    $17
    $284
    $182
    $227
    .$62
    $286,414
    $67,154
    
    $17
    $1,198
    $19,107
    $30,955
    Subcategorv E through I
    BAT 2
    BATS
    BAT 4
    
    PSES 1
    PSES 2
    PSES 3
    PSES 4
    $0.40
    $0.69
    $7.01
    2,609
    2,618
    2,615
    $90
    $154
    $1,564
    
    $18.79
    $102.09
    $83.68
    $110.20
    76,890
    . 78,831
    78,855
    78,813
    $143
    $756
    $619
    $816
    $90
    $18,512
    ($1,261,372)
    
    $143
    $25,036
    ($440,522)
    ($367,437)
    Subcategorv J
    BAT 2
    BATS
    BAT 4
    
    $0.55
    $5.80
    $6.31
    1,550
    1,621
    1,553
    $208
    $2,089
    $2,370
    $208
    $43,028
    ($4,333)
    
                       F-9
    

    -------
                Table F-3 (cbnt.)
        Results of Cost Effective Analysis
    Upper-Bound Costs for Nonsmall Facilities
    Regulatory
    Option
    PSES1
    PSES2
    PSES3
    PSES4
    Pretax Annualized
    Costs
    (Millions of
    $1999)
    $1.33
    $23.25
    $27.91
    $29.22
    Pollutant
    Removals
    (Pounds
    Equivalent)
    3,918
    4,983
    5,112
    4,951
    Pretax Average
    Cost Effectiveness
    ($1981 Per Pound
    :-; Equivalent
    Removed)
    $198
    $2,723
    $3,185
    $3,443
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    Removed)
    $198
    $12,011
    $21,075
    ($4,757)
    Subcategory K
    BAT 2
    BATS
    BAT 4
    BATS
    
    PSES1
    PSES2
    PSES3
    PSES4
    $4.82
    $48.37
    $61.25
    • $66.09
    
    $10.84
    $188.95
    $133.01
    $136.54
    63,192
    64,094
    64,029
    65,169
    
    377,651
    382,550
    382,735
    381,751
    $45
    $440
    $558
    $592
    
    $17
    $288
    $203
    $209
    Subcategory L
    BAT 2
    BATS
    BAT 4
    BATS
    $0.30
    $2.95
    $4.32
    $3.85
    . 373
    383
    371
    398
    $472
    $4,494
    $6,796
    $5,645
    $45
    $28,181
    ($115,860)
    , ' $2,479
    
    $17
    $21,212
    ($176,292)
    ($2,093)
    
    $472
    $160,314
    ($70,689)
    ($10,190)
    
    PSES1
    PSES2
    PSES3
    PSES4
    $15.26
    $105.33
    $74.56
    $94.11
    49,950
    51,257
    51,367
    51,237
    $178
    $1,199
    $847
    $1,072
    $178
    $40,224
    ($162,814)
    ($87,885)
                      F-10
    

    -------
           Average CE tables for non-small direct and indirect dischargers based on retrofit costs are
    presented in Table F-4.4 Option BAT 2 under Subcategory K has the lowest average GE value for a direct
    discharger at $45 and BAT 5 under Subcategory L has the highest average CE at more than $5,600.
    Among indirect dischargers, PSES 1 for Subcategories A through D and K has the lowest average CE at
    $17 and PSES 4 under Subcategory J has the highest at $2,900.
    
           Table F-5 shows the average and incremental CE figures for small direct and indirect dischargers
    in all subcategories using upper-bound costs.5 For small direct dischargers, CE values range from a low of
    $300 under BAT 2 for Subcategory A through D to a high of more than $31 million for BAT 3 hi the same
    subcategory.  Cost effectiveness values for small indirect dischargers range from a low of $39 under PSES
     1 for Subcategory K to a high of $802 million under PSES 3 for Subcategory E through I.
    
            Detailed tables containing toxic pollutant removals and baseline loads for nonsmall and small
     facilities for each subcategory and both discharge types can be found in Section F.5.
            F.2.6.2 Industry Cost Effectiveness
    
            For the proposed options, EPA selected BAT 3 for all direct discharging nonsmall facilities in
     Subcategories A through D, E through I, K and L, and BAT 2 for Subcategory J.  For small direct
     dischargers in subcategories K and L, EPA selected option BAT 1.  Table F-6 lists the incremental
     annualized cost and the incremental removals under the proposed options for each subcategory using the
     upper-bound costs. The incremental costs and removals are then totaled, and costs divided by removals to
     calculate the industry cost effectiveness ratio. For all direct dischargers, the industry CE ratio is about
     $21,900 per incremental pound equivalent removed based on upper-bound costs.
         4 Upgrade costs were estimated for options 3 and 4 only. Hence, incremental CE values could not be calculated
      for upgrade costs and average CE values are presented instead.
         5 EPA did not estimate retrofit costs for small facilities. The incremental CE of option 2 is undefined in some
      subcategories because incremental removals for the option are zero.
                                                    F-ll
    

    -------
                 Table F-4
     Results of Cost Effective Analysis
    Retrofit Costs for Nonsmall Facilities
    Regulatory
    Option
    Pretax Annualized
    Costs
    (Millions of $1999)
    Subcategory A through D
    BAT 2
    BATS
    BAT 4
    $9.93
    $42.25
    $73.53
    Pollutant Removals
    (Pounds Equivalent) ,
    
    93,586
    93,687
    94,195
    Pretax Average Cost
    Effectiveness ($1981
    per Pound
    Equivalent Removed)
    
    $62
    $263
    $455'
    
    PSES1
    PSES2
    PSES3
    PSES4
    $7.05
    $151.49
    $86.42
    $105.86
    240,421
    310,768
    309,081
    309,541
    Subcategory E through I
    BAT 2
    BATS
    BAT 4
    $0.40
    $0.54
    $3.53
    2,609
    2,618
    2,615
    $17
    $284
    $163
    $200
    
    $90
    $120
    $787
    '
    PSES1
    PSES2
    PSES3
    PSES4
    $18.79
    $102.09
    $83.25
    $109.82
    76,890
    78,831
    78,855'
    78,813
    $143
    $756
    $616
    $813
    Subcategory J
    BAT 2
    BATS
    BAT 4
    $0.55
    $4.28
    $4.98
    1,550
    1,621
    1,553
    $208
    $1,540
    $1,871
    
                    F-12
    

    -------
             Table F-4 (cont.)
     Results of Cost Effective Analysis
    Retrofit Costs for Nonsmall Facilities
    Regulatory
    Option
    PSES 1
    PSES 2
    PSES 3
    II PSES 4
    Subcategory K
    BAT 2
    BAT 3
    BAT .4
    BATS
    
    PSES 1
    PSES 2
    PSES 3
    PSES 4
    Subcategory L
    BAT 2
    BAT 3
    BAT 4
    BAT 5
    
    PSES 1
    PSES 2
    PSES 3
    PSES 4
    Pretax Annualized
    Costs
    (Millions of $1999)
    $1.33
    $23.25
    $23.09
    $24.78
    
    $4.82
    $34.46
    $44.21
    $66.09
    
    $10.84
    $188.95
    $126.00
    $131.39
    Pollutant Removals
    (Pounds Equivalent)
    3,918
    4,983
    5,112
    4,951
    
    63,192
    64,094
    64,029
    65,169
    
    377,651
    382,550
    382,735
    381,751
    Pretax Average Cost
    Effectiveness ($1981
    per Pound
    Equivalent Removed)
    $198
    $2,723
    $2,635
    $2,920
    
    $45
    $314
    . $403
    $592
    
    $17
    $288
    $192
    $201
    
    $0.30
    $2.18
    $3.03
    • $3.85
    373
    383
    371
    398
    
    $15.26
    $105.33
    $74.25
    $93 89
    49,950
    51,257
    51,367
    51,237
    $472
    $3,329
    $4,769
    $5,645
    
    . $178
    $1,199
    $843
    $1,069
                     F-13
    

    -------
                 Table F-5
       Results of Cost Effective Analysis
    Upper-Bound Costs for Small Facilities
    Regulatory
    Option
    Pretax Annualized
    Costs
    (Millions of
    $1999)
    Pollutant
    Removals
    (Pounds
    Equivalent)
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    "Removed)
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    Removed)
    Subcategorv A through D
    BAT1
    BAT 2
    BATS
    
    PSES 1
    PSE'S 2
    PSES 3
    PSES 4 '
    $0.03
    $0.51
    ' $4.30
    53.5
    53.5
    53.6
    $318
    $5,534
    $46,767
    $318
    Undefined
    $31,294,686
    
    $29.99
    $162.40
    $152.53
    $172.79
    2,819
    3,315
    3,299
    3,304
    $6,207
    $28,577
    $26,972
    $30,514
    $6,207
    $155,629
    $355,314
    $2,659,229
    Subcategory E through I
    BAT1
    BAT 2
    BATS
    $0.02
    $0.29
    $0.57
    2.9
    2.9
    2.9
    $3,843
    $57,940
    $113,831
    $3,843
    Undefined
    $3,429,962
    
    PSES1
    PSES 2
    PSES 3
    PSES 4
    $121.64
    $436.51
    $478.35
    $529.33
    1,489
    1,538
    1,538
    1,537
    $47,655
    $165,580
    $181,448
    $200,870
    $47,655
    $3,759,913
    $802,022,349
    ($46,264,959)
    Subcategorv J .,_
    BAT1
    BAT 2
    BATS
    $0.00
    $0.17
    $1.77
    596
    596
    624
    $0
    $169
    $1,659
    Undefined
    Undefined
    $33,007
    
                     F-14
    

    -------
              Table F-5 (cont.)
       Results of Cost Effective Analysis
    Upper-Bound Costs for Small Facilities
    Regulatory
    Option
    PSES 1
    PSES2
    PSES 3
    PSES 4
    Pretax Annualized
    Costs
    (Millions of
    $1999)
    $0.81
    $10.64
    $7.59
    $7.89
    Pollutant
    Removals
    (Pounds
    Equivalent)
    10,348
    10,654
    10,657
    10,644
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    Removed)
    $46
    $583
    $416
    $432
    Pretax
    Incremental Cost
    Effectiveness
    ($1981 Per Pound
    Equivalent
    • Removed) •
    $46
    $18,737
    ($571,582)
    ($13,042)
    Subcategory K •
    BAT1
    BAT 2
    BAT 3
    NA'
    NA
    NA
    NA
    NA
    NA
    NA
    NA
    NA.
    NA
    NA
    NA
    
    PSES 1
    PSES 2
    PSES 3
    PSES 4
    . $1.42
    $6.02
    $6.62
    $7.40
    21,071
    21,079
    21,080
    21,078
    Subcategory L
    BAT1
    BAT 2
    BAT 3
    $0.003
    $0.03
    $0.21
    1.4
    1.4
    1.4
    
    PSES 1
    PSES 2
    PSES 3
    PSES 4
    $27.29
    $101.36
    $94.67
    $10462
    1,034
    1,053
    1,054
    1.052
    $39
    $167
    $183
    $205
    
    $1,299
    $11,932
    $85,033
    
    $15,398
    $56,182
    $52,403
    $58.001
    $39
    $327,850
    $1,140,580
    ($317,057)
    
    $1,299
    Undefined
    $8,811,023
    
    $15,398
    $2,320,089
    ($2,957,132)
    C$3.750.193)
                     F-15
    

    -------
                               Table F-6
    Incremental Cost Effectiveness of Proposed Pollutant Control Options
                   Upper-Bound for Direct Dischargers
    Size
    
    Regulatory
    Option
    Incremental
    Pretax
    Annualized Cost
    (Millions of $1999)
    Pounds Equivalent
    Removed
    
    Cost Effectiveness
    ($1981/Pounds
    Equivalent)
    Subcategorv A through D
    Nonsmalls
    BATS
    $49.59
    101
    $286,414
    Subcategorv E through I
    Non-Small
    Subcategory J
    Nonsmalls
    Subcawgory K
    Nonsmalls
    Smalls
    Subcategory L
    Nonsmalls
    Smalls
    
    BATS
    $0.29
    9
    
    BAT 2
    $0.55
    1,550
    
    BAT 3
    BAT1
    $43.55
    NA
    902
    NA
    
    BATS
    BAT1
    
    $2.65
    $0.003
    $96.62
    10
    1
    2,573
    $18,512
    
    $208,
    
    $28,181
    •NA
    
    $160,314
    $1,299
    $21,897
                                   F-16
    

    -------
           Table F-7 calculates and compares the industry average cost effectiveness values for the proposed
    pollutant control options using upper-bound costs and retrofit costs for non-small facilities.  The average
    CE ratio for the industry is $401 per pound-equivalent using the upper-bound costs, and $287 per pound
    based on retrofit costs.
    
           Table F-8 summarizes the cost effectiveness of the proposed option for direct dischargers in the
    meat products industry relative to that of other industries.
    F.3     COST REASONABLENESS ANALYSIS
    
            F.3.1  Pollutants of Concern and Methodology
    
            EPA selected four noncprsyentional pollutants to perform the cost reasonableness analysis:
    chemical oxygen demand (COD), hexane extractable material (HEM), nitrate/nitrite, and total nitrogen.
    Table F-9 presents the nonconventional pollutant chosen for each option under the different subcategories.
    EPA calculates cost reasonableness as the average cost per pound removed of the selected pollutant under
    each regulatory option.  Cost reasonableness applies to direct discharging subcategories only. EPA has
    historically considered ratios as high as $37 per pound to be cost reasonable.
            F.3.2   Results
    
            Table F-10 presents the cost reasonableness results using both upper-bound and retrofit costs for
     nonsmall facilities in all subcategories. Based on upper-bound costs, BAT 4 in Subcategory L has the
     highest cost reasonableness value of almost $14 per pound of pollutant removed (in 1999 dollars). The use
     of retrofit costs lowers that value to about $10 per pound. The lowest cost per pound removed occurs
     under BAT 2 in Subcategory-J at about $0.03 per pound, which is the proposed option for this
     subcategory. Under the proposed option BAT 3 in all subcategories except J, cost reasonableness figures
     range from $6.60 to $9.60 per pound in subcategories K and L, to less than $1.60 in subcategories A
     through D and E through I.
                                                   F-17
    

    -------
    
    
    
    
    W "53
    (U M
    0 O
    it
    SI
    
    bฃ """^
    2 So
    >• TH
    •^ri ^"
    - . -
    
    
    
    
    **
    0
    .2
    4)
    
    a
    
    o
    it
    O)
    a,
    
    •4.1
    es 13
    tn ง S
    13 ^2 ^
    pj w ^
    I'll
    
    
    1
    Oป T— 1
    .N •ป
    C8 *S
    11
    m *"5
    **3 ^^
    u
    PH
    
    
    
    •**
    a
    1
    
    
    
    •a
    s
    o
    M
    ซ
    04
    P
    1?
    e
    ซ a
    ' If
    & O
    
    .1
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Q
    .ซ
    eo
    •*-*
    rr1
    Subcategory *
    CO
    \O
    cs
    •fee-
    
    
    
    
    t—
    CO
    fee-
    
    
    
    
    
    r-
    oo
    CO
    
    cs
    
    
    
    
    
    10
    ON
    fee-
    
    
    
    
    
    
    
    CO
    PQ
    
    Nonsmalls
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    >— H
    -s:
    r
    w
    Subcategory
    o
    ON
    
    
    
    
    
    
    . — i
    fee-
    
    
    
    
    
    .00
    (N
    
    O
    C5
    fee-
    
    
    
    
    
    VO
    c5
    
    
    
    
    
    
    
    
    CO
    PQ
    
    Non-Small
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    *-,
    Subcategory
    oo
    c^l
    •fee-
    
    
    
    
    oo
    fee-
    
    
    
    
    
    o
    >o
    1—1
    
    
    -------
                     Table F-8
    Industry Comparison of BAT Cost Effectiveness
               For Direct Dischargers .
    ':' ;.;/'*' ••". •• •' : .;'•-; :;•• ;
    -•-• •" "'•'-. • • '' ' .• -:' ,;:/ •• :
    Industry
    
    
    
    
    
    Coil Coating 	
    
    
    Electronics II
    Foundries
    
    Inorganic Chemicals II
    ron & Steel
    Leather Tanning
    Meat Products (Proposed)
    Metal Finishing 	 ,,,,.._
    
    
    Nonferrous Metals Mfg I 	
    Nonferrous Metals Mfg II 	
    Oil and Gas: Offshore1"
    Coastal— Produced Water/T
    
    Pesticides
    Pharmaceuticals0 A/C
    
    
    
    
    Textile Mills
    
    TPounds Equivalent ;._'•;. ;
    Currently Discharged
    1,340
    4,126
    12
    3,372
    BAT-BPT
    2289
    70
    9
    NA
    2,308
    32,503
    
    1,740
    259
    169
    3,305
    140
    34
    6,653
    1,004
    3,809
    1C 951
    54,225
    2,461
    897
    90
    44
    1,086
    BAT-BPT
    61,713
    BAT-BPT
    BAT=BPT
    Pounds Equivalent
    ; Remaining at Selected
    ' ; • „• "Option ;
    (thousands) • - B
    90
    : . • - - 5
    0.2
    1,261-1,267
    BAT=BPT
    9
    8
    3
    NA
    39
    1,290
    '27
    1,214
    112
    7
    3,268
    70
    , 2
    313
    12
    2,328
    239
    BAT = Current Practice
    9,735
    371
    47
    0.5
    41
    63
    BAT=BPT
    2,628
    BAT=BPT
    BAT=BPT
    ;,,; Incremental
    Cost Effectiveness of
    " Selected Option(s)
    ($ /founds Equivalent
    ; ~: Removed) -
    121
    2
    10
    5-7
    BAT=BPT
    49
    27'
    404
    . NA •
    84
    <1
    6
    66
    BAT=BPT
    $21,900
    12
    50
    ' 69
    4
    6
    33
    35
    BAT = Current Practice
    . 5
    14
    47
    96
    BAT=BPT
    6
    BAT=BPT
    39 |
    BAT=BPT I
    BAT=BPT 1
    ^Although toxic weighting factors for priority pollutants varied across these rules, this table reflects the cost-effectiveness at the tune of regulation.
    'Produced water only; for produced sand and drilling fluids and drill cuttings, BAT=NSPS.
    ND: Nondisclosed due to business confidentiality. ' . • •
                          F-19
    

    -------
                        Table F-9
    Pollutants Selected for Cost-Reasonableness Analysis
    legulatory Option
    •Pollutant.
    Subcategory A through D
    BAT 2
    BATS
    BAT 4
    HEM
    Nitrate/Nitrite
    Nitrate/Nitrite
    Subcategory E through I
    BAT 2
    BATS
    BAT 4
    HEM
    Total Nitrogen
    Total Nitrogen
    Subcategory J
    BAT 2
    BATS
    BAT 4
    COD
    COD
    COD
    Subcategory K
    BAT 2
    BATS
    BAT 4
    BAT5
    COD
    Total Nitrogen
    Total Nitrogen
    Total Nitrogen
    Subcategory L
    BAT 2
    BATS
    BAT 4
    BATS
    HEM
    Total Nitrogen
    Total Nitrogen
    Total Nitrogen
                            F-20
    

    -------
             Table F-10
    Cost Reasonableness Estimates
     Nonsmall Direct Dischargers
    
    =^==^==P=
    Regulatory
    Option
    =====;=
    Removals
    (Millions
    oflbs.)
    ——=—====== —
    Retrofit Costs
    Pretax Total
    Annualized
    Cost (Millions
    of $1999)'
    Average
    Cost/Pound
    Removal ($/lb.)
    Upper-Bound Costs
    Pretax Total
    Annualized
    Cost (Millions
    of $1999)
    Average
    Cost/Pound
    Removal ($/lb.)
    Subcategorv A through D 	 	 : 	
    
    
    
    BAT 2
    BAT 3
    BAT4
    12.30
    38.70
    41.00
    $9.9
    $42.2
    $73.5
    $0.81
    $1.09
    $1.79
    $9.9
    $59.5
    $118.0
    $0.81
    $1.54
    $2.88
    Suhcategorv E through I ._,_ 	 _
    BAT 2
    BATS
    BAT 4 •
    0.25
    2.01
    • 2.02
    $0=4-
    $0.5
    $3.5
    $159
    $0.27
    $1.74
    $0.4
    $0.7
    . $7.0
    $1,59
    $0.34
    $3.47 {I
    Subcategorv J 	 , 	 1|
    BAT 2
    BATS
    [BAT 4
    18.30
    18.30
    18.10
    $0.6
    $4.3
    $5.0
    .$0.03
    $0.23
    $0.27
    $0.6
    $5.8
    $6.3
    $0.03 ||
    $0.32
    $0.35
    Subcatepory K 	
    BAT 2
    BATS
    BAT 4
    BAT 5
    1.63
    '7.32
    8.10
    8.00
    $4.8
    $34.5
    $44.2
    $66.1
    $2.95
    $4.71
    $5.46
    $8.23
    $4.8
    $48.4
    $61.3
    $66.1
    $2.95
    $6.61
    $7.56
    $8.26
    SubcategorvL _ , 	
    BAT 2
    BATS
    BAT 4
    [RAT 5
    0.09
    031
    0.32
    0.32
    $0.3
    . $2.2
    $3.0
    $3.9
    $3.28
    $7.11
    $9.54
    $11.97
    $0.3
    $2.9
    $4.3
    $3.9
    _ __— ^— -^B— — =
    $3.28
    $9.60
    $13.59
    	
    $11.97
                  F-21
    

    -------
    F.4     COST EFFECTIVENESS METHODOLOGY AND RESULTS: NUTRIENTS
    
            In addition to conducting a standard CE analysis for selected toxic pollutants (Section F.2), EPA
    also evaluates the cost effectiveness of removing selected nonconventional pollutants: nutrients, primarily
    nitrogen and phosphorus. The methodology for this analysis has been drawn from the economic impact
    analysis of the Concentrated Animal Feeding Operations Industry (U.S. EPA, 2001).
    
            The nutrient cost effectiveness analysis does not follow the methodological approach of a standard
    CE analysis.  Instead, this analysis compares the estimated compliance cost per pound of pollutant removed
    to benchmarks, such as those reported in available cost effectiveness studies. A review of this literature is
    provided in Section F.4.1. EPA uses these estimates to evaluate the efficiency of regulatory options in
    removing nutrients and to compare the proposed BAT options to other regulatory alternatives (Section
    F.4.2).
            F.4.1   Review of Literature
    
            EPA has reviewed the available information on pollutant removal costs for nutrients. This
    research can be broadly grouped according to estimates derived for industrial point sources (PS) and
    various nonpoint sources (NFS), including agricultural operations.  In general, the PS research provides
    information on technology and retrofitting costs — and in some cases, cost per pound of pollutant removed
    — at municipal facilities, including publicly owned treatment works (POTWs) and wastewater treatment
    plants (WWTPs). This research utilizes actual cost data collected at a particular facility undergoing an
    upgrade. Other cost effectiveness research is based on the effectiveness of various nonpoint source
    controls, such as Best Management Practices (BMPs) and other pollutant control technologies that are
    commonly used to control runoff from agricultural lands. This research typically uses a modeling approach
    and simulates costs for a representative facility. The latter studies are less relevant to the proposed meat
    products industry effluent guidelines.
    
            EPA reviewed the literature on nutrient cost-effectiveness; Table F-l 1 summarizes the cost
    effectiveness values reported in these studies. These studies estimate a wide range of costs per pound of
                                                 'F-22
    

    -------
                                                  Table F-ll
                      Summary of Pollutant Removal Cost Estimates and Benchmarks
    Type of
    Pollutant
    Total
    Nitrogen
    (TN)
    Total
    Phosphorus
    (TP)
    Low
    Estimate
    High
    Estimate
    ($per pound removed)
    ($0.79)
    —
    $0.91
    $9.64
    $270.34
    $2.72
    $5.92
    $3.64
    $9.53
    $165.00
    $1,179.35
    $135.17
    ^•C^-j^^ryv'' -^
    :r-:- .,; ':•. -Treatment ^ -:V';:f
    ^vv^iv/.-;-,;!^.^'':^-^'.-'.:;-.1
    WWTPs
    WWTPs
    Aerobic Lagoon
    Ag.(low) to municipal
    Large Point Source
    Aerobic Lagoon
    Literature
    • Sources
    Randall et al (1999)
    Wiedeman (2000)
    Tippett and Dodd (1995)
    NEWWT 1994
    LCBP (1995)
    Tippett and Dodd (1995)
    WWTPs = Waste Water Treatment Plants; POTWs = Publicly owned treatment works.
    Full citations are provided in references. Timeframe of dollar values shown vary by source (shown below).
    Notes summarize timeframe of analysis, study assumptions (where available), and range of sources/treatment.
    Randall (2000):  1995-1998; 6% interest and 20-year capital renewal; BNR retrofits at^WWIP only.
    NEWWT (1994): 5% interest and 20-year capital renewal; low bound is agricultural BMPs and higher bound is
    municipal treatment facilities.
    McCarthy, et. al. (19961: No discount rate was applied and annual cost equals total lifetime costs adjusted by
    design life (varies by practice); study also examined agricultural land application (both with varying increasing
    over-application of land applied manure under pre-existing conditions). Cost-effectiveness values that assume
    direct discharge of animal wastes are not shown.                                          .             '
    LCBP (2000): 1995: No discount rate was applied and annual cost equals total lifetime costs adjusted by design
    life (varies by practice); study also examined agricultural BMPs.
                                                    F-23
    

    -------
    pollutant removed, spanning both point source and nonpoint sources, as well as a" range of municipal,
    urban, and agricultural practices. Annualized costs also vary widely depending on a variety of factors,
    including the type of treatment system or practice evaluated, and whether the costs are evaluated as a
    retrofit to an existing operation or as construction of a new facility.
    
            Researchers at Virginia Tech compiled a series of case studies that evaluated total costs for
    biological nutrient removal (BNR) retrofits at WWTPs throughout the Chesapeake Bay Watershed
    (Randall et al., 1999). These case studies estimated a range of costs per pound of nitrogen removed at
    these facilities. This research was commissioned by EPA's Chesapeake Bay Program and was conducted
    with the assistance of the Maryland Department of the Environment and the Public Utilities Division of
    Anne Arundel County. As part of this work, the researchers estimate BNR retrofit costs for 51 WWTPs
    located in Maryland, Pennsylvania, Virginia, and New York.  The final report in this series compares these
    costs to the projected change in effluent total nitrogen concentrations, assuming that the influent flow meets
    the design or projected flow after 20 years (Randall, et al., 1999).
    
            As shown in Table F-ll, this study  concludes that the costs of nitrogen removal are very plant-
    specific arid the costs per pound of addition  nitrogen removal ranged from a projected savings of $0.79 per
    pound to a cost of 5.92 per pound (Randall  et al., 1999).6 The range of these estimates is comparatively
    narrow given that the study examines  a single retrofit category across similar facilities.  This study assumes
    a 20-year capital renewal period and interest and inflation rates of 6 and 3 percent, respectively (Randall,
    2000).  The primary emphasis in this study is nitrogen, since the cost to upgrade for phosphorus removal is
    both configuration- and site-specific (Randall, 2000).7 Based on this analysis and other data from the
    Maryland Department of the Environment, EPA's Chesapeake Bay Program Office derived a cost
    effectiveness value for BNR of $3.64 per pound of nitrogen removed (Wiedeman, 1998).
       6 The costs per pound of additional nitrogen removed were flow-weighted to determine the average for each state
     and for all plants evaluated.
        1 For conventional piug-fiow activated sludge configurations, all that is required for phosphorous removal is the
     installation of relatively low-cost baffles and mixers; for oxidation ditches, the addition of an anaerobic reactor separate
     from the ditch is needed (Randall, 2000).
                                                   F-24
    

    -------
            A number of other studies have assessed the cost effectiveness of various state-level programs to
    reduce nutrients in Wisconsin (NEWWT, 1994) and Vermont (LCBP, 2000). In Wisconsin, a series of
    studies compared the cost effectiveness of point and nonpoint source controls across 41 sub watersheds in
    the Fox-Wolf watershed in Wisconsin (NEWWT, 1994). These studies estimated the cost of reducing
    phosphorus and suspended solids (TSS) loads from municipal treatment facilities and agricultural sources.
    Baseline projections were compared to necessary reductions to meet future water quality objectives (as
    mandated by that State's current regulations). Phosphorus removal costs for rural sources are estimated to
    be $9.64 per pound; while municipal treatment facilities have an estimated average annual cost of $165 per.
    pound of phosphorus removed (NEWWT, 1994).
    
            The Lake Champlain Basin Program (LCBP) conducted a similar study to evaluate costs to meet
    Vermont's water quality goals.  This study estimated phosphorus removal costs ranging from $270 to more
    than $1,000 per pound at a large municipal facility, compared to $440 to $544 per pound of phosphorus
    ;.~.r,;ir>ved using agricultural BMPs (LCBP, 2000). In addition, researchers at Virginia Tech ซvhc estimated
    removal costs for nitrogen at WWTPs conclude that it will cost about the same to remove a pound of
    phosphorus as it costs to remove a pound of nitrogen, if removing only one nutrient. If the facility is
    upgraded to remove both nitrogen and phosphorus, the cost typically will be only slightly more than the
    cost to remove nitrogen alone (Randall, 2000).
           F.4.2   Results of Nutrient Cost-Effective Analysis
    
           Tables F-12 and F-13 present the cost per pound of total nitrogen removals by subcategory and
    option.8 For direct dischargers, the average cost per pound of nitrogen removed ranges from $0.34 under
    BAT 3 in Subcategory E through I, to more than $15 (upper-bound costs) under BAT 3 in Subcategory J
    (Table F-12). For indirect dischargers, the average cost per pound of nitrogen removed ranges from $0.16
    under PSES 1 in Subcategory J, to about $40 (upper-bound costs) under PSES 4 in Subcategory E through I
      8 No nitrogen is removed under option 2. The technology for option 2 includes nitrification but not denitrification.
    Therefore nitrogen is not removed from the wastewater but is instead converted to nitrate/nitrite (see Development
    Document, U.S. EPA, 2002)."
                                                  F-25
    

    -------
    
    
    
    
    
    
    
    
    
    
    a
    S-
    p
    u
    •*"*
    งS2
    3 &
    O I-
    ซr.3
    is
    O Q
    a tj
    C<~cj
    .^T* W
    O ฃ
    "2S
    *S *••
    
    o f2
    ซ l
    •*_> a
    a o
    g & &
    " o c2
    s**
    •*3 13 CO
    ,C3 5 O
    H |0
    ง*!
    fiS
    CM ซ
    oซ
    8*8
    a a
    S-e
    •••* s
    "8 9
    QJ O
    SSM
    W ฃ
    ,!_) CU
    WQ.
    - ป"*4
    5^
    CU
    Cฃ
    2
    o
    >•
    <
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    •
    CO
    CU
    
    tu
    ^3 "co"
    ซ2 g
    W ฎ
    •ซJ C^
    at
    jy ^H
    1^
    •<
    
    
    
    
    
    o
    _!-
    
    ^S
    tf
    
    
    
    -e
    o
    i
    V '
    c,
    o.
    p
    
    
    
    
    CO
    " "O _
    a ^3
    o ^
    2: s
    IS
    H
    
    
    
    
    
    
    
    •4-*
    <ง^
    ^"^ o\
    'O O\
    a^
    es Q
    g ง
    
    
    5
    
    
    
    
    
    
    
    
    
    ^_,
    a
    o
    •^-*
    ^
    
    
    
    
    
    •s
    s
    o
    
    CU
    a
    P
    
    
    ^
    1*
    _2
    "3 .ฐ
    
    W) Q^
    ซo
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Q
    60
    . a
    p
    •s
    b
    Q
    tf
    ซ
    5J
    -c>
    ซs
    g
    
    
    
    
    
    
    
    
    
    T3
    U
    W
    o
    3
    
    
    
    
    O
    
    
    
    
    
    
    
    
    
    
    S
    
    
    
    
    
    
    
    
    
    CO
    c\
    ฃ
    
    
    
    
    
    
    
    
    
    
    
    04
    J^.
    J\J
    PQ
    t— 1
    T—5
    •fee-
    
    
    
    
    
    
    
    
    VO
    *4
    
    
    
    
    
    0
    CO
    of
    CX
    "•H
    oo"
    CO
    
    
    
    
    
    •n
    01
    o5
    f/^
    
    
    
    
    
    
    
    
    
    -------
                                      g
                        r-
                                                                            Os
                          t
        1
        O
    
     
    -------
      C
      o
      ex
    
      IS
      o ts
      g 2
    
      II
      o A
      Jl
    
    ail
    
    ^ฃ<2
    ซ 5 <"
    S ง 42
        8
      •-S  a
      ,9i 13
      la  B
      ta  ?
      *^  o
      •S ca
    
      a  g
       2
    1
    
    
    .
    w
    B
    V
    •^ TZ
    1 u a
    SB ง
    W g
    ||
    f^
    •rff*
    
    
    •
    1'
    
    
    s
    o
    42
    *2j
    PM
    
    
    
    ,9
    O
    I
    fc.
    o<
    p
    • ,
    en
    If
    o s
    s ,5
    ^ f*4
    H
    
    
    
    CO
    
    If tl} O\
    ป 6M
    i i
    111
    &
    
    
    
    
    
    
    
    O-
    Ui
    s
    ซ
    
    
    
    •e
    G
    O
    ฃ
    a<
    P^
    
    
    o
    13 a
    l>t
    | a> n<
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Q
    
    2
    o
    x.
    c
    ^
    c
    c?
    <
    
    
    
    
    
    
    
    
    
    Undefined
    
    
    
    
    o
    
    
    
    
    
    g
    
    
    
    
    
    
    V
    ฃ
    
    
    
    
    
    
    
    ,— I
    CO
    PL
    c/:
    CM
    ฃ
    
    
    
    
    
    
    
    
    
    Undefined 1
    
    
    
    
    o
    
    
    
    
    
    S
    
    
    
    
    
    
    ง
    V,
    •6=
    
    
    
    
    
    
    
    CM
    CO
    ff.
    c/:
    PH
    CM
    fee
    
    
    
    
    
    
    
    
    q
    en
    
    
    
    
    s
    oo
    1 — 1
    *-H
    en
    
    
    
    oo
    
    
    
    
    
    
    CN
    vc
    
    
    
    
    
    
    
    en
    CO
    p:
    CO
    PH
    CM
    en
    
    
    
    
    
    
    
    
    
    en
    
    
    ^u
    
    en
    VO
    G\
    OO
    CM"
    en
    
    
    
    00
    8
    
    
    
    
    
    
    S
    c:
    
    
    
    
    
    
    
    T^
    or
    PH
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    -si
    a
    ^
    DrT
    il
    jj
    j
    c?
    g
    
    
    
    
    
    
    
    
    
    en
    
    
    
    
    in
    o\
    i-H
    aC
    o
    r— i
    
    
    
    
    S
    
    
    
    
    
    
    gj
    oc
    
    
    
    
    
    
    
    ^H
    cr
    g
    
    
    
    
    
    
    
    
    
    Undefined
    
    
    
    
    o
    
    
    
    
    
    2
    
    
    
    
    
    
    o\
    c-i
    
    
    
    
    
    
    
    CM
    CO
    PL
    CO
    PH
    O
    en
    
    
    
    
    
    
    
    
    oo
    d>
    en
    
    
    
    
    en
    oq^
    en
    CM"
    
    
    
    
    cQ
    en
    oo
    
    
    
    
    
    
    00
    \c
    er
    00
    -se-
    
    
    
    
    
    
    
    en
    CO
    tt
    co
    PH
    a(
    en
    
    
    
    
    
    
    
    
    en
    6O-
    
    
    
    
    00
    i — i
    *—,
    oo
    CM"
    
    
    
    
    CM
    oq
    O
    T-H
    
    
    
    
    
    
    c^
    c
    &
    
    
    
    
    
    
    
    •*l
    c<
    PH
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    ฃ
    1
    c?
    g
    
    
    
    
    
    
    
    
    
    i
    
    
    
    
    00
    oo
    o
    oo"
    
    
    
    
    g
    
    
    
    
    
    
    en
    er
    ^
    
    
    
    
    
    
    
    <— '
    (/
    S
    g
    
    
    
    
    
    
    
    
    
    Undefined
    
    
    
    
    o
    
    
    
    
    
    g
    
    
    
    
    
    
    CM
    en
    
    
    
    
    
    
    
    CM
    1
    PH
    O
    en
    CM
    ^
    
    
    
    
    
    
    
    
    OO
    IT-
    CM
    
    
    
    
    en
    en
    s
    o"
    F— H
    
    
    
    q
    
    
    
    
    
    
    S
    si
    •se-
    
    
    
    
    
    
    
    en
    c<
    S
    en
    CM
    fee
    
    
    
    
    
    
    
    
    vo
    oo
    CM
    
    
    
    
    1
    o"
    1-1
    
    
    
    OO
    I
    
    
    
    
    
    
    CM'
    CM
    s
    
    
    
    
    
    
    
    ^3"
    CO
    pq
    CO
    oo
    CN
    

    -------
       S
       1  ซ
       O I— l
       O a
       **  05
    't' w  S
    
    
          I
    ta
    a     s
    •S  2 u
     ซ ฃ *-
    H S ซ3
    
        en ^5.
        = ~
    
       ฃ•8
       •C  eg
    
       I  a
       W  ง
       •s pa
        O)
        Si)
        ซ
    
    
    
    
    
    S3 •'
    
    -------
     (Table F-13).  The cost per pound .of nitrogen removed is generally much lower for direct dischargers than
     indirect dischargers.
    
            Under the proposed options (BAT 3 for all subcategories except J, for which BAT 2 was selected),
     cost per pound of nitrogen removed ranges from $6.60 to almost $10 (upper-bound costs) in subcategories
     K and L ($5 to $7 for retrofit costs).  In subcategories A through D, and E through I, the cost is less than
     $1.56 per pound. No nitrogen is removed under the proposed option for Subcategory J.
    
            Tables F-14 and F-15 present the cost per pound of total phosphorus removals by subcategory and
     option. For direct dischargers, the average cost of phosphorus removals ranges from $5 per pound under
     BAT 1 in Subcategory A through D, to $311 per pound (upper-bound costs) under BAT 5 in Subcategory
     L (Table F-14).  For indirect dischargers, the average cost per pound of phosphorus removed ranges from
     about $7 under PSES 1 hi Subcategory K, to $180 (upper-bound costs) under PSES 4 in Subcategory J
     (Table F-15).  For all options except 3 and 4 in subcategories K and L, the cost per pound of phosphorus
     removed is lower for direct dischargers 2ian indirect dischargers.
    
            Under the proposed options (BAT 3 for all subcategories except J, for which BAT 2 was selected),
     the cost of phosphorus removals is the highest for Subcategory L ($225 per pound, upper-bound costs;
     $167 per pound retrofit costs) and Subcategory K ($46 per pound, upper-bound costs; $33 per pound
     retrofit costs).  In subcategories A through D, E through I, and J, the costs are less than $13 per pound
     (upper-bound costs) and $9 per pound (retrofit costs).
    
           Tables F-16 and F-17 present the cost per pound of total nutrient removals by subcategory and
     option. In all subcategories, the cost per pound of nutrients removed is lower for direct dischargers than
     for indirect dischargers, often substantially lower. Among direct dischargers the cost of total nutrient
    removals is less than $1.50 per pound in Subcategory A through D and E through I, and less than $7.00
    per pound for Subcategory J under the proposed options. The highest cost per pound under the proposed
    options is found in Subcategory L, and that does not exceed $10; for Subcategory K, the cost is less than
    $6 per pound of nutrients removed.
                                                 F-30
    

    -------
    
    
    
    
    
    
    
    3
    o
    IB
    1 &
    gl
    a u
    O ซJ
    *"* V?
    O "8 '
    
    S ***
    •*•* c^
    5,1
    jj
    fe <ฃซ2
    •2 "O J2
    ,O ซu In
    ta tn e
    H ง,U
    2ฃ
    BH 0
    'S -S
    O o>
    " *<•
    1 "a
    g> ซ
    •j3 -g
    w a
    fg ง
    W cq
    ^* W
    s a
    0 a
    ซ D
    SJD
    2
    aป
    -^
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    -.s
    ,v S
    m 5
    si
    ||
    
    ซ's~'
    1
    
    
    
    
    
    
    1
    1
    
    
    
    F
    -"8
    C
    o
    ?
    V
    n.
    P
    
    
    
    
    
    
    ll
    |8
    H
    
    
    
    
    
    
    
    
    
    a""
    ss
    frt ^^
    5,2.
    •PM "5^*
    lo
    II
    ง.
    
    ฃ
    
    
    
    
    
    
    
    
    
    
    
    -w
    1
    V
    P"4
    
    
    
    ts
    0
    o
    M
    •
    ซ
    a,
    &4
    P
    
    
    1
    "ง O
    Sf'U
    ซ0
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    Q
    ^
    60
    s
    g
    •ซ
    
    ^%
    ^
    6Q
    a
    1
    co
    
    *
    
    
    
    
    
    VI
    
    
    
    
    
    
    
    cs
    1 — 1
    o
    T-H
    
    
    
    
    
    $
    
    
    
    
    
    
    
    
    CO
    ON
    
    
    
    
    
    
    
    
    
    
    
    CM
    ^
    pa
    ON
    
    
    
    
    
    
    CO
    
    
    
    
    
    
    
    _
    ^5
    ^^
    VO"
    cs
    VO^
    
    
    
    
    
    ^,
    ^g*
    60
    S
    g
    .Jg
    to
    i5
    'Q
    ^
    CO
    
    Z
    
    
    
    
    
    i — 1
    ' T— H
    
    
    
    
    
    
    
    o
    o
    t-;
    CO
    
    
    
    
    
    CO
    ON
    ON
    
    
    
    
    
    
    
    
    0
    C5
    
    
    
    
    
    
    
    
    
    
    fN
    ^
    m
    
    &
    
    
    
    
    
    t~~
    &*
    
    
    0
    
    
    
    
    CO
    
    t"-*
    S
    
    
    
    
    
    a'
    
    
    
    
    
    
    
    
    ON
    VO
    O
    e<3-
    
    
    
    
    
    
    
    
    
    
    CO
    ^
    pa
    vo
    CO
    
    
    
    
    
    S
    
    
    
    
    
    
    
    vo
    
    ^^
    ON
    
    
    
    
    
    CO
    CO
    ^3.
    
    
    
    
    
    
    
    o
    fe
    
    
    
    
    
    
    
    
    
    
    •*
    ^
    pa
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    ^*i
    ^
    60
    ซ
    1
    =0
    
    *
    
    
    
    
    
    ป
    
    
    
    
    
    
    
    cs
    r-
    r-
    vd"
    oo
    
    
    
    
    
    ro
    ON
    ON
    $&
    
    
    
    
    
    
    
    vl
    S
    
    
    
    
    
    
    
    
    
    
    CJ
    ฃj3
    pa
    T-M
    S
    
    
    
    
    
    vo
    
    
    
    
    
    
    
    00
    00
    CO
    co"
    o
    1— 1
    
    
    
    
    
    00
    ฃ/^
    
    
    
    
    
    
    
    o
    oq
    iri
    
    
    
    
    
    
    
    
    
    
    CO
    ^
    pa
    ^
    ^
    
    
    
    
    
    vo
    
    
    
    
    
    
    
    in
    C1^
    •^"
    ON
    
    
    
    
    
    3
    ^^~
    
    
    
    
    
    
    
    ซ — i
    CO
    VO
    •ae-
    
    
    
    
    
    
    
    
    
    
    •ซ*
    ^
    pa
    CO
    

    -------
    1
    a-
    
    
    
    
    
    ft*
    a
    ซ
    y 'TJj
    1* H
    5*^* S
    *€
    O ง}
    O c\
    a> 31
    1~
    "*^
    
    
    
    
    
    •4^
    'S
    !*
    ซ
    rt
    
    
    
    
    
    
    
    "O
    a
    0
    PQ
    0)
    a
    a
    1*
    5'i
    ฃ.1
    IS
    H
    
    
    EO
    1g|
    H So
    ง•8
    |1
    it
    t.
    CL,
    
    
    
    
    
    *S
    a)
    
    
    
    
    "O
    ง
    O
    P5
    •
    a
    A
    P
    (4
    -S
    ซ g
    If
    ^0*
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    1
    -|
    eg
    <
    ^
    
    
    
    
    
    
    
    
    
    
    
    •se-
    
    
    
    
    
    
    CO
    CO
    oo
    cC
    ง
    
    
    
    CO
    
    
    
    
    
    CO
    oo
    a
    
    
    
    
    
    
    
    ซ
    8
    •ae-
    
    
    
    
    
    
    
    
    
    
    
    s
    
    
    
    
    
    
    ง
    T— 1
    
    
    CO
    
    
    
    
    
    CO
    s
    
    
    
    
    
    
    
    CO
    00
    fee-
    
    
    
    
    
    
    
    
    
    
    
    CD
    oo
    fee
    
    
    
    
    
    oo
    "1
    oo
    ปvo
    f-^
    
    
    
    1 — 1
    co
    
    
    
    
    
    c^
    1 — 1
    VO
    fee
    
    
    
    
    
    
    
    Pง
    <
    ^
    
    
    
    
    
    
    
    
    
    
    
    g
    to-
    
    
    
    
    
    CN
    ct
    oo
    
    
    
    CO
    ฐ>
    
    
    
    
    
    1
    fee-
    
    
    
    
    
    
    
    ซ
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    '
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    L ^
    
    S
    1
    ^
    <
    2
    
    
    
    
    
    
    
    
    
    
    
    1
    
    *!
    
    -------
    
    
    
    
    
    
    
    
    
    
    U3
    2
    o
    •ง,ซ
    tfl o\
    1 tf.
    0* es
    1 JS
    = 1
    •25
    a **
    Ov
    o
    s|
    •fci S
    S i-<
    3 1
    3l
    ซ1 s
    25IZ;
    s*ซ •
    S'Ss
    eB ' to tซ
    H 2,9
    n,cj
    *i
    'S 1
    m a>
    cn ป^
    •S ป*
    n
    > CS
    1-s .
    ^ ง
    H ,5
    -*^ ^i
    an t<
    O S
    U2T
    P<
    &^
    2
    ^
    <
    
    
    
    
    
    
    
    
    
    
    
    
    m
    a
     S3
    S J
    "i ป
    r i ^*
    25
    ,SปS;
    -1
    
    
    
    
    
    
    
    
    
    1
    "S
    K
    
    
    
    
    
    
    •e
    o
    
    1
    Qi4
    (2
    
    
    
    
    n
    1 i
    — i
    
    H
    
    
    
    
    
    
    i ^
    "O งN
    N &>
    yz Q^^
    CS o
    s ป
    i ง
    M a
    (S
    
    
    
    
    
    
    
    S
    |
    i
    
    
    
    
    
    
    T3
    1
    o3
    eu
    B,
    ^
    
    ^
    *•<
    ^
    'C5 fl
    "a .2
    P