United States
         Environmental Protection
         Agency
            Office of Air Quality
            Planning and Standards
            Research Triangle Park, NC 27711
DRAFT REPORT
EPA-452/D-96-005
June 1996
vvEPA
         Air
                                175
ECONOMIC IMPACT ANALYSIS
AND REGULATORY FLEXIBILITY
ANALYSIS OF AIR POLLUTION
REGULATIONS: ARCHITECTURAL
AND INDUSTRIAL MAINTENANCE
COATINGS

-------
                           DISCLOSURE

      This report presents the results  of  an  economic  impact
 analysis and regulatory flexibility analysis of  a  proposed
 federal rule for volatile organic  compounds  (VOC)  for the
 architectural and industrial maintenance  (AIM) coatings
 industry.   All monetary impacts estimated in this  report  are
 reported in real (base year 1991)  dollars.    The economic and
 regulatory impacts  reported in the preamble  for  this  rule are
 reported in real 1996  dollars.   The base-1991 dollar  estimates
 were  converted to base-1996 dollar estimates by  multiplying
 the base-1991 estimates by the ratio of 1990/1995  gross
 domestic product (GDP)  price deflator, thus  converting from
 end-of-1990/beginning-of-1991 dollars  to  end-of-
 1995/beginning-of-1996  dollars.  The 1995 deflator was
 estimated by computing  the average  annual change in the GDP
 deflator from 1990-1993  and using  this annual change  estimate
 to project a 1995 deflator using the observed 1993 deflator.
      This  report is available in the public docket upon
 proposal of the  rule.   Copies of the report and other material
 supporting the proposal  are in  Docket A-??-?? at EPA's Air  and
 Radiation  Docket and Information Center,  Waterside Mall,  Room
M1500,  Central Mall, 401 M.  Street, Washington,  D.C.  20460.
The EPA may charge a reasonable fee  for copying.

-------
                          TABLE  OF  CONTENTS

Section                                                         Page

             List  of  Figures   	    vii
             List  of  Tables	   ix

   1          Introduction, Regulatory  Background,  and
             Industry Profile  	  1-1
             1.1  Introduction	  1-1
             1.2  Regulatory  Background  	  1-2
                 1.2.1  Regulatory Structure  	  1-3
             1.3  Industry Profile 	  1-5
                 1.3.1  Commodities and VOC Content   	1-5
                 1.3.2  Demand for AIM Coatings   	1-7
                        1.3.2.1  Conceptual View  of  Coating
                                 Decision   	1-7
                        1.3.2.2  Aggregate Demand and
                                 Substitution  Effects   ....  1-9
                        1.3.2.3  Substitution  Possibilities  . 1-11
                        1.3.2.4  Characteristics  of  Latex
                                 and  Alkyd Paints  	1-13
                        1.3.2.5  Substitutability Between
                                 Latex and Alkyd  Paints   .  .  . 1-16
                        1.3.2.6  Coating Users  	 1-20
                 1.3.3  Production of  AIM Coatings	1-22
                        1.3.3.1  Raw  Material  Inputs  	 1-22
                        1.3.3.2  Formulations   	 1-25
                        1.3.3.3  Manufacturers' Substitution
                                 Options and New
                                 Technologies   	 1-27
                 1.3.4  Industry Conditions    	 1-30
                        1.3.4.1  Shipments and Manufacturer
                                 Specialization   	 1-30
                        1.3.4.2  Company Size and Industry
                                 Structure	1-34

  2          Cost  and  Economic Impacts of Proposed Regulation  .  2-1
             2.1 Background	2-1
             2.2 Overview  of Response  Options  	  2-2
                2.2.1  Supply	2-2
                2.2.2  Demand	2-4
             2.3 Cost  Analysis	2-5
                2.3.1  Reformulation  Costs    	  2-5
                                ill

-------
                    TABLE  OF  CONTENTS  (continued)

Section                                                          page

                        2.3.1.1   Product-Level  Reformulation
                                  Cost  Estimates  	2-5
                        2.3.1.2   Total Reformulation Costs
                                  and VOC  Reductions  	2-10
                 2.3.2  Exceedance Fee Option   	2-16
                 2.3.3  Reformulation versus Exceedance
                        Fee:   Least-Cost Analysis   	  2-16
             2.4  Market Analysis	2-23
                 2.4.1  Potential Firm Responses and
                        Market Effects with Product
                        Reformulation   	  2-23
                        2.4.1.1   Firm-Level Model Extension   .  2-24
                 2.4.2  Model Execution and Results   	  2-26
                        2.4.2.1   Baseline  	  2-26
                        2.4.2.2   Quantifying Market Shocks .   .  2-29
                 2.4.3  Market and Welfare Effects Results  .   .  2-32
             2.5  Social Cost-Effectiveness  	  2-40
             2.6  Employment Impacts  	  2-44
             2.7  Summary	2-45

  3          Regulatory Flexibility Analysis   	  3-1
             3.1  Background,  Affected Entities,  and
                 Regulatory Requirements  	  3-1
                 3.1.1  Potentially Affected Entities   ....  3-2
                 3.1.2  Regulatory Requirements    	  3-4
             3.2  Analysis	  3-4
                 3.2.1  Baseline Market Presence of
                        Small AIM Coatings Producers	  3-4
                 3.2.2  Costs Associated With Regulatory
                        Compliance	3-15
                 3.2.3  Reformulation Cost Impact Estimates    .  3-15
                        3.2.3.1   Small  Business Costs Under
                                  "Reformulation-Only"
                                  Option	3-19
                 3.2.4  Potential Mitigating Effects,  Fee
                        Option and Small Product Exemption  .     3-22
                 3.2.5  Market Effects	3-24
             3.3  Summary	3-27

  4          References	  4-1
                                 IV

-------
Appendices

   A          Market Definition,  Demand Estimation,
              and Data	A-l

   B          Methodology for Computing Market and Welfare
              Adjustments	B-l

   C          Volatile Organic Compound Content Levels and
              Emission Reductions for Select Architectural
              Coatings from the 1990 Survey Data	C-l

   D          Sensitivity Analysis of National Cost  Estimate  	 D-l

-------
                           LIST OF FIGURES

Number                                                          Page

  1-1   Comprehensive classification of emissions
        from consumer and commercial products	1-2
  1-2   Inputs generally used in the manufacture
        of a solventborne coating	1-26
  1-3   Inputs generally used in the manufacture
        of a waterborne coating	1-26
  1-4   Approximate volume relationships  of  coating
        ingredients	1-27
  1-5   Location of manufacturing establishments
        in the paints  and allied products
        industry in 1987:   SIC 2851	1-36
  2-1   Basic  stages  of AIM  coating  reformulation
        (prototype  firm and  product)   	   2-4
  2-2   Fee versus  reformulation   	2-17
  3-1  Chain  of  ownership	   3-3
                                VI1

-------
                           LIST OF TABLES

Number                                                           Page
  1-1      Average VOC  Content  for AIM Coatings
           Expected to  Be Covered by Proposed Regulation   .  .  .   1-6
  1-2      Examples of  Users' Substitution Possibilities   .  .  .  1-12
  1-3      Paint Characteristics    	  1-15
  1-4      Comparison of  Characteristics  Between Latex
           and Alkyd Paint  	1-17
  1-5      Consumers of Architectural Coatings   	  1-21
  1-6      Commodity Use  in  1982:  Paint and Allied
           Products	1-22
  1-7      Percentage of  Solvent  in Conventional and
           Reformulated Organic Solventborne Coatings  	  1-30
  1-8      Value Shipped  of  Potentially Regulated Paint
           and Allied Products: 1981 Through 1991 ($106)     .  .  .  1-31
  1-9      Number of Companies, Establishments,  and
           Producer Specialization—Paint and Allied
           Products:  1987  	1-33
  1-10     Commodity Production in 1982:  Paint and
           Allied Products  	  1-34
  1-11     Large Firm Dominance and Numbers of Companies
           and Establishments in  the Paint and Allied
           Products Industry:  1987   	  1-35
  1-12     Number of Companies and Establishments in
           the Coatings Industry,  Selected Years,
           1972-1991   	1-38
  1-13     Recent Acquisitions in  the Coatings Industry  ....  1-38

  2-1      Reformulation  Cost Estimation  Uncertainties   ....  2-8
  2-2      Table of Standards	2-12
  2-3      Reformulation  Costs and VOC Reductions
           (Survey Population)   	  2-15
  2-4      Fee versus Reformulation:   Least-Cost Analysis
           for the Survey Population  	2-21
  2-5      AIM Coatings Market Segments Baseline Data  	  2-27
  2-6      Market Model Results -  Scenario 1:  Standard  ....  2-33
  2-7      Market Model Results -  Scenario 2:  Reformulation
           Plus Fee	2-34
  2-8      Market Model Results -  Scenario 3:  Low Fee Rate .  .  2-35
  2-9      Market Model Results -  Scenario 4:  High-Cost
           Reformulation   	2-36'
                                 IX

-------
                     LIST OF TABLES  (continued)

Number                                                           Page

  2-10     Social Cost-Effectiveness  Estimates   	  2-43
  2-11     Estimated Employment  Effects  	  2-44
  3-1      Small Business Presence  in  the AIM Coatings
           Market	   3-5
  3-2      Baseline VOC Content	3-10
  3-3      Specialization-Based  Average VOC Content:
           Small Business Products   	  3-12
  3-4      Small Business Reformulation Costs, 1997  	  3-16
  3-5      Average Regulatory Impact by Firm Size    	3-21
  3-6      Market-Level Impacts  and Small Business
           Presence	3-26
                                 x

-------
                            SECTION 1
    INTRODUCTION,  REGULATORY BACKGROUND,  AND INDUSTRY  PROFILE
 1.1  INTRODUCTION

      Under Title I of the  Clean Air Act of 1990,  the U.S.
 Environmental Protection Agency (EPA)  is developing
 regulations to reduce volatile  organic compound (VOC)
 emissions from various commercial  and  consumer  products.   The
 first category of commercial  and consumer products  to be
 regulated is architectural and  industrial maintenance (AIM)
 coatings.
      This report analyzes  the economic impacts  of these
 proposed  federal regulatory strategies.   Section 183  (e) (1) (B)
 of  the Clean Air Act  Amendments  of  1990  defines a consumer or
 commercial product  as
      any  substance, product (including paints,  consumer
      and  commercial products,  and solvents), or article
      (including any container or packaging) held by any
      person,  the use,  consumption,   storage, disposal,
      destruction, or  decomposition  of  which may result in
      the  release of volatile organic compounds.
 Thus,  the  general purpose of the regulation is to reduce the
 flow  of VOCs  into the  atmosphere from  consumption and disposal
 of products  that contain VOCs.  Figure 1-1 shows the
 dissipative  emissions  and the disposal emissions into the  air
 that  are  the  target of this regulation.1  These emissions  are
distinguished  from the manufacturing-related emissions that
are controlled by other forms  of regulation (as are emissions
                              1-1

-------
                   MM  jmi
                    Milirxli UK I coniumption (pelhilinti
                   •mbodiad in coflimircul ind contumn goodi)
                                             toodtlufli
                                           application to sorit o( agro-ch
        Pom!- and atta-aoutca
        production amutiona
Distipativa
•miationi
                     //  Y\
^

11
Dupoul •mitiioM

ft n
1


}} (I-
i

  Ar*a-»ouic»
agtKullural «nitsioni
                           I  Stwaga  I I  Incmifation  I I   Landlil   I
                            AIR LAND. AND WATER
  Figure  1-1.   Comprehensive classification of emissions  from
                consumer  and commercial  products.
  Source: Adapted from Stigliani, William M.  Chemical Emissions  from
         the Processing and Use of Materials:  the Need for an
         Integrated Emissions Accounting System.   Ecological Economics
         2(4) :325-341.  1990.  (Figure 2).

to land and water) .  The regulatory structure is  presented
here followed by an overview of the AIM coatings  industry.

1.2  REGULATORY BACKGROUND

     Section 183(e)(3)(A)  directs  the  EPA to create categories
of consumer or commercial products  whose use accounts for at
least  80  percent of VOC  emissions  in ozone nonattainment
areas.  The EPA will divide this category list into four
groups  and establish priorities for regulation.   AIM coatings
is the  first group of products to be regulated.
     The  design of regulatory strategies to reduce VOCs
emitted by AIM coatings  is shaped  in specific ways by the
                                1-2

-------
 Clean Air Act as amended.   Two components of the legislation
 are of particular importance:
      •  determining regulated entities and
      •  establishing best  available controls.
      Regulations developed under Section 183(e) may be imposed
 only with respect to "manufacturers, processors, wholesale
 distributors, or importers of consumer or commercial products
 for sale or distribution in interstate commerce in the United
 States" or certain entities that supply such products to the
 former Sections 183(e)(l)(C)  and 183(e)(3)(B).   The definition
 of regulated entities excludes retailers and users.
      The regulations affecting AIM coatings will require best
 available controls.   The EPA Administrator,  on the basis of
 "technological and economic feasibility,  health, environ-
 mental,  and energy impacts,"  will determine the desired degree
 of emissions reduction that
      is achievable through the application  of the most
      effective equipment,  measures,  processes,  methods,
      systems or techniques,  including  chemical
      reformulation,  product or feedstock substitution,
      repackaging,  and directions  for use, consumption,
      storage,  or disposal.   (Section 183[e][l]).
 The requirement for  best available  controls  establishes  the
 general  environmental  goal  of  regulation, not the means  by
 which regulated entities will  comply.
 1.2.1 Regulatory Structure
      The EPA surveyed  116 AIM  coatings  manufacturers for
 products manufactured  and their VOC  contents.2  The
 Architectural  and Industrial Maintenance Surface Coatings VOC
 Emissions Inventory  Survey  (the survey) provides average VCC
 emissions and  1990 sales quantities  by  product.  Based on
 these data,  EPA  is proposing VOC content limit standards,
which will go  into effect in 1997, to regulate VOC emissions
 from  65  AIM  coatings products.  Beginning July 1, 1997,
manufacturers  and importers of AIM coatings subject to the
regulation must  limit the VOC content per liter of coating to
                              1-3

-------
 the  standards specified for each coating product  they
 manufacture.   The EPA is also considering an option of
 allowing manufacturers and importers that do not  meet the
 limits  of the standard to pay an exceedance fee.  The VOC
 content limits are presented in the Table of Standards  (TOS)
 for  1997 in Section 2 (Table 2-2).   The limits specified in
 this table were used in this economic impact analysis.  They
 cover all the major AIM coatings categories as well as certain
 special purpose coating products for which a less stringent
 limit is granted in order for the coating to adequately
 perform its designed purpose (e.g.,  high-temperature
 coatings).
      AIM coatings manufacturers who do not limit the VOC
 content of  their products to meet  the 1997 standards will pay
 a fee on the  VOC content of the product in excess of the
 limit.a  The proposed fee  rate  is $5,000  (1995  dollars,
 adjusted to $4,500 in 1991 dollars)  per metric ton (Mg)  of
 excess  VOCs.   Fees will  be paid quarterly and used by the
 Administrator to support the administration of the regulation
 or to promote additional VOC emission reductions from AIM
 coatings through technological  development grants, award
 programs, or  other means.
      This report includes  an overview of  the AIM coatings
 industry, products,  and  technologies and an analysis  of  the
 economic impacts on the  affected entities and the industry as
 a result of the  proposed TOS VOC content  limits and fees.   An
 economic model of  the AIM coatings  industry is  developed to
 obtain  estimates of the  potential price and quantity  changes
 associated  with  the proposed TOS and exceedance fees.  In
 addition, a Regulatory Flexibility Analysis  is  conducted,
which estimates  the impacts  of  the regulation on small
     *Excess VOCs are defined as  the maximum VOC content of the coating,
as applied, in grams per liter of  coating, less water and exempt compounds,
minus the applicable VOC standard.

                              1-4

-------
 businesses and presents alternatives  that may be implemented
 to mitigate those impacts.

 1.3  INDUSTRY PROFILE

      This profile of the AIM coatings industry describes
 commodities and VOC content, demand for AIM coatings,
 production of AIM coatings, and industry conditions.
 1.3.1  Commodities and VOC Content
      Paint products are manufactured in four main product
 categories:  architectural coatings, finishes  for original
 equipment manufacturers (OEM),  special purpose coatings, and
 miscellaneous allied paint products.3   Because the regulation
 of consumer and commercial products does not apply to the OEM
 category because the products  are for industrial uses, it will
 not be discussed in this profile.   Paint product categories
 under consideration for regulation via a national regulatory
 negotiation rulemaking are listed in Table 1-1.4   The products
 are grouped into the three Census  categories in which they are
 found.b  The  largest quantity of potentially regulated
 coatings is included in the architectural coatings category,
 but some coatings  are  classified with  the special purpose and
 allied paint  products  categories,  which also include other
 products not  covered by the AIM coatings regulations such as
 marine  paints  and  putty.
      Examples  of architectural  coatings,  all of which are
 represented in Table 1-1,  include  exterior and interior
 organic solventborne and waterborne  tints, enamels,  under-
 coats,  clear  finishes,  stains,  and architectural  lacquers.
Architectural  coatings  are  used for  general  purpose  on-site
application to residential,  commercial,  institutional,  and
     bSee Appendix A for a detailed explanation of products proposed for
regulation and their corresponding Census classification.

                              1-5

-------
        TABLE  1-1.   AVERAGE VOC  CONTENT FOR AIM  COATINGS
                TO  BE  COVERED BY PROPOSED REGULATION
                                               Sales-weighted average VOC content
                                               	(g/L)         	
 Product category
Organic solvent
Waterborne
 Architectural coatings
    Exterior flat  architectural coatings
    Exterior nonflat  architectural coatings
    Interior flat  architectural coatings
    Interior nonflat  architectural coatings
    Semitransparent stains
    Opaque stains
    Undercoaters
    Primers
    Sealers
    Waterproofing  sealers, clear
    Waterproofing  sealers, opaque
    Quick dry undercoaters, primers,  and
      sealers
    Bituminous coatings
    High performance  architectural coatings
    Roof coatings
    Lacquer
    Varnish

 Special purpose/industrial maintenance
    Coatings
    Swimming  pool  coatings
    Dry  fog coatings
    Mastic texture coatings
    Metallic  pigmented csatings
    Fire retardant coatings
    Antigraffiti
    Concrete  curing compounds
    Form release compounds
    Graphic arts coatings
    High-temperature coatings
    Industrial maintenance coatings
    Multicolored coatings
    Pretreatment wash primers
    Sanding sealers
    Shellacs
    Traffic marking paints

 Allied  paint products
    Below  ground wood preservatives
    Semitransparent wood preservatives
    Clear  wood preservatives
    Opaque wood preservatives
      336
      404
      315
      413
      527
      429
      379
      374
      607
      659
      242
      441

      290
      431
      269
      667
      481
      554
      365
      278
      461
       a
      577
      717
      601
      386
      560
      392
      321
      718
      531
      539
      398
     541
     591
     493
     446
    68
    76
    48
    74
    85
    56
    41
    48
    41
   200
    a
    31

    4
   113
    28
   300
   143
    a
   149
   107
    a
   23
   131
   71
    a
   42
    a
   112
    a
    a
   192
    a
   85
    a
   67
   419
    a
*  Sales-weighted average VOC content not available.

Source: Industry Insights.  Architectural and Industrial Maintenance  Surface
        Coatings VOC  Emissions Inventory Survey. Prepared for the National
        Paint and Coatings Association in Cooperation with the AIM Regulatory
        Negotiation Industry Caucus.  Final draft report.  1993.
                                     1-6

-------
 industrial structures.  They are  intended for ordinary use and
 exposure and provide protection and  decoration.
      Special purpose coatings are similar to architectural
 coatings in that they can be classified  as stock  or  shelf
 goods, rather than formulated to customer specifications,  as
 are OEM coatings.  The difference is  that they are formulated
 for special applications or environmental conditions  such as
 extreme temperatures, chemicals, fumes,  fungi, or corrosive
 conditions.
      The VOC content of AIM coatings varies  substantially
 between specific types of products.  Most of this variety is
 due to the type of solvent used in the coating and the  ratio
 of  the solvent to other ingredients in the formulation.   The
 sales-weighted average VOC contents for potentially regulated
 coating products are listed in Table 1-1.Cf5
 1.3.2  Demand for AIM Coatings
      1.3.2.1  Conceptual View of Coating  Decision.  The demand
 for AIM coatings derives from the demand  for  the  treatment  of
 architectural and industrial (AI)  surfaces.   Surface treatment
 services include not only coating treatment, but  also
 noncoating treatment alternatives such as wallpaper or
 exterior siding.   While the choice among coating alternatives
 is  emphasized below,  it is implicitly recognized  that the
 substitution between coating and noncoating surface treatments
 is  possible as well.
      The coatings themselves are an input into the production
 of  surface treatment services,  the final product of interest.
 Each AI surface possesses  certain attributes that affect the
      °Sales -weighted average VOC content is
                      (VOC Content )i -  ( Sales )i
                            n
                           £ (Sales).
                           i-l

where VOC content is equal to the percentage by weight, sales are measured
in pounds per year, and n equals the number of product categories.

                              1-7

-------
demand for  surface treatment.   These  include surface material
 (substrate),  age,  exposure (e.g., weather,  chemicals),  and
other  physical  factors  that  intrinsically affect  the relative
performance of  treatment  alternatives.
     In an  economic decisionmaking  context,  we  think of the
owner  of the surface as seeking to  maximize the utility
derived from the  services provided  by the surface (i.e.,
shelter,  decoration,  etc.).  If we  abstract to  the  situation
where  all nontreatment  attributes of  the  surface  are given,  we
can characterize  the treatment  decision in the  context  of  a
utility maximization problem.   We assert  that the owner of a
particular  surface seeks  an optimal treatment alternative — one
that maximizes  utility, subject to  constraints  on the cost of
treatment.
     Let process  i indicate the activity  of  treating a  surface
defined by  the  attributes  above.  Through this process,  labor,
capital,  and materials  are employed to treat the  surface.
Thus we can characterize  the production of  a unit of  surface
treatment through  process  i as  follows:
where Qi is the surface area unit (e.g.,  1,000 ft2)  treated
using process i and L, K, and Xi are the  quantities of labor,
capital, and material  (e.g., coatings) used to produce Qi.
     For the processes that include coatings application, we
assume a fixed proportions relationship between each input and
output, determined by the type of coating being used.  For
example, process A requires 1 gallon of coating A, 40 hours of
labor, and 10 units of capital to cover a unit area of a given
surface type.  Therefore, for a given set of input prices,  we
can compute the (constant) per-unit cost  of treatment.  Costs
of noncoating alternatives can be similarly computed.
     Considering all n possible treatment alternative for a
given surface generates an array of costs (ClfC2, . . . ,Cn) .
                              1-8

-------
      Each owner/consumer places  a subjective  value on the
 outcome of each treatment alternative.   This  value derives
 from such factors as innate  preferences  for the  visual appeal
 of treatment alternatives and  perceptions  of  the structural
 quality and durability.   For example, consumer A may prefer
 the look of glossy solvent-based coatings  to  flat water-based
 coatings and/or may perceive other differences in product
 quality.  The consumer explicitly or  implicitly  monetizes
 these preferences,  and the associated monetary values for each
 of the n alternatives comprise the array of perceived benefits
 for (BlfB2,	Bn) .
      In evaluating the choice  among treatment alternatives,
 the consumer weighs each alternative's monetized benefit,  Bi,
 against the cost of treatment, Ci.  The subjective payoff  from
 each alternative can be  expressed as

                         *i = Bi  - Ci  '

      The consumer maximizes  utility with respect  to  the
 surface treatment choice  by  selecting the alternative  with the
 highest payoff.   This of  course presumes that at  least one of
 the payoffs is not  negative.    If  all potential payoffs are
 negative,  the consumer is better  off by choosing no  surface
 treatment  at all.
      1.3.2.2 Aggregate Demand and Substitution Effects.   If
 all consumers' preferences were identical and all AI surfaces
 to  be  treated possessed the same characteristics, the  consumer
 choice model above would predict only one optimal type of
 surface treatment throughout  the economy.  A wide array of
 treatments  and coatings are applied,  indicating a variety  of
 surfaces with different characteristics  as  well as individual
preferences  that vary across  consumers.
     The purpose of this  analysis is  to  determine how
consumption  choices may change  in response  to  any price and
product quality effects  of the  proposed  regulations.   If the
                              1-9

-------
regulations  induce  a  change  in  the price schedule of various
AIM  coatings,  the unit  costs  of  treatment alternatives will be
directly affected.  Furthermore, the regulations  may induce a
change  in the  structural characteristics of the coating that
alters  the application  technology.  For example,  a different
VOC  content  may change  the volume of the coating  that must be
applied and  the amount  of labor  and capital necessary to
achieve the  same surface area treatment; consequently,  the
technological  parameters may change with the new  VOC
requirements.  Therefore, treatment costs will be affected
jointly by what we  call the factor price effect and the
technology effect.  If, for example,  the VOC-content
regulations  would raise the price of the affected coatings  and
reduce  the technological efficiency of the treatment process
(e.g.,  more  coats necessary), then both the factor  price
effect  and technology effect would combine to increase  the
cost of the  affected treatment alternatives,  generating a new
set  of  treatment costs  (Cl> , . . .,Cn' ) •
     VOC-content regulations may also affect consumer
valuation  of the treatment alternatives through a change in
visual  characteristics and altered perceptions of quality or
durability.  These changes generate a new set of  subjective
values  for the treatment alternatives of (B11, ...,Bn' ) .   As a
result,   evaluating the new arrays of  benefits and costs
produces a new array of treatment payoffs,  (^l' . .  . , itn' ) .  The
consumer can again be expected to select the  treatment
alternative with the highest payoff.   This  situation may
produce a different optimal  selection than  the no-regulation
case.   The consumer may in fact choose a noncoating
alternative or no-treatment  alternative,  where coating
treatment would be selected  without the regulation.
     Aggregating over all consumers and all surfaces, we can
see how the regulatory changes can induce substitution among
treatment alternatives and changes  in aggregate demand for the
affected coatings.   These aggregate changes in demand and the
                             1-10

-------
 associated effect on consumer welfare are where we focus our
 attention in this study.
      1.3.2.3  Substitution Possibilities.   The total reduction
 in emissions as a result  of reformulation is  influenced by the
 availability of substitutes.   Because of  the  potential for
 users to substitute one product for another in some
 applications,  the total effect on VOC emissions of a change in
 formulation will also include an indirect  effeet—mediated
 through relative price changes—due to the change  in product
 mix.   For example,  one could  choose from  five product classes
 to paint the exterior of  a building (waterborne paint,  organic
 solventborne paint,  waterborne stain,  organic solventborne
 stain,  and wood preservatives) .   Regulation is likely to
 affect the formulations and manufacturing  costs of those
 products differently,  leading to  product substitution by
 users.
      All coatings users have  three  substitution possibilities
 that  will lead to a  reduction in  the  amount of  VOCs  emitted
 from  coatings.   They may  substitute a  coating  that  contains a
 smaller percentage of organic solvent  for  a coating  that
 contains a larger percentage  of organic solvent.   They may
 substitute a waterborne coating for an organic  solventborne
 coating.   Alternatively,  users could choose a coating-free
 surface and avoid using a coating altogether.   Table  1-2
provides examples of  these substitution possibilities.
Materials  such  as aluminum, vinyl siding,  and glass are
additional  exterior substitution possibilities for
architectural coatings.
      The substitution of one coating for another may occur
during  the  initial application or subsequent maintenance or
remodeling.  However, certain decision-making  constraints may
be imposed  on the consumer.  For example,  although stain is a
substitute  for paint, it may only be a practical option where
a rustic, weathered look is desired.  This option may be only
for particular surfaces, house styles, or  settings; therefore,
the decision to choose paint or stain is likely to be

                              1-11

-------
   TABLE  1-2.   EXAMPLES OF USERS'  SUBSTITUTION POSSIBILITIES

Substitute:
For:
Substitute
organic
solvents
Varnish
(organic
solvent)
Lacquer
(organic
solvent)
Substitute
types of
solvent
Flat exterior
paint
(waterborne)
Flat exterior
paint (organic
solvent)
Substitute
surfaces
Wallpaper, wood
paneling3
Nonflat interior
paint (organic
solvent)
   Difference in
   VOC contentb
1.55
2.23
3.44
  a VOC-containing products used to apply these substitute surfaces not
   considered for difference in VOC content.
  b Average VOC content measured in pounds per gallon.

influenced  by practical circumstances as well as price
differentials,  brand characteristics,  and VOC content.
Practical considerations are also  important when a coating has
been applied previously.  For example,  if the previous  owner
of a house  used paint, the present owner might be unlikely to
switch  to stain, regardless of price  or his/her preference for
stain.d
     In comparison to architectural coatings, special purpose
coatings are much less uniform in  protective function and
offer fewer substitution possibilities  between product  types.
Although, as shown with some of the special purpose/industrial
maintenance products in Table 1-1,  substitution between
waterborne  and organic solventborne products may be an  option.
Furthermore,  a coating application is actually a system of
components  such as primer and topcoat.   Therefore,  a full
     dlf  the difference  in price between paint and stain was  so great that
it outweighed the cost of preparing a painted house for  stain, it could
influence the coating decision.

                              1-12

-------
 analysis of substitution in response to regulation should
 consider the potential for paired effects--simultaneous change
 in both primer and topcoat.  For example,  a consumer who needs
 to paint bare wood typically uses a solventborne primer for
 better adherence because wood absorbs the  oil-based solvent
 better than it absorbs the waterborne solvent.   A waterborne
 or solventborne topcoat may then be applied.6
      The substitution for another surface  occurs only during
 new construction or major remodeling.  An  architect or an
 engineer will frequently decide about substituting another
 surface,  whereas a nonprofessional is likely to decide about
 the substitution for another coating.  Substitution decisions
 will be made by two different types of decision makers using
 different sets of criteria.
      1.3.2.4  Characteristics of Latex and Alkvd Paints.
 Since the late 1940s,  manufacturers have substituted water for
 some or all of the organic solvents in several  architectural
 coatings.   Waterborne  coatings are commonly referred to as
 latex and organic solventbornes as alkyds.   Some advantages  of
 waterborne coatings are that water is cheaper than organic
 solvents  and waterborne coatings are easier to  clean,  are
 almost odorless,  have  reduced toxicity and flammability, and
 emit fewer VOCs.   Waterborne coatings account for  over 65
 percent of all architectural coatings,7 but only 30 percent of
 VOC  emissions from architectural coatings.8  At the present
 time,  waterborne  coatings  also have a shorter shelf-life,
 lower gloss,  and  lower abrasion resistance  compared  to organic
 solventborne coatings,  and they should not  be stored  outdoors
 where temperatures fall below freezing.
      As suggested in Table  1-1,  although it is possible to
 substitute latex  coatings  for  alkyd and achieve a reduction  in
VOCs,  the  degree  of substitutability can vary depending on the
application.   Therefore, latex paints may be less acceptable
 to some users  for  certain applications  than alkyds.  One point
of controversy is  whether  there  are  certain applications for
which latex  architectural paints cannot be  substituted for

                              1-13

-------
alkyds  without sacrificing performance  quality.9  Paint
manufacturers market  a variety of  paint products, both  latex
and  alkyd,  to meet  consumer demand for  various characteristics
other than  price, including performance, appearance, and
convenience characteristics listed in Table 1-3.10~14  Examples
include "Sun-Proof  Gloss,"  "All-Weather Gloss," and
"Durability Plus  Latex Stain."15"17  Most products are
distinguished by  latex or  alkyd type, gloss level, and brand
series.
     The characteristics of paint  products provide consumers
with utility,  and several  products  on the market within
brands,  between brands, and between latex and alkyd types
share many  characteristics.  The fact that a coating
application is actually a  system of components explains the
point that  a combination of goods  (primer and topcoat)  can
provide a characteristic (i.e., corrosion resistance) that
neither provides  (at  least  very well) alone.  In choosing a
paint product,  the  consumer is  interested in the array of
desired characteristics provided by each good,  and each
characteristic may  carry different weights for different
consumers.
     Several of the key characteristics that are important to
coatings  consumers  are  listed in Table 1-3.   Some
characteristics such  as gloss and hardness complement each
other,  whereas others such  as leveling and resistance to
sagging oppose each other,   so that a single paint is not
likely  to have both characteristics simultaneously.   Choosing
paint involves trade-offs,   and a consumer's choice depends on
how  he/she  ranks preferences for these characteristics.   The
degree  of substitutability possible between one paint product
and  another  depends on how  closely the characteristics  offered
by each are  substitutable.   No one paint,  whether solventborne
or waterborne,  can achieve all characteristics  well,  and
overall paint  quality, regardless of paint type,  is  an
                             1-14

-------
                  TABLE 1-3.  PAINT CHARACTERISTICS
 Characteristic
                  Definition
Typea
 Ease  of
   application

 Adhesion


 Leveling

 Hardness


 Scrubbing

 Sagging


 Water resistance

 Rust  resistance

 Blocking



 Gloss



 Price
 Ease of brushing and spreadability.             C


 Ability to stick well to poorly primed or      P,C
 chalky surface.

 Smooth finish,  free of brush marks.             A

 Paint film dries hard and resists denting,      P
 scratching,  marring.

 Dried surface resists repeated scrubbings.      P

 Tendency for paint to run,  drip,  or sag         P
 like a curtain during application.

 Impervious to standing water.                    P

 Rust resistance  without a primer.               P

 Tendency for paint to remain tacky              P
 and  sticky even  after it is dry and adhere
 to things  placed on it.

 Measurement  based on a gloss meter: flat       A
 being the  dullest,  followed by eggshell,
 satin,  semigloss,  gloss,  and high gloss.

 Manufacturers' suggested retail price per     N/A
 gallon.	
 Color-specific

  Hiding


  Fading
  resistance


  Mildew
  resistance

  Chalking
Ability for paint  to hide  substrate well       P
with one or two coats.

How well paint retains original color and     A,P
gloss when exposed to sunlight  and the
elements.

How well a paint resists the buildup of        P
mildew.

Erosion of resin binders,  which leaves        P,A
behind a light powdery dust  (pigment).   Can
be a good characteristic for white paints
in dirty areas but never advantageous  for
colors.
"Type  = Appearance  (A), Convenience (C),  and Performance (P) .

N/A =  Not Applicable.

Sources: Consumer Reports Magazine.  House  Paints and Stains.   52(61=365-374.
          1987.
        Consumer Reports Magazine.  High-Gloss Enamels,  p.  173. March 1988.
        Consumer Reports Magazine.
          1989.
        Consumer Reports Magazine.  Paints for Finishing Touches.
          September 1990.
        Consumer Reports Magazine.  Interior Latex Paints,  p.  333.  May
          1991.
              Interior Semigloss Paints,  p. 317.   May

                                          p.  619.
                                  1-15

-------
 important factor.   Price differences generally are accurate
 indicators of  quality differences.18  Many customers have
 problems with  paint because they use cheap  or  inappropriate
 paint  and are  not  properly prepared.19  Paints may be evaluated
 based  on the characteristics in Table 1-3 to determine how
 substitutable  a latex product may be for an alkyd  for  a
 particular application.
     1.3.2.5  Substitutability Between  Latex and Alkvd Paints.
 Substitutability of the  characteristics listed in  Table 1-3
 between latex  paints and alkyd paints is the focus of  the
 following discussion,  although a brand  line also may offer
 several paints in  which  these characteristics  vary widely.
 Analysis of ratings articles from Consumer Reports Magazine
 summarized in  Table 1-4  shows characteristics  in which
 solventborne paints tend to consistently perform better and
 characteristics  in which waterborne paints tend to consis-
 tently perform better.6'21"24  For the analysis  in Table 1-4 a.
 product possesses  each characteristic if it was rated
 excellent or very  good in the Consumer  Reports testing.  Their
 ratings also indicate, in general, how  many paints  a consumer
 would  have to  choose from if he/she were seeking in a  latex a
 characteristic that is usually found  in an alkyd.   Conclusions
 may be drawn from  the  analysis  regarding whether there  are
 certain applications for  which  latex  paints on the market
 cannot perform as  well as  alkyds.  Table 1-4 summarizes  the
 substitution options between latex and  alkyd paints for
 exterior trim paints,  interior  semigloss,  high-gloss enamels,
 and house paints and stains  analyzed  in Consumer Reports.
 These  specific product categories are where substitution
between a latex and alkyd paint is most likely.  Latex  flat
wall paint is  the  most popular  choice for general interior
painting (76 percent of  interior sales)  , and Consumer Reports
did not  test any flat  alkyd  paints in the articles pertaining
     eAn article in Sunset Magazine20 concurs with the conclusions drawn
from the Consumer Reports ratings.

                             1-16

-------








EH
2
H
£
ALKYD
Q

03,

B
E-<
3



w
1^*
t^
W
CQ
CO
U
H
EH
CO
H
rv*
DH
M
EH
u
1
£
u

fa
o
2
o
(J
CO
M
CH
<
CU
S
o
0


»»
1
1
1-t

W
J
ffl
g






















j?
•o
0*

^-
c
01
>

^1
o
AJ 01
c u
01 -H
rH AJ
rH 01
4) *H
U kl
X 01
4> AJ
U
•O 10
t) kl
AJ 10
10 X
ki U
01 01
AJ AJ
C kl
•H O
10 a
a o>
a:
-o
> M
rV 01
rH E
10 3
01
•O C
C O
15 u

X
Ol
AJ
10
rH
VM
0
kl

•s
z





















c
o
-»4
AJ
a
u
>H
rH
a
to
rH
a
u
«H
1!
•I
0 °
Jj rH
13 «
* S
•
n
a
o
rH
O


- «g
0 0 5
AJ rH "H
ffi
01
U
C AJ
10 n
AJ S
01 K
•»4
01
01 ki
K a
i


1 Ol
Ol C
<0 *r<
W C
t
A 01
3 C
kl *r<
U A
w

« 01
|i
£


'
||
01
J

ra <=
S 0
!•

i
*? i
0 3 c
... -H C
SaS
"J a
U Ml

ki -O
« t
£ I
3 01
Z AJ



AJ C
c :
•H T
a o
a, .





x
^1

§
•o
o
i
E
.r4
kl
AJ
Interior























































uterior
rH
E
- § e oJ e"
CO kl 01 - " 71
k, .C .C 01 " "2 £ !1
O AJ O ki e « O AJ
0 $ AJ 0 " £ W F
"O •" -H ° Sen
•* "O a E .t S
Tl Oi O
0 C - - 01 " » M
•rt 10 W £ c- t ^ P J-
C£ -2 01 J301AJAJ
««r>^ vj kl ffl kl "4 (0 01
§££ ££ 2-HCJ3AJ
S?U S -301-. 0101
OOJ? 01AJ ^OI=-CC
SAJW to3 TJUHW-H-H
isi l« fe.m
-Sm. ^4 CJ3
gSiiSS -3S"g|S
||j5t« ^6'I!'S
^.S-S^ S5 Sg-ls s
ESSOXS |MiJS>X(2
£3 VO UJ 0 « O" °°
° — „ „ « 
S3 N S N fH -H

< < m
zz "



-oo 2* S- |



<• < < < <
55 -; -^ m o 5
Z Z Z Z 2

J, -. Iff
rH „ * * »*> 0 *
ZZ -° 2



"S °S

<< ^^ ^
^ ^ 22 22 2

^* rf  in _, »
rH CM rH CM rH ^ rH




U Oil 1J
O rH 0
llOllTJX-4 -OX OlIOI'OX-HOIAJl'O
rjOi>,01 kl£i>lOI W £ >, 11 klOldrS
golJ?AJ 5^j£AJjao]irjJi:Aj oia-H-*:
MrHlrHIO AJklrHQ OtrHJCrHlO AjO«rH
3 d< J x !!< J ^ d 51 < J * a °3 <
wi aa u








•a
o
i
r<
o
^4
kl
tl
AJ
X
U
in n n
rs v-i ts
i i i
LO « n
rH rH «"«

0 * *
"•* *** "*fc
rt Z Z



< < <
^ »» »»
z z z



< < <
"^ ^ ***
z z z

•c < <
^ ^* ^
z z z


_ < <
°J ^ --.
- z z

< < <
»» -V >»
z z z

< < •«
-^ "^ *^
z z z



< <
O "^ ^
z z


M ^
VO rH "^





s «•» •>
M




oo n «
(S rH rH




vj
O tH "i|
x .H a c -o x
0) kl Ol -H > t)
AJ v 
10 kl
- £. t)
01 • A
AJ >, r- • E
ki 01 oo m oi
Of. • en co AJ
a AJ C rH c* a
41 O rH 01
£t D — < -CO
01 AJ . 00 >,
kl 3 -H 9 00 10
oi 10 c r- <* X •
E O "4 m rH CD
3 01 vu | rH
u ft o> m si • vo
c TJ \o o r-
O 01 (-1 kl rH •
O oi • •• (0 m a
CO) — s
O -H AJ VO
AJ rH kl — a •
O (Nl • 01
O1T3 Q, ml m 0)
B C Oi r- • £
-H 10 Ct rH 01 U
•O ki AJ 3
kl A kl 01 • C O
oo; c a-4 E-
O 01 E -r4 10
o £ 3 e a, 01
e AJ 01 AJ • c
c co m in ^H
01 VW O rH 01 X
01O U TJ 4) o 01
O C 6 "H •"•
rH >a o a 10 01 c
Ol C AJ c -H T<
oi m u E Ou
kl Ol AJ 0>
O ki C C 01 W ki
01 -r« -rl 01 O
01 £ TJ 10 O ki vu
01 Ol kl O. rH O
O -4 O O ^4 01
rH X U 01 1 kl AJ
Ol U 01 .C 01 C
01 <0 3 Ol AJ -H
a j: o -H c o
Cn AJ 01 i x M a,
— , 01
.= e o
O rH . ....
Tiki Ol 01 D 01 D 11
au -H 5, cccc
£ E S -r( ~4 -rl -r<
01 0) AJ N N N N
AJ AJ 01 10 10 10 10
<0 C AJ Ol Ol Ol Ol
.C-rl 01 C 1010(010
AJ(O k. .H zaexx
a oi Q
TJ 3 a 01 01 01 01
t> 0) AJ AJ AJ AJ
AJAJ AJ O klklklkl
010)*<0 AJ OOOO
VOIAJ.C a S, a a
AJAJ01AJ 41 OlOJtlO)
oi rH K re ce DC
01 n AJ 01 ja
AJAJ AJ O klklklkl
Cki>>C U 11 01 4) 01
-* 2 8 "2 -H EEEE
10 a £ 10 rH 3333
aoiAja a oioioim
a. a c c c c
«J nvu o OOOO
OkiOlO OtJOO
0) -< AJ
kl E AJ kl O
tl 3 -rt 01 Z 01
A W rH ^ 41
E c 10 £ n u
3033 ki
z o cr z < 3
-^ o
• o u z en
1-17

-------
 to  products used for general  interior wall  painting.25'26   Flat
 alkyds may be desirable  in certain cases  for  interior walls
 because they are slightly  more washable and abrasion
 resistant.27
      Table 1-4 shows that  different characteristics  are
 important to different types  of paint.  For example,  hardness
 was a characteristic rated only for high-gloss enamels.   This
 suggests that the type of  surface  or application will
 influence a consumer's set of  preferred characteristics.
 Table 1-4 identifies typical  applications for each paint
 product.   According  to the Consumer Reports tests, a  consumer
 looking for good leveling  in  a latex paint would have little
 to  choose from,  except with the interior semigloss for which
 six were rated very  good or excellent out of 21 tested.   For
 resistance  to blocking, however, at least one latex paint in
 each product  category performs  well.  The summary does not
 show how many latex  products rated well in both blocking
 resistance  and leveling,  which  is  important to consider if
 they are popular  consumer priorities.
      As  Table 1-4 shows,  the high-gloss enamels product
 category may  be  the most difficult  in which to find latex
 substitutes.   When tested  in 1988,   Consumer Reports found
 alkyd high-gloss  enamels more prevalent in the marketplace and
 only tested four  latex products in  this category.   A primary
 attribute of  alkyd high-gloss enamels is a glass-like  finish,
which implies  excellent leveling.  A  Popular Science article
 stated when referring to high-gloss enamels that "Currently
 there is  no substitute—either a reformulated alkyd or a
 latex—that exhibits its qualities of smoothness and high
 initial gloss, although acrylic enamels are coming close and
 actually  retain their sheen longer."28
      Some color-specific  characteristics,  not  represented in
Table 1-4,  are hiding ability and resistance to fading,
mildew, and chalking.  Performance in these areas  varies by
color, but, in general,  latex paints resist mildew and fading
                             1-18

-------
 and chalking from exposure to ultraviolet  (UV) sunlight and
 weather better than alkyd paints.
      Most of the characteristics in which waterborne paints
 have the advantage tend to be convenience related, such as
 ease of application,  decreased toxicity and flammability,
 quick drying time, ability to apply to a damp surface, low
 odor,  and easy clean-up.  With organic solventborne paints,
 the advantages tend to be performance or appearance related,
 such as resistance to blocking,  good leveling, and high gloss.
 An  EPA Region IX report on nonflat architectural coatings
 (e.g.,  eggshell,  high gloss)  supports the characteristic
 ratings for alkyd and latex paints found by Consumer Reports.29
 The report concludes  that,  for most uses of nonflat
 architectural coatings,  latex paints are available that meet
 consumer expectations and can substitute for alkyd paints.
 They noted similar areas where nonflat latexes tend not to
 perform as well as alkyds:  high-gloss applications,  resistance
 to  blocking,  leveling,  and adhesion.  Some latex paints that
 perform well  in these areas are  available.   The EPA report
 also notes that latexes  retain their gloss  better over time
 and will adhere as well  as  alkyds to a properly prepared
 surface.   It  is possible, however,  that a consumer's  set of
 preferred characteristics may place high initial gloss over
 lasting gloss and a priority  on  reduced priming.   For their
 report  the EPA surveyed  48  paint  contractors in California,
 and a minority (21 percent) used  more  nonflat  latex  than alkyd
 paint.   Contractors suggested that,  although latex convenience
 characteristics are advantageous,  they are  not  the determining
 factors  behind their  choice of paint.   Contractors were more
 concerned with the appearance and performance  characteristics
 exhibited most often  in  alkyd paints.f
     Consumer Reports advises that,  for applications where
performance characteristics between the two  types  of paints
     fThe report acknowledges that contractors surveyed had opposing views
regarding characteristics of latex and alkyd paints.

                             1-19

-------
are  about even,  such  as  interior walls  and exterior  surfaces
in good condition,  the consumer would be  better  off  with the
latex paint because it is  always more convenient to  use.  An
alkyd paint may  be  preferred  when application  to a poorly
prepared surface is necessary,  when  initial high gloss and
smoothness are important,  or  where the  surface to be painted
must repel frequent standing  water.   Another case where  alkyds
are  preferred is priming bare wood.   Bare  wood absorbs organic
solvents better  than  water, and many painters  choose a
solvent-based primer  in combination  with a water-based
topcoat.30  If an application demands a  combination  of  the
characteristics  found in an organic  solventborne  paint,  the
consumer has  a few  latex substitution possibilities.  Consumer
Reports  tests indicate that an  informed consumer  will
generally be  able to  find  a couple of waterborne  products to
meet performance characteristics most often found in organic
solventborne  paints.  Good resistance to blocking may be more
important to  a consumer than  easy  clean-up, but consumers have
incentive to  find paints on the market that do both.
      1.3.2.6  Coating Users.  Users  of coatings can be divided
into two groups:  professionals and  nonprofessionals.  The
nonprofessional  is  typically  a  "do-it-yourselfer" who pur-
chases only a small amount of coatings each year.  The
application of coatings by nonprofessionals is limited
primarily to  residential architectural coatings.   Professional
users of coatings may be professional painters or contractor/
builders.   These professionals apply coatings to a broad array
of surfaces in residential, commercial,  institutional,  and
industrial  settings.  Table 1-5 shows that in 1989 do-it-
yourselfers purchased two-thirds of all  residential  archi-
tectural coatings.31  It seems reasonable to assume that
contractors purchased all of the nonresidential architectural
coatings and  thus accounted for 60 percent of the use of  all
architectural coatings.
                             1-20

-------
         TABLE  1-5.   CONSUMERS OF ARCHITECTURAL COATINGS

                                   Percentage of total
       	gallons in 1989 (%)	
        Residential
          Do-it-yourselfers                 41
          Contractors                       20
        Nonresidential*                     39
        Total                              100
        * Commercial,  institutional, light industrial.
        Source: SRI International.  U.S.  Paint Industry Data Base.
               Menlo Park, CA.  1990.

      Data that break down coating consumption by user group
 were published for  the year 1982.9  Table 1-6  shows that,
 including imports,  37 percent of domestic trade of paints and
 allied products would be  considered a potentially regulated
 usage given  the 1982 data.32   Commercial users represent the
 largest potentially  regulated category, using  30.7 percent.
 Potentially  exempt uses in 1982 accounted for  63.1 percent of
 total domestic trade in paints and allied products.   Coatings
 for original equipment manufacturers are included in these
 statistics;  however,  because they are used during the
 manufacturing process,  the entire category would be considered
 an  exempt use.  Approximately 29.3  percent of  the total used
 by  industry can be attributed to OEM coatings,  leaving 42.8
 percent of industry  uses  for architectural,  special  purpose,
 and allied paint products.   Exports in 1982 comprised 3.1
 percent of total domestic  trade.
     9These data include consumption of all products produced by SIC 2851
Paint and Allied Products, which includes AIM coatings, special purpose
coatings, marine paints, and allied paint products  (i.e.,  brush cleaner).

                               1-21

-------
  TABLE 1-6. COMMODITY USE IN 1982:  PAINT AND ALLIED PRODUCTS
 Users
 Value3
 ($106)
Percentage  of  total
domestic tradeb  (%)
 Potentially regulated
    Household consumers
    Commercial
    Government
    Total
 Potentially exempt
  323.0
2,629.1
  212.5
3,164.6
        3.8
       30.7
        2.5
       37.0
Industry
Exports
Other0
Total
Total domestic trade
5,197.7
262.6
-47.7
5.412.6
8,577.2
60.6
3.1
-0.6
63.1
100.0
a Commodity output measured at producers' prices.
b Domestic trade is domestic output plus imports.
c May include special industries, final use change in business inventories,
  and gross private fixed investment.
Source:  U.S. Department of Commerce.  The 1982 Benchmark Input-Output
        Accounts of the United States. Washington, DC:  U.S. Government
        Printing Office.  1991.
1.3.3   Production of  AIM Coatings
      1.3.3.1  Raw Material Inputs.   Coatings comprise four
basic  types of materials:   pigment, resin  (binder),  solvent,
and additives.  Pigment is the solid component  consisting of
uniform particles of  a  controlled size that are insoluble in
the vehicle (the liquid portion of the coating).   Pigments are
used  in coatings to decorate and protect and as fillers.33
Pigmentation,  although  it  varies depending on desired
properties, is similar  in  both waterborne and solventborne
formulations.
     Film-forming binders  surround and hold together the
elements of the coating film and make up the nonvolatile
portion of the vehicle.  Resins aid in adhesion; determine the
cohesiveness of the dried  film; affect gloss; and  provide
                               1-22

-------
 resistance to chemicals,  water,  and acids.   Natural and
 synthetic resins and oils,  along with certain additives such
 as driers and plasticizers,  serve as binders in coatings and
 are one of three types:   multiuse resins (acrylics,  vinyls,
 urethanes, polyesters);  thermoset resins (alkyds,  epoxies);
 and oils  (drying oils,  bodied oils).
      The vehicles in organic solventborne and waterborne
 paints differ not only by the type of resin used,  but also in
 the way they form a film and dry (or cure) .   Alkyd paints are
 oxidizing film formers in which  the  drying  oils react with the
 oxygen in the air when the paint dries.   The chemical reaction
 binds the molecules of the vehicle into  a hard,  dry film.
 Alkyd coatings continue  to oxidize long  after they dry and
 eventually provide a rock hard surface.   Latexes consist of
 tiny,  heat-sensitive plastic particles (latex)  that  are
 dispersed but not dissolved  in water  along with the  pigment.
 As the water evaporates,  a layer of  closely  packed plastic
 particles and pigment is  left behind.  The softened  plastic
 particles then lose their shape  and molecules  diffuse and
 reattach to form a binding film.34  The chemical
 characteristics of latex  and alkyd paint  influence some of
 their  characteristics, such  as gloss  and  resistance  to
 blocking and water.   Heat-sensitive plasticizers in  latex
 paint  cause the residual  tackiness called blocking,  which is
 more of  a problem in glossy  latex paints  where  the ratio of
 resin  to pigment is  higher.   Precise  control of particle shape
 and size in the film former  is necessary  to  increase gloss.
 The plastic mesh also breathes better, allowing water and air
 to pass  through it.   The oxidizing process of alkyds forms a
 smooth (thus  glossier) , watertight skin of hardened resin that
provides  durability  and water resistance.
     Petroleum  distillates in alkyd paints and ,water in latex
paint  function  as  the carrier, or volatile vehicle, that
disperses  the pigment and resin and provides the necessary
fluidity  for  applying the coating.  Basically there are two
types  of  solvents:  water and organic.  In alkyd paints

                              1-23

-------
organic solvents dissolve the components  of  the  film  former,
keeping them in solution.  In latex paints,  water  separates
and  suspends the droplets of film  former.  Following
application,  the evaporation rate  of  the  particular solvent
controls the rate at which the film forms, leaving the pigment
and  resin bonded to the surface.   Latent  solvents, which
dissolve the film former when combined with  true solvents,  and
diluents may be added to the true  solvent.35   Diluents can be
blended with the dissolved solution to extend the true and
latent  solvents.  Water is the true solvent  used in latex
paints  but may function as a diluent  in alkyd formulations.
Three types  of organic solvents are used  in  coatings:
hydrocarbons  (aliphatic, aromatic); oxygenated solvents
(alcohols, esters,  ketones,  glycol ethers);  and chlorinated
solvents (1,1,1-trichloroethane,  methyl chloroform).36
Architectural solventborne paints are mainly formulated with
aliphatic hydrocarbons.
     Additives are used in relatively small amounts in both
organic  solventborne and waterborne formulations to provide
additional necessary properties or augment the properties of
other inputs.  They may be added to the film former,  solvent,
or pigment.  Waterborne paints in particular may use additives
such as  agents to reduce foaming or bubbling of paint when  it
is shaken and applied;  wetting agents, which can improve
pigment  dispersion or adhesion;  freeze-thaw agents,  which
reduce  the temperature at which the paint will freeze to
prevent  coagulation;  and coalescing agents, which aid the flow
of the  latex particles to form a more continuous  film.37  VOC
contents  in latex paints (4  to 10 percent, or 50  to 200 g/L)
are due  to the additives used.38  Solvents  such as alcohols  and
ethylene  glycols are added as  co-solvents to  waterborne
formulations.  They are  often  necessary to allow  the  plastic
particles  to soften and  be mobile enough to bind  into a
continuous film.39
     The  additives  used  in the  largest volume are thickeners,
fungicides and preservatives,  plasticizers, and defearners.40

                             1-24

-------
 Figures 1-2 and 1-3 show the principal  raw material
 ingredients discussed above as they  are used in organic
 solventborne and waterborne coating  formulations .
      1.3.3.2  Formulations.  One of  the distinguishing
 characteristics of each coating is the  relative amount of the
 three main material inputs contained in the coating:  pigment,
 binder,  and solvent.  Different formulations,  particularly
 different ratios of pigmentation in  the dried film to total
 volume of the dried film (pigment-volume concentration),  will
 lead to correspondingly different protective and decorative
 functions.41  For example,  a coating  designed to hide surface
 irregularities (like a mastic texture coating)  has a higher
 pigment-volume concentration than a  gloss varnish  whose
 decorative function is to impart a shiny transparent or  semi-
 transparent coating.  Low pigment-volume concentrations  have
 an increased resin content and in general have  high
 durability,  gloss,  and washability.   The  ratio  of  solvent to
 nonvolatile components ("solids")  also characterizes types of
 coatings.h Penetrating  stains have a low solids-to-solvent
 ratio,  and, when the solvent  evaporates, virtually no  film is
 left  behind.
      Figure 1-4  shows  typical  formulations and  average VOC
 contents  for  a  few  architectural  coatings.42  The  coatings in
 Figure  1-4 with  higher solvent content also have higher VOC
 content.   A low  solids-to-solvent  ratio, as with semi-
 transparent stain,  is  associated with high "VOC  content in
 coatings with organic  solvents because VOCs are contained
 almost exclusively  in  the solvent  portion of the coating.  Two
ways  to reduce the  amount of VOCs  released from coatings are
 to increase the  solids-to-solvent  ratio and to substitute
water for  an  organic solvent.
     Nonvolatile components are often referred to as  the "solids"  portion
of the coating, which includes pigments, resins, and other additives,
although resins are not really solid until the film forms and are
considered part of the nonvolatile vehicle, or liquid portion of the
formulation.

                             1-25

-------
            Alkyd
            Resin
           Aliphatic    Titanium
         Hydrocarbons   Dioxide
   Talc
           Nonvolatile        Volatile
           Component       Component    Hiding
          (Film Former)        (Carrier)    Pigment          Extender
                    Additives /            X  Additives
                                                       All Other Inputs
 Figure  1-2.
                         Packaged Coating

 Inputs  generally used in  the  manufacture  of  a
        solventborne  coating.
               Acrylic
                Resin
             Water
                     Titanium
                     Dioxide
Tata
               Nonvolatile       Volatile
               Component      Component   Hiding
              (Film Former)      (Carrier)    Pigment        Extender
                      Additives /         \  Additives
                                                    All Other Inputs
                                         Packaged Coating
Figure 1-3.
Inputs  generally used in  the manufacture  of  a
         waterborne  coating.
                                   1-26

-------
       Semitransparent
            Stain
         Exterior Nonflat
             Paint
                       General Primer or
                         Undercoater
\
Solvent
Pigment — 7
Binder — -, 1
I '
f
b
Average VOC Content
     Organic solvent  527
     Waterborne
85
                                Solvent
                                Pigment
                                Binder

Average VOC Content
Organic solvent 404
Waterborne     76
                                 Solvent
                                                    Pigment
                                                     Binder
Average VOC Content
Organic solvent  374
Waterborne     48
           Figure  1-4.
    Approximate  volume relationships of
     coating ingredients.
Note:      VOC content in grams per liter from Table 1-1.

Source:     Whittington, Trevellyan V.  Paint  Fundamentals.  In Paint
           Handbook.  Guy E. Weismantel (ed.).  New York, McGraw-Hill.
           Pp. 1-1 to 1-23.  1981.  (Adapted  from Figure 1.4)
      1.3.3.3  Manufacturers'  Substitution Options and New
Technologies.  Manufacturers  face two substitution
possibilities to  reduce VOC emissions from coatings.   They may
reformulate  the coating to increase the solids-to-solvent
ratio.   Alternatively,  manufacturers may reformulate  the
coating so that it contains the  same amount of solvent but
emits fewer VOCs  during application (i.e., substitute water
for  an organic solvent).   Certain coatings such as interior
flat wall paint,  interior semigloss,  and exterior house and
trim paint have been formulated  using water for several years.
Between 1950 and  1980,  waterborne coatings replaced approxi-
mately 70 percent of solventborne coatings.43   The performance.
                                1-27

-------
of  latex paints often meets and even exceeds  alkyd counter-
parts;  therefore, manufacturers may choose  to discontinue
organic solventborne paints in these product  classes.
      Other products, including stains, clears,  high-gloss
enamels,  outdoor varnishes, and some special  purpose  coatings,
are more difficult to reformulate.  Currently clear coatings
have  two problems associated with them:  waterbornes  are
transparent to UV radiation,  whereas organic  solventbornes
absorb  UV rays thus protecting the substrate;  and  waterborne
acrylic polymers are not strong enough.44  Quality performance
in  reformulated products is currently possible, but the cost
may be  very high.45  For example,  polyurethane clear floor
finishes have a high organic solvent content,  and  manufac-
turers  have developed waterbornes that maintain excellent
abrasion resistance, but the best products  sell for $50.00  per
gallon.1'47  As new technologies become more refined, new  resin
systems,  such as alkyd systems once used only  in
solventbornes,  will be used in more coatings,  so prices will
become  more competitive.
      High solids content formulation is an alternative
technology to waterborne formulations that manufacturers  have
employed to reduce VOC emissions from coatings.  A high solids
coating is formulated with a  high solids-to-solvent ratio.j
Since a smaller percentage of solvent is contained in the
coating,  fewer VOCs are released during application.  Table
1-7 shows example reduced  solvent contents of three different
types of  reformulated organic solventborne products.48
      Disadvantages of  higher  solids organic solventborne
paints  include increased viscosity,  longer drying  time,
reduced durability,  and generally higher prices.49
Reformulated organic solventbornes are thicker, which makes
     1Flecto's Varathane Diamond Finish, BonaKemi USA's Woodline
Waterborne Urethane, and Pratt & Lambert's  Fabulon Division Crystal Finish
are  three examples of durable waterborne clear finishes.46

     jNote that the definition of high solids varies by coating type.

                              1-28

-------
      TABLE 1-7. PERCENTAGE  OF  SOLVENT IN CONVENTIONAL AND
           REFORMULATED ORGANIC SOLVENTBORNE  COATINGS
 Product
Conventional solvent
    content (%)
Reformulated solvent
    content  (%)
 Interior semigloss
        60
        47
Clear coatings
Stains
55 - 62
72 - 85
35 - 37
30 - 35
 Source:  Bakke, Timothy 0.  Clean Air Paints.  Popular Science.  237;85.
        August 1990.
 them harder to apply and extends drying time,  but they may
 offer greater protection.  Durability may be compromised
 because of the reduced strength of shorter chain alkyd
 molecules substituted for longer chain molecules to improve
 flow.50  Reformulated alkyd products do  offer some advantages
 however.   Durability may be traded for flexibility,  which
 provides  increased resistance to cracking and peeling.
 Presealers may not be necessary for wood substrates  because
 the  thicker coatings penetrate more evenly.   Reduced VOC
 emissions,  lower odor,  and reduced toxicity and flammability
 are  other benefits.
      One  manufacturer,  Glidden (Cleveland),  has developed and
 marketed  a VOC-free paint line.   The company believes that low
 odor in addition to environmental safety will  also be an
 important selling point.   The VOC-free paints  tested well in
 durability,  hiding power,  and ease of  application tests.   One
 drawback  is they currently are only available  in a few colors,
mainly  off-whites and dull greens and  grays.51
      Low  and VOC-free coatings have also been developed for
 the  light-duty applications in industrial maintenance.   Porter
Paints  (Louisville,  KY) has introduced a waterborne  mainten-

                              1-29

-------
 ance paint for use directly over solventbornes,  thus avoiding
 the use of VOCs to strip the old paint.   In the industrial
 maintenance area Glidden is researching  resins to develop new
 ones compatible with water without  using additives.52
      Raw material suppliers are expanding and improving upon
 existing technologies to meet demand  for performance in new
 waterborne and high solids formulations.   Solvents for use in
 waterborne formulations (i.e.,  glycol  ethers)  and high solids
 (keytones,  esters)  are replacing many  of the  hydrocarbon
 solvents used in solventborne formulations.   Resins are being
 developed with a goal toward improved  performance in new low-
 VOC formulations;  similarly additives  are being  developed to
 improve flow and leveling characteristics of  the new resins.
 Hulls America has  developed a line  of  colorants  that is
 virtually VOC free by removing surfactants  and glycols.53
 Additional  low-VOC technologies  are reactive  diluent
 technology,  radiation curing  technologies,  and powder
 coatings,  which currently are mainly used in manufacturing
 applications.
 1.3.4   Industry Conditions
     1.3.4.1   Shipments  and Manufacturer  Specialization.   In
 1991, the architectural  coatings segment  of the  paint  and
 allied  products industry shipped $4,881.9 million  in poten-
 tially  regulated products  (Table 1-854"63) .  The value of
 shipments has  steadily increased by approximately  59 percent
 since 1981, with a  slight decrease between 1990  and  1991.  The
 strong  construction market  throughout the 1980s helped
 contribute  to  this  growth, but the industry as a whole is
 generally considered  to  be maturing.64  In 1991,  the  market
 share for architectural  coatings in relation to  the  total
 coatings  industry was  37.8 percent.   New products are
 important to the paint and allied products industry, because
 growth  for  individual producers is predicted to come from
market  share expansion, new product introductions, and
 improvements in  established products.65
                             1-30

-------


CO
B
Q
o
CH
Cu
Q
W
M


fCQ
EH

CU
Q vc~~
W 0
EH £
OrH
cc cn
r-l
J 0
M 0
12 £G

0 £

O
Q
W
Cu
CU
M
K
W
U
3
<
>
co
r-l
W
. T
CQ











n

















.
1'
C

<
OJ
CO
m
en

c
en

in
CD

a
en

r-
CD


CO

en
CO
en


en

en

w
fl
U
4J
Produ
j
i *
2-0
3 °
2 u
*£ vo p» in vo •-* 01
in en en m o r- m
O V£) r- TT 0 rH
cn oj co p- r-i TJ- *r
O ^O P" ^" rH
CD ^j« co CN r- vo en
o 01 m o o r- m
m ui CD v 01

in r- co in ri
m *H
CD CD «H P- in OJ *T
o c- »H TT m cn 01
co r- co in v rH
O m co in en oj o»
o in en m rH I-H co
rH \o cn m cn co rH
o p* co in ^ oj
^« en r* o rn r* en
in vo o) m o rH o
^ «-H m o oj co p-
»
co p* m ^T oj o* vo
OJ rH rH OJ r- CO O
 rH rH
—
rH ff^ O ^* rH CO P-

cn co o o* co m r-
rH vD vo r- '(J1 rn n
CD in \o «^ in co p*
CO CO r-t VO O
01 CO
O> 0) 0) U
C C C 0)
-H H Of i^ OJ 3 a
1C £1 l-i -Q U U ^
O *J O AJ O <0
U C XJ C X) -H W
O)OlO)OttDO)O>Ot U O  D .n ^j^
5 £ JS 	 ur^-H •
^ ^ ,H miouiMBiramwwoiococotii
OOOOOOOOOOM 2.
aiciaiaiaiaioDaiuu-cc
010 itKattactctctixctiTjneo

^^^ IOIDIOIOIDIOIOIOIOIDJ-1-
C C.C.C.C.C.C.C.C.C. UlN
n n fN ui i-i vc c
O^Of*l . . . . . . , . . .CIOOO
y. c C^Oi^O^Ol^OlOojOptO^OoiOmOCT^k-i^^co

O OOQ) -*^ ^0)Dft)ll)HftlWftlft)C:WOt"'O

« 8 •§ 8 1 « wto ^'JwO«°w°W°w°M°W°ui°ui°W=.2?«)0
W M t-!02S3S333333=lOOlOiJ
,-. C O. n:
_. u uj 01 W •-<
o .* O C — 3 «
2jt"i -.-oat-
M n n . «
~* ~ -H " in " "
l/l l/l L/l . U V
CO CO CO ^ C ^ ^"
01 M M 2. 5 2
1-31

-------
      Sales  in  the  architectural  sector generally reflect
activity in house  redecoration,  maintenance and repair, as
well  as  sales  of existing homes, new home building, and,  to  a
lesser extent,  commercial and  industrial construction.  Among
interior and exterior  architectural coatings, the waterborne
coatings market dominates the  sector and experienced a  larger
percentage  increase  in growth  during 1981 than did organic
solventborne coatings.   Interior waterbornes grew the most,
88.4  percent from  1981 through 1991.  Seventy-six percent of
interior coatings  and  57.6 percent of exterior coatings are
waterborne.  In response to environmental regulations aimed  at
the reduction  of VOC emissions, the industry has shifted from
manufacturing  conventional organic solventborne paints  in
favor of paints with high solids-to-solvent ratios and
waterborne  and solventless paints.66
      Although  the  historical Census data do not identify value
of shipments for paint products within the four product
classes,  other sources indicated that the majority of interior
wall  and exterior  siding paint jobs use waterborne prod-
ucts.67"69  Therefore, the exterior  and  interior  solventborne
shares probably account for mainly coatings used on exterior
and interior trim,  floors,  decks, and high-gloss enamels.
      Industrial new construction and maintenance paints and
traffic  marking paints are classified by the Census as special
purpose  coatings,   which comprised 22 percent of  the total
coatings market in 1991.   Market shares for industrial
maintenance  and traffic marking paints  within the  special
purpose  segment were 28 percent and 4.6 percent,  respectively.
Growth prospects for this segment are expected to  be above
average, especially for industrial  and  machinery maintenance
coatings.
      For all companies classified in the paints  and allied
products industry  in 1987,  98 percent of their value of
shipments was generated from the manufacture of  paints and
                             1-32

-------
 allied products  (Table 1-9) .k>7°  Only 3 percent of  the value
 of paints and  allied products shipped were  manufactured by
 companies outside  the industry.  The top three secondary
 producers of paint and allied products account for  about half
 the value produced as secondary products in other industries
 and are shown  in Table 1-10:   adhesives and sealants,  plastics
 materials and  resins,  and printing  ink.69   Because coating
 products often function as  sealants,  the adhesives and
 sealants industry  is a logical secondary producing  industry.
  TABLE  1-9.   NUMBER OF COMPANIES,  ESTABLISHMENTS,  AND PRODUCER
         SPECIALIZATION--PAINT AND ALLIED PRODUCTS:  1987

SIC
code
2851


28511

28513



Industry/
primary Number of
product class companies
Paints and 1,123
allied
products
Architectural
coatings
Special
purpose
coatings

Industry
Primary
Number of product Coverage
establish- specialization ratio
ments ratio (%)" (%)b
1,426 98 97


282

131

Product
class
Total made
in all
industries
($106)
12,078.8






* Value of primary products for the industry divided by the sum of the value of
  primary products produced by the industry and the value of secondary products
  produced by the industry.

b Value of primary products for the industry divided by the total value of products
  for that industry produced in any industry.

Source: U.S. Department of Commerce.   1987 Census of Manufactures,  Industry Series:
       Paint and Allied Products.  Washington, DC, U.S. Government Printing
       Office.  1990.
       Industry statistics,  unless otherwise noted,  include figures for all
segments  of the paint and allied products industry,  not  just those to be
regulated.

                                1-33

-------
  TABLE 1-10.
COMMODITY  PRODUCTION IN 1982:  PAINT AND ALLIED
             PRODUCTS
        (SIC 2851}
    Product examples
        Producing industries
Value3
($106)
Percentage
 produced
 Interior and exterior
 paint,  lacquers, and
 varnishes;  OEM
 coatings;  industrial
 new construction and
 maintenance paints,
 traffic paints,
 automotive  refinish
 paints,  marine paints,
 aerosol coatings,
 paint and varnish
 removers,  thinners,
 putty and glazing
 compounds,  brush
 cleaners
        Primary               8,243.3

        All secondary
         producers               303 .2

        All producers          8,546.5
        Top  three secondary:      142.6
         Adhesives and
          sealants                68.8
         Plastics materials
          and resins              45.8
         Printing ink             28.0


        All  other secondary      160.6
        producers	
             96.5
             1.7

             0.8

             0.5
             0.3


             1.9
* Measured at producers' prices.

Source: U.S. Department of Commerce.  The 1982 Benchmark Input-Output
       Accounts of the United  States. Washington, DC, U.S. Government
       Printing Office.  1991.
      1.3.4.2  Company Size  and Industry Structure.   In 1987,
the paint and allied products  industry comprised 1,121
companies owning a total of 1,428 establishments (Table
l-ll72'73).   Single establishments were held by approximately
77 percent of the companies, and they had an average value
added of  $1.1 million.  The multiestablishment companies had
an average value added of $20.4  million and produced almost 85
percent of the total value  added for the industry.   Also shown
in Table  1-11,  the 50 largest  companies in 1987  produced 66
percent of the total value  of  shipments for the  industry.
Data  from the Small Business Administration (SBA) indicate
that  in 1991 there were 1,152  companies and approximately 98
percent of those were classified as  small businesses as
defined by having fewer than 500 employees.74   Figure 1-5
                               1-34

-------







H
s
W M
*C* ±J
6 C
HM flj
w I
H a
iJ '2
*^ -C
CQ || w
rij II
E"
CO |
W ||
g£
v3 Oi
**• rH
MH
O
a
it
^


cn
w ..
2 X
iERS OF COMPA*
DUCTS INDUSTR
H O
5 *
2
I8
^ H
<« t-3
w 3
u *
% e
r*7 *M
3 2
I-H ^
S c_





0

>,
JO

V4
O
UH
•0
01
JJ
c
o
u
o
<0
0>
Oi
ID
JJ
C
01
u
}H
01
a.

*-" in
i"
?s -
« ex
J f
..
s °


m W
m W
o>e
si
o
° 8
CM U

jj n
IA 0)
0) -rl
rn c
M*
J&
0
CO U

JJ 01
n o>
01 -H
Oi C
rl 10
rd a
U E
O
•» u
1^~
5s
Hi2
0) 0 „
Di «T3«,
T3 3 4) o
ii *i "° ^
J J T3  > 10 —
4.

01 TJ^~
3 01 o
SIS


«i
0 z 1 ° 2
Q M *2
s •< 3-83
9 PH E JJ C
OS 3 n 01
H g
W ^
I5
rH
rH
1
rH
W
•J
CQ
•<
EH

ndustry Number of
CTI- 9B«;i » comoanies
|| M ^«







n
in





o
<•





r~
CM
^< o\
CM rH
o in
r- vo
CM" o
rH rH
m f
m o
CM
r- <->
0 rH
CM r~
CM CM
\o in



CO ^0
CM *&
•9 in
rH
aint and allied 1,121
product producers
ultiestablishment 259
- 0. S



o



















m
o
m
o
CM
-1
rH
m
a\
«*
o\



CM
«5
CO
companies
ingle-establishment 862
nomnanies
m

i-i
O

o
OJ
CO

M
o
'
CQ
0)
•rH
(0

g
O
u

JJ
en
01
>f the four largi
: the industry.
Lpments c
pments oi
C -H
^- "^
W J3
CQ
U-l
O «M
O
0)
3 a)
r— t 3
ro r-l
> <0
U-l
IO r-l
*
g JJ
3 O
W JJ
Percentages consist of the
companies) , divided by the
1/2























(Q
0>
. .J
f compan:
o
M
0)
I
3
C
i
JQ
T3
Q)
T3
•rH
>
•H
TJ
-d

co e
c c
01 rH
U 0)
c- o
oo O

-------


TJ
C

C
-H
(C
a
0)
f-t
•£*
4-1
C
•H
W
4->
C
0)
1
in
• H
r-t
XI
(0
JJ
w
0)
Dl
c
-H
w
3
U
U
(0
I4_l
3
C
(C
e

4-1
O

c
o
-H
U
(0
U
0
J












•
1-^
in
00
/•o
\ N
U
M
w
••
r~
00
1
M
4J
I/I
3
73
C
•H

U]
JJ
U
3
*O
O
V^
a

-o
0)
•H
1— 1
l-(
(0


a>
w
(0
01
i— i
a>
cc
,
w
0
-H
U
01
W
4J
U
O
a
01
a
*,
T-l

(D
3
r^
O
>

.
uses
to
c
0)
0

U
-H
E
o
c
o
U
U
r-
oo
o
rH


0)
U
k-l
1)

s
0
U




i-H

&\
i-i


O
±
U
r-
OD
O
S
0>
1— 1
-H
M-l
[/)
JJ
c

*
U
o
§ ,
M
0)
>J
                                            C  3
                                            0)  C
                                     3
                                     D)
                                            (0 >w
                                            a  o
                                            0)
                                               W
                                               C
                                               0)
                                               U
                                            01
                                            U
                                            M

                                            o
                                            t/1
1-36

-------
 displays the location of manufacturing establishments in the
 paint and allied products industry by state.75  California has
 the greatest number,  201,  followed by Illinois  with 118.
 Paint manufacturing is fairly well represented  in most states
 east of the Mississippi River.
      Over the past decade,  consolidation has been a major
 trend in the paint and allied products industry.   The maturity
 of  the industry and increased technology requirements are
 factors contributing to the restructuring.   A large number  of
 mergers and acquisitions took place in response to pressure
 from the higher cost of paint ingredients,  intense industry
 competition,  compliance with government regulations,  and low
 profit margins.76  Other companies are divesting their paint
 and coating operations to  focus  on other businesses or as an
 alternative to making the  capital  and research  and development
 (R&D)  commitments required to remain competitive.   The number
 of  coating manufacturers and the number of  establishments
 operated by these manufacturers  have decreased  over the past
 two decades.  As indicated  in Table 1-12,  from 1972  to 1991,
 the number of companies decreased  by 12  percent, and the
 number of manufacturing establishments  decreased by over  20
 percent.77
      On average,  35  to 40 mergers  or acquisitions have taken
 place each year in the coatings  industry for  the past few
 years.78  A transaction involves  the transfer of production
 capacity from one  company to  another but  does not necessarily
 indicate the  dissolution of  the  company making  the  transfer.
 The  selling company could sell only  a division  or product line
 and  remain in business.  Some of the large recent acquisitions
 reported in trade  journals, by the press, and in companies'
 annual  reports  are listed in Table 1-13.79
      Most of  the  larger  companies produce architectural, OEM,
 and  special purpose coatings.  Several of the largest  coatings
producers  are chemical corporations;  however, paint manufac-
 turing  represents  only a small part  of their overall busi-
ness.80  In 1991,  merger activity slowed down and  left the

                              1-37

-------
   TABLE 1-12.   NUMBER  OF COMPANIES AND  ESTABLISHMENTS  IN THE
           COATINGS INDUSTRY, SELECTED YEARS,  1972-1991
Year
1972
1977
1982
1987
1991
% change 1972-1991
Number of
establishments
1,599
1,579
1,441
1,426
l,400a
-12.4%
Number of companies
1,317
1,288
1,170
1,123
1, 030a
-21.8%
* 1991 figures are from Finishers'  Management.  The U.S. Paint and Coatings
  Industry.  pp. 23-25.  April 1991.

Source: U.S. Department of Commerce.   1987  Census of Manufactures,
        Industry Series: Paints and Allied  Products.  Washington, DC,  U.S.
        Government Printing Office.  1990.
   TABLE 1-13.   RECENT  ACQUISITIONS IN  THE COATINGS INDUSTRY
 Selling company    Acquiring company
                       Division sold
 DeSoto

 Whittaker Corp.

 Azko Coatings
   Inc.

 DeSoto
 Clorox Co.
Sherwin Williams

Morton International

Reliance Universal
  Inc.
Valspar
PPG Industries, Inc.
Consumer Paint  Operation

Specialty Chemicals
  Operation

Buyout
Coil Coatings Operation

Olympic and Lucite
  finishes
Source:  Loesel, Andrew.  Coatings Industry Faces New Mix.  In Chemical
        Marketing Reporter.  238(18):SR3-SRB. New York, Schnell Publishing
        Co.  1990.
                                 1-38

-------
industry basically divided into two groups:  a few, well-
financed and highly diversified multinationals and a large
number of regional paint companies.81
                             1-39

-------
                           SECTION 2
        COST AND ECONOMIC  IMPACTS  OF  PROPOSED  REGULATION
     This section assesses the costs that must be incurred  to
comply with  the proposed AIM coatings regulation and examines
the economic  impacts of these costs as they are absorbed by
producers and consumers of the regulated products through
market processes.  Although the effects of this regulation  may
be felt outside of the markets in which the regulated products
are traded  (e.g., in markets for coating substitutes and
coating input markets  [including labor markets]), the analysis
focuses entirely on the AIM coatings product markets
themselves.

2.1  BACKGROUND

     The EPA  plans to control VOC emissions from AIM coatings
using a combined regulatory approach:   (1)  product category-
specific VOC  content limits and (2)  an option for producers of
products over the proposed content limits to pay a fee on the
VOC content in excess of the limit.   Using reformulation cost
estimates and a proposed fee rate,  the potential impacts of
the proposed  regulation are analyzed,  first in static analyses
of the reformulation,  fee,  and withdrawal options and second
with a dynamic market analysis that  estimates changes in
prices and quantities and welfare costs.
                              2-1

-------
2.2  OVERVIEW OF RESPONSE OPTIONS

     The regulation to reduce the VOC content of AIM coatings
will affect both production decisions for the suppliers of  the
coatings and consumption decisions on the demand side.  Before
developing a formal economic model to analyze these
regulations, the scope of responses available to producers  and
consumers are briefly characterized.
2.2.1  Supply
     The EPA is proposing a set of limits for the VOC content
in specific product categories to be met by 1997.  Firms that
produce products exceeding the proposed VOC limits essentially
have three potential options:   reformulate the products so
that they comply with the standard,  potentially paying a fee
on the excess VOC content over the standard,  or remove the
product from the market.
     The firm will presumably choose the option that maximizes
net benefits,  as measured by the expected (discounted)  value
of the profits generated under each option,  less the
(discounted) cost of each option.   However,  in the short run,
firms may be unable to reformulate their coatings to meet the
content limits requirement because of the timing of the
regulation or constraints on the resources that can be
allocated to reformulation.   They may remove the product from
the market temporarily until the VOC standard can be achieved
or potential pay the exceedance fee.  If technological and
market conditions allow,  they may increase their production of
within-limit products to compensate for the reduction in over-
limit products until the firm makes long-run adjustments to
adapt to the product-line restrictions.   This strategy may be
temporary until the over-limit product(s)  can be successfully
reformulated,  or it may reflect a permanent response to the
regulation.
     The compliance strategy decision is likely to be
complicated by issues other than cost that relate to the
                              2-2

-------
 profitability of reformulation.  If a product  serves a narrow
 market niche,  reformulation may fundamentally  alter the
 product's attributes and erode the niche position.   In such a
 case,  the producer may not find choosing reformulation
 profitable.   Although concerns regarding the regulation's
 constraints  on product differentiability are undoubtedly real
 in  some cases,  this complexity is not explicitly  addressed in
 the quantitative analysis that follows, primarily because of
 the difficulty in observing both levels of and changes in
 product differentiation.   Moreover,  the proposed  limits have
 been determined technologically feasible,  subject to Best
 Available Controls, which should imply that the reformulation
 will not fundamentally alter the functions that the product
 can perform.
      Often,  product reformulation involves an  investment in
 research and development  (R&D)  to develop a compliant  product.
 The extent of  the reformulation necessary to bring  a product
 into compliance can vary  from product to product.   In  some
 cases,  compliance can be  achieved for a particular  product
 without large  R&D investments because the product is similar
 enough to an existing formula or other product undergoing
 reformulation.   A major reformulation,  as  discussed here,
 typically requires a significant resource  and  time  commitment.
 The process  can take several years  and is  divided into  a
 number of different stages.   Figure  2-1 identifies  the  basic
 reformulation  stages for  a  prototype AIM paint (other  coatings
 such as varnishes may have  fewer stages) .82  The firm may need
 to  alter its capital equipment  to produce  the reformulated
 product,  but these physical  capital  adjustments are  usually
 small  compared  to developing the intellectual capital  to
 devise the new  formula.3
    aOne member of the Regulatory Negotiation Committee raised the point
that substituting  away from solvents  in the new formulations introduces  the
problem of  disposing of old solvent storage containers.  Another member
mentioned disposing of obsolete labels as a potential problem as well.83
                              2-3

-------
                             Formulate a White Paint
                        Conduct Field Testing and Color Formulation
                              Formulate the Bases
                              (that will be tinted)
                           Determine Prescription for Retail
                           	Stores	
                              Develop Information
                          (technical and safety sheets, marketing
                         	information, labels)	
                        Scale Up Production. Introduce, and Distribute
Source:
                 Estimated Effort     2 '° 3 person-years
                 Estimated Elapsed Time 1 5 to 5 5 years

Figure 2-1.  Basic stages of AIM coating reformulation
             (prototype  firm and  product).

    AIM Coatings  Regulatory Negotiation Committee meeting,  July 28-
    30, 1993, Washington,  DC.   Meeting Summary.
2.2.2   Demand
     The regulations can  be expected to induce changes in the
prices of the affected products.   A consumer may alter his/her
selection of  coatings based on the relative  prices of  coating
products and  on the relative prices of coating versus
noncoating alternatives.   For example, consumers may opt for  a
waterborne coating rather than its solventborne alternative if
the  regulation-induced change in prices increases the  relative
price  of the  solventborne product.   Moreover,  a potential user
of a high-VOC coating product facing reformulation may even
opt  for a noncoating alternative if the price rises too much.
     The reformulated products may also possess different
characteristics that affect their  demand.  For instance,  VOC
content reduction in a typically high-VOC product may  change   .
                                 2-4

-------
consumers'  perceptions  of the product's performance,  dura-
bility,  and ease  of  application.   The lower VOC content may
also work as a  signaling  device for the "green" consumer in
pursuit  of products  deemed more friendly to the environment.13
These  factors collectively affect  the benefit consumers derive
from using the  product  and thus affect their willingness to
pay for  the reformulated  product versus other product
alternatives.

2.3  COST ANALYSIS

     The regulations will cause firms who make products that
exceed the allowable VOC  content specified in the Table of
Standards (TOS) to either reformulate,  pull the product from
the market or pay an exceedance fee.   This section evaluates
the costs imposed on manufacturers  to reformulate non-
compliant products and  the option  to  pay an exceedance  fee.
Section  2.4  incorporates  the  option of withdrawing products
from the market into the  decision process to evaluate national
market efforts.
2.3.1  Reformulation Costs
     The estimated national cost for  the regulation is  based
on information  developed  by industry  representatives  during
the regulatory  negotiation.   The assumption in estimating
these costs  was that coating  technologies would need  to be
researched and  developed  in the laboratories of resin
manufacturers/suppliers and paint manufacturers in order to
meet VOC  requirements.  Although the  proposal is significantly
less stringent  than the three-phases  of  requirements  discussed
during negotiations,  the  EPA  has relied on these same
reformulation cost estimates  for calculating the national cost
of the proposed rule.  Given  that the rule has similar  VOC
content  requirements to State rules which have been enforced
    fcSome manufacturers currently produce zero-VOC-content coatings that
are marketed as "clean  air"  coatings.
                              2-5

-------
since 1990, the EPA believes  the reformulation estimates used
may be overstated.  Since the proposed rule is implementing
available resin technologies, the cost to comply for those
manufacturers who have or will reformulate their higher VOC
coatings is expected to be partially reduced through the
assistance of resin manufacturers/suppliers.   Upon request,
most resin suppliers are willing to share information and
sample low VOC coating formulations with interested paint
manufacturers, both large and small.  In addition,  another
limitation in the cost data is that no distinction for
reformulation cost is made between categories (i.e.,  the
reformulation cost in one category is the same as the
reformulation cost in any other category),  or in relation to
the required VOC content reduction (i.e., it  does not
distinguish between coatings at different VOC levels above the
limit).   In the preamble for this proposed rule,  the EPA
requests comment and technical information on previous (since
1990)  or potential reformulation costs specific to each
category and VOC content level change.  In addition,
information is requested on any changes in variable (e.g.,  raw
material) costs or disposal costs associated  with
manufacturing coatings to meet the proposed VOC levels.
     2.3.1.1  Product-Level Reformulation Cost Estimates.
Whether reformulation is a feasible response  for the proposed
VOC content limit regulations is an empirical issue that
varies by firm and product-specific factors as well as by
timing specifications of the regulation.   Since there is
insufficient data for this study to observe these firm and
product-specific factors,  the analysis that follows discusses
costs involved for a representative firm facing a typical
reformulation in response to the regulations.
     The new formula represents a different mix of  the four
coating components:   resins,  solvents, pigments,  and addi-
tives.   For solventborne products,  a new formula may involve
increasing the ratio of solids (resins)  to solvents to reduce
                              2-6

-------
 the  solvent's  contribution  to VOC  emissions.   As  explained
 previously,  a  reformulation may be "major,"  involving
 significant  investment  in R&D or it may  require relatively
 little  investment  cost.  The discussion  of reformulation  costs
 that follows refers  to  major reformulations.
      A  presentation  to  the  AIM Regulatory Negotiation
 Committee  indicated  the level of effort  required  for develop-
 ing  a new  coating  product.  Costs  were estimated  based  on this
 level of effort.84   These cost estimates  are  for developing a
 new  coating  product, which  may be  more difficult  and more
 expensive  than reformulating an existing product.  However,
 these data provide the  only available estimates for reformula-
 tion costs in  this study.   Some of the upward  bias implied by
 using new  product  development costs as a proxy for
 reformulation  costs  may be  offset  by downward  biases resulting
 from omission  of certain non-R&D costs in the  reformulation
 effort  (e.g.,  test materials,  non-R&D labor).  Raw materials
 costs also may increase  because reformulated products often
 have a  higher  solids-to-solvent ratio, and higher solid resins
 are  more expensive than  solvents.85 The various uncertainties
 and  potential  biases surrounding the reformulation costs  esti-
 mates are  presented  in Table 2-1.  Since these  biases cannot
 generally  be quantified, the costs derived in  this section  are
 used as a  qualified  best estimate of reformulation costs.
      The levels of effort for developing a new product
 indicated  by the Regulatory Negotiation Committee presentation
 are
      •  2  to 3 scientist years and
      •  1.5  to 5.5 years elapsed time.
Given the  above time ranges, a midpoint estimate of the level
of effort  is 2.5 scientist years over a 3-year period.   Esti-
mating  the cost of a scientist year at $100,000 gives the
total cost of  a major reformulation effort as  $250,000,
assuming these costs are incurred evenly over  the 3-year
                              2-7

-------
     TABLE 2-1   REFORMULATION COST ESTIMATION UNCERTAINTIES
   Assumptions

      •  Reformulation effort:  2.5 scientist years over 3-year
         period
         Cost per scientist-year (1993 $)  = $100,000
      •  Discount rate:  7 percent
      •  Initial lump-sum cost annualized on perpetual "capital
         rental" basis (discount rate •  reformulation  cost)

   Potential upward  bias  factors

      •  Cost estimate is for new product;  reformulated product
         may be lower.
      •  Cost estimate is for "major" reformulation; minor
         adjustments may be feasible.
         Major reformulation effort is based on expectations of
         the originally proposed three-phase Table of  Standards,
         with each phase progressively more stringent.  The
         reformulation effort may be less  intensive under the
         current single phase proposal.
      •  Costs may fall over time as new technology is developed,
         disseminated.
      •  Costs may be partly borne by material suppliers.

   Potential downward bias  factors

      •  Perpetual capital rental basis  for cost annualization
         assumes reformulation benefits  accrue forever.   Finite-
         lived benefits would produce higher annualized costs.
         Ignores nonresearch costs of reformulation (e.g.,
         materials).
      •  Multiple products may be lumped together as one in the
         survey.  Therefore,  multiple reformulations may be
         necessary in some cases where a single reformulation is
         proj ected.

   Potential factors  with unknown bias

      •  Estimate is for a white paint;  other products may
         differ.
      •  Reformulation may positively or negatively affect
         variable production costs (e.g., materials).
         Effects on product quality (i.e.,  "cost"  of reduced
         quality).
         Costs may rise/fall based on amount of "excess  VOC"  to
         reduce.                        	
period  ($83,333 per  year).  Since  the effects  of  the

regulation are annual  reductions in VOC emissions,  placing  the

corresponding costs  on an annualized basis is  appropriate.
                                2-8

-------
 First,  the  cost  of  the  initial  reformulation  is  valued in  the
 same year  in  which  the  benefit  (emissions reduction)  stream
 begins  (i.e.,  at the  end of  the  3-year reformulation  period).
 Thus the future  value of the 3-year expenditure  is  computed at
 the end of  the third  year, when  the reformulation is  complete
 and the benefits of the reformulation commence.  Here and
 throughout  the analysis a 7  percent discount  rate is  used.
 Valuing each  annual expenditure  at the end of the year gives
 the future  value of

  FV =  83,333  •  (1.07)2  + 83,333  •  (1.07) + 83,333
     =  $267,908   .                                        (2-1)

 To compute  the annualized value, we took the product  of the
 discount rate  and the initial cost

       A =   r  • FV =  (.07) (267,907) = $18,754 per year  .  (2-2)

 This annualization formula is based on the notion that  the
 reformulation  produces  a perpetual stream of emissions
 reduction benefits.  Thus the annualized cost of the
 investment  can be viewed as  the  foregone perpetual annual
 yield of the  funds applied to alternative uses.
     In the cost  and market  analyses that follow, all  dollar
 values  are  expressed in their 1991 equivalents,  to match the
 price data  that were obtained for the market analysis.  The
 annualized  cost  figure  in Equation (2-2)  is based on  a 1993
 cost estimate.  This value is deflated to 1991 dollars using
 the GDP price  deflator  and derive an annualized cost  of
 $17,772 per reformulation.
     As a point of comparison, estimates of the cost  of AIM
 coatings reformulation are provided in a study conducted for
 the South Coast Air Quality Management District (SCAQMD) to
address economic  impacts of VOC content regulations in
California.86   This study identified costs associated  with
                              2-9

-------
product reformulation and temporary and permanent product
sales losses.  Reformulation costs varied depending on the
extent of the reformulation necessary.  Most of the small
firms surveyed indicated that they did not have full-time R&D
employees.  Costs for additional research and development due
to the regulation ranged from $1,000 to $5,000 annually for
firms with few products affected by Rule 1113 and more than
$50,000 for firms with many affected products and little or no
research staff.
     The SCAQMD study also indicated other compliance costs
not related to R&D.  Rough estimates of the cost of equipment
adjustments necessary to accommodate reformulation ranged from
$5,000 to $35,000 per firm.   Costs attributed to temporarily
or permanently discontinued products ranged from zero to
$3,000 for firms with few affected products to more than
$75,000 for firms with many affected products.  Employment
changes for the surveyed firms in the SCAQMD study were
expected to be minimal,  affecting only the possible addition
of R&D chemists.
     Because the timing,  number of reformulated products,  cost
components,  and regulatory structure associated with each
SCAQMD cost estimate is not  apparent from the report,  they
cannot be combined with the  Regulatory Negotiation Committee
estimates given above in any meaningful fashion to provide a
better estimate of regulatory costs.  Therefore,  the estimates
provided by the Committee are used in the analysis performed
here as the best estimate of reformulation costs,  recognizing
the limitations discussed above and outlined in Table 2-1.   To
evaluate the sensitivity of  the resulting economic impacts of
the regulations, the analysis that follows also evaluates
alternative reformulation and exceedance fee values.
     2.3.1.2  Total Reformulation Costs and VOC Reductions.
At the industry level,  some  producers will respond to the
regulations by reformulating products that exceed the VOC
content limits,  some will pay an exceedance fee,  and others
                             2-10

-------
will  remove  noncompliant products  from  the market.   The
analysis  begins  with  the conservative assumption  that  only  the
first option,  reformulation,  is allowed.  Then, the  cost  of
reformulation  is recomputed  assuming all products over the  TOS
limits will  reformulate.   In this  section, aggregate
reformulation  costs are for  the products reported in the
Architectural  and Industrial Maintenance Surface Coatings VOC
Emissions Inventory Survey (the survey).87  The  survey
population represents roughly three-fourths of  total industry
output.   In  Section 2.4, the analysis is extended to the
national  industry level to calculate market effects  of the
regulations.
      To estimate reformulation costs for the entire  survey
population,  a  determination  of the number of AIM products that
will  need reformulation to comply with  the standards is made.
This  number  depends on the number of AIM products with a VOC
content exceeding the proposed standards for the respective
product categories.  Table 2-2 provides the proposed TOS.
      Next, the number of products in the survey that exceed
the limits imposed by the  TOS is determined.   The survey
reports the  number of products,  sales volume,  and average VOC
content for  specific VOC content ranges (e.g.,  0 to  50 g/L,  51
to 100, 101  to 150) within specific product groups (e.g.,
exterior  flat waterborne,   exterior flat solventborne,  interior
flat waterborne).  Knowing the limits imposed by the TOS, the
number, volume,  and average VOC content of products over the
limit can be derived using the survey data.   These data can be
used  to generate  estimates of the expected cost of
reformulating products subject to the TOS,  as  well as the
associated reduction in emissions accomplished by the
reformulations.   Using these cost and emission reduction
estimates, the cost-effectiveness per metric  ton (Mg) of VOC
reductions is analyzed.
                             2-11

-------
                TABLE 2-2.   TABLE OF  STANDARDS
                                              VOC content limit
	AIM  coating	(g/L)	
 Antenna coatings                                     500
 Antifouling coatings                                450
 Antigraffiti coatings                                600
 Bituminous coatings  and mastics                      500
 Bond breakers                                       600
 Chalkboard resurfacers                               450
 Concrete curing compounds                            350
 Concrete protective  coatings                         400
 Dry fog coatings                                     400
 Extreme high durability coatings                     800
 Fire-retardant/resistive coatings
   Clear                                             850
   Opaque                                            450
 Flat coatings,  N.O.S.
   Exterior                                          250
   Interior                                          250
 Floor coatings                                       400
 Flow coatings                                       650
 Form release compounds                               450
 Graphic arts coatings  (sign paints)                  500
 Heat reactive coatings                               420
 High-temperature  coatings                            650
 Impacted immersion coatings                          780
 Industrial maintenance  coatings                      450
 Lacquers (including  lacquer sanding                  680
 sealers)
 Magnesite cement  coatings                            600
 Mastic texture  coatings                             300
 Metallic pigmented coatings                          500
 Multicolor coatings	580	
                                                      (continued)
                              2-12

-------
          TABLE  2-2.   TABLE OF  STANDARDS  (CONTINUED)

                                               VOC content limit
	AIM coating	(g/L)	
 Nonferrous  ornamental metal lacquers                 870
 Nonflat  coatings, N.O.S.
    Exterior                                           380
    Interior                                           380
 Nuclear  power plant coatings                         450
 Pretreatment wash primers                            780
 Primers  and undercoaters,  N.O.S.                     350
 Quick dry coatings
    Enamels                                            450
    Primers,  sealers, and undercoaters                 450
 Repair and  maintenance thermoplastic
 coatings                                             650
 Roof coatings                                        250
 Rust preventive coatings                             400
 Sanding  sealers                                      550
 Sealers                                              400
 Shellacs
    Clear                                              650
    Opaque                                             550
 Stains
    Opaque                                             350
    Clear  and semitransparent                          550
    Waterborne low solids                              120
 Swimming pool coatings                               600
 Thermoplastic rubber coatings  and mastics            550
 Traffic marking paints                               150
 Varnishes                                            450
 Waterproofing sealers and  treatments
    Clear                                              600
    Opaque                                             400
 Wood preservatives
    Below  ground                                       550
    Clear  and semitransparent                          550
    Opaque	                         300
N.O.S. = Not otherwise specified.
                              2-13

-------
     Table  2-3  presents  the results of the analysis for the
TOS.88  Table 2-3 reports  reformulation costs and emissions
reduction summed across  all survey products.   A breakdown of
costs  and emissions  reduction by product category is provided
in Appendix C.   Aggregate  cost estimates for the survey
population  were derived  by multiplying the number of products
in the survey facing reformulation by the annualized cost of
reformulation.   As Table 2-3  indicates,  1,729 products from
the survey  exceed the 1997 limits,  which is 36 percent of the
total  number of products in the survey (4,846).c  The  same
presentation to the  Regulatory Negotiation Committee that
serves as the basis  for  reformulation cost estimates also
indicated that  roughly one in three products  that exceed the
proposed limits would not  need a major reformulation,
primarily because the product lines are similar to  others that
will be reformulated.  Thus,  the costs are assessed for the
remaining two-thirds of  products over the limit to  computing
the aggregate cost estimate.   After reducing  the number of
products, the estimated  number of reformulations is 1,153,
yielding an aggregate cost of reformulation of $20.5 million
dollars (1991 dollars) for the survey population.
     To compute the  average cost per Mg of emissions
reduction,  an estimate of  the expected emissions reduction for
the survey  population is needed.   Total baseline emissions for
the survey  population is 344,059 Mg.   The 1997  TOS  is
projected to reduce  aggregate VOC emissions by approximately
20 percent.89 Applying these  percentages to the baseline
emissions figure  for the survey population yields estimated
reduction emissions  of 68,812 Mg.
    cThe actual survey total number of products is 4,920.  However,
throughout  Section 2 we use 4,846 as the  total number (and the
corresponding quantity and emissions) because product-level data were
unavailable for 74 products in the survey.
                              2-14

-------








s
o
M
£-1

J
ID
(J-i
O
CU
>H
W
>
&
D
CO
-— '

CO
^
£-*
o
M
EH
U
B
w
CH
CJ
o
>
Q
S
rtj

CO
EH
CO
O
U

§
M
EH
<
j

h-j
2
CM
0
Cn
W
f^j
Pi


B
m
l
05

TABLE





j_j
w —
O D
U Dl 2
2 --.
Q) -H
O> l-i cn
(0 Q) en
M Q. 
«} O —
Jj 4-1
O QJ
EH H

M— J
o
tn1 §
X -H
0 £*
u * ^
T3 -H en
QJ g en
M C rH
2 oi2
^ ^
2 QJ
1 *
*^


O 0
JJ
kl W -H
QJ JJ E
JJ. U -H
Sr-< ^
tJ rH
3 -O
2 O
^
a







CO
en
CN







(N
CO
CO
VO






if
(N
m
in
CO
•a1
o
(N





CN
r~
r^
r~
t-i











ro
in
<— i
T-H







en
(M
[-

rH



C
O
-H
JJ
03
' '
D
E
l-i
0
4-1
OJ
1-1

i-l
o
•r— i
(0
E

03
T3
Q)
Q)
C
jj
O
c

o
T3
jj
-H
E
•H
i— 1
QJ
JS
JJ
M
QJ
>
o
w
jj
u
3
•a
0
S-l
a
4-(
0

•a
'H
-H
x;
jj
i
QJ
C
0

jj
CO
x;
jj
c
o
-H
JJ
a
E
3
W
tn
oo
Q)
x; •
jj —
jj
c x
O 0)
Jj
TJ
0) QJ
W QJ
03 W
CQ —

ID







>l
J3
T3
Q)
-H
rH
a
-H
JJ
rH
3
E
„
M
TD
VH
CO
T3
C
05
JJ
W

4-1
0
Q)
rH
XI
03
E-
r^
CTl
cn
rH
QJ
x:
Jj
VH •
0 >i
4-1 QJ
>
dP In
0 3
CM W
4-1 QJ
o x;
jj
w
Q) £
JJ O
03 l-i
Dl E4-i
C -H
•H JJ Dl
-U MS
C Q)
3 en
O Cm
lH O 0
-H
O JJ ^
JJ U •*
3 n
01 T3
3 QJ 4-1
T3 l-i O

4-1 VI U)
4-1 C C
O O O
-H -H
0) cn cn
X! cn cn
•H -H
>i E E
co 0)0)
c i
Q) ^l
l-l O
(Xi JJ
03
r-l
• 3
>i Dl
0) QJ
> OS
l-l
3 2
CO M
<
>1
l-l QJ
o x;
jj jj
c
Q) x:
> jj
C -H
M 3
W C
C O
O -H
•H JJ
W 03
cn 1-1
•H QJ
E a
W 0
o
0 U •
O C*O
> C 0>
•rH CA
W iH
Dl C
C 0
•rH -rH
JJ JJ JJ
CO CO l-l
0 -H 0
cj u a
O QJ
QJ cn oi
u cn
03 < JJ
4-1 4-1
VH cn co
3 D> ^
W C Q
•H
2 -U rH
M CO CO
,r-t ^
Industr;
Nationa
Industr;

Q)
u
lH
3
O
W
2-15

-------
      Combining the cost and emissions reduction estimates  for
the  survey population, an estimate of the  cost  per Mg of VOCs
reduced is derived as $298/Mg.
2.3.2   Exceedance Fee Option
      As an alternative to strict adherence  to the limits
imposed by the TOS,  the EPA is taking comment on an option to
provide AIM coatings producers the alternative  of paying a fee
per  unit of output for products that exceed the limit.   The
per-unit fee will be computed as follows:

            fee = (VOC content - VOC limit)  • rate   .       (2-3)

VOC  content is measured in grams per liter  (less water and
exempt  compounds)  and the fee rate is paid  on the grams per
liter in excess of the limit.   The proposed fee rate is $0.005
per  excess gram per  liter (1995 dollars) with annual adjust-
ments based on the gross domestic product  (GDP)  price defla-
tor.  Total fee payment per product simply  equals the per-
liter fee times total liters of production.
2.3.3   Reformulation versus Exceedance Fee:  Least-Cost
        Analysis
      In this section,  an expansion of the producers'
compliance options is considered by allowing for the fee
option.   The premise of the analysis conducted  in this  section
is that AIM coatings producers will choose  the  less costly of
the  reformulation and exceedance fee options as  a compliance
strategy.d  The choice  is based  largely  on two product-
specific factors:  quantity of output produced  and the
"excess"  VOC per unit.
     The diagram in  Figure 2-2 helps explain the effect that
output  quantity has  on the choice between reformulating the
    dThe coatings manufacturers may undertake the reformulation activity
themselves,  though, in some cases, manufacturers who allocate few resources
to R&D may  rely on material suppliers  to provide compliant formulations.
Although the reformulation activity may be performed by another firm, the
costs incurred will presumably be borne by the manufacturers facing  the
regulatory  constraints.
                              2-16

-------
                                        Average Cost of
                                         Reformulation
                  QT'
QT
Output Level
                 Favors Fee        Favors Reformulation
             Figure 2-2.  Fee versus reformulation.

product and  paying an exceedance fee.   Since the cost of
reformulation  is  a fixed cost (i.e.,  it is independent of  the
level of output)  the  average reformulation cost per unit of
output falls as output  levels increase.  This situation is
represented  by the downward-sloping line in Figure 2-2.
However, the exceedance fee  per  unit  of output is constant
with respect to the output levels.   Let F be the fee per unit
of output; the flat line extending  from F on the vertical axis
indicates that the fee  rate  is constant.   In Figure 2-2,  for
all output levels  less  than  QT the average  cost  of  reformula-
tion is higher than the per-unit fee,  and for all output
levels above QT,  the average cost is below  the fee.   This
relationship indicates  that  the  fee is the less  costly
alternative  when output is below QT and reformulation is  the
less costly  alternative when output is above QT.  Thus  the  fee
is more likely to  be  chosen  by small volume producers,  all
else equal.  As Figure  2-2 illustrates,  the existence of a fee
                              2-17

-------
places an upper limit on the per-unit costs of complying with
the regulation.
     Figure 2-2 also allows us to see the effect of different
fee rates on the "threshold point" of quantity, below which
the fee is the preferred option.  If the fee were F' instead
of F, reflecting either a higher assessment rate per Mg of
emissions or a higher amount of excess VOC per unit, the
threshold point would be lower.  Thus for higher excess VOC
categories and for higher assessment rates we would expect to
see fewer producers selecting the fee option,  all else equal.
Because the fee will be more cost-effective only for lower
volume products and lower excess VOC categories,  allowing the
fee option should have a relatively small impact on variation
from the aggregate emissions reduction targets as long as the
fee assessment rate is not set at an extremely low level.
Sensitivity to the fee rate is also analyzed below.
     The least-cost analysis presented here determines which
option (fee or reformulation)  would impose a lower cost to the
manufacturer for specific products within a certain VOC
content range from the survey.   For the purpose of this
analysis,  a product stratum is defined as all  products
existing in a specific VOC content range for a specific
product category.   An example of a stratum would be all
exterior flat waterborne products in the 101 to 150 g/L VOC
content range.   Thus,  for the TOS,  all strata  in the survey
are examined to determine those that exceed limits for their
respective product categories.   As indicated above,  the survey
includes data on the number of products,  sales volume,  and
baseline VOC emissions for each stratum.   These data are used
to compute average sales volume per product for all strata
exceeding the TOS limits.   Then,  the average (per liter)  cost
of reformulation is computed for each stratum  by dividing the
annualized reformulation cost per product by the average
volume per products.   Suppose the average sales volume per
product for one stratum is 100,000 L/yr.   Given the annualized
                             2-18

-------
 reformulation cost estimates  from the previous section
 ($17,772 per product,  1991  dollars),  the average reformulation
 cost for this stratum  is  $0.17/L.
      To determine the  exceedance fee  for each stratum,  the
 midpoint of the VOC content range is  used as an estimate of
 average VOC for the stratum.   This measure was used to  compute
 excess VOC content because  it  is consistent with the regula-
 tory definition of VOC content (grams per liter less water and
 exempt compounds)  and  is  available for each stratum.  By
 contrast,  the "actual VOC"  data  in the survey are not consis-
 tent with the regulatory  definition of VOC content and  are not
 available for each stratum  in  the  survey.
      First the fee rate is  adjusted to 1991 dollars by
 multiplying the proposed  fee rate  (in 1995 dollars) of  0.005
 by  the ratio of GDP price deflators for 1991 over the 1995
 (projected).  The resulting rate is 0.0045.  Suppose the
 midpoint of the stratum is  75  g/L  above the proposed limit.
 The  associated fee per unit would  be  75 •  $0.0045  = $0.3375/L.
 This fee is higher than the average reformulation cost  per
 liter ($0.17/L).   Under these  conditions,  it is assumed that
 products in this  stratum would reformulate rather than  pay the
 exceedance fee.6   This  decision would be reversed  if,  for
 instance,  the stratum exceedance were 25  g/L,  in which  case
 the  fee would be  $0.1125/L, which  is  less  than the average
 reformulation cost ($0.17/L),   or if average sales volume were
 25,000  liters per  product instead,  in which case average
 reformulation cost would be $0.68/L.
      These average reformulation cost and  fee  per-liter
 calculations are performed  for each stratum above the proposed
    eBy conducting the fee-versus-reformulation decision at the stratum
level, and basing the decision on average cost and fee for each stratum,  it
is implied that all products within the stratum are identical  to the mean
values.  In reality, there will be some variation around the mean so that
some products may find one alternative less costly while others find the
other alternative less costly.  This analysis is unable to capture this
heterogeneity with the available data, but presumably these effects are
smoothed out as the analysis compares means across the hundreds of strata
in the survey.
                              2-19

-------
TOS limits  to determine  the  relative frequency of
reformulation/fee  selections and their impact on costs and
emissions reduction.   Results are presented in Table 2-4 and
can be summarized  as  follows:   (1)  the fee option is chosen
for a sizable minority of  products exceeding the limit, and as
a result,  (2) there is a substantial reduction in aggregate
compliance  costs;  however,  (3)  overall product volume for
products selecting fee is  very small.91  As explained shortly,
the combined result of factors (1),  (2),  and (3)  is that the
fee option  provides a  means  to reduce compliance costs with
little negative effect on  overall emissions reduction results.
     Under  the standard  fee  rate of  $0.0045 (1991 dollars), we
estimate that the  fee  is the preferred alternative for 323 of
the 1,153 products (28 percent)  facing the reformulation
versus fee  decision/   However, these products only  account
for 26.1 million liters  of output, about  1 percent of total
coating volume for the survey population.   Total fee payments
for those products is  just under $4.0 million (average:
$0.15/L); however, the estimated avoided  reformulation costs
for the 323 products choosing  the fee is  over $5.7 million
($0.22/L) for a net aggregate  savings to  producers of about
$1.8 million.  Moreover,  because the fee  payment is simply a
transfer from one  sector of  society  (AIM  coatings producers)
to another  (the government),  the net social cost savings are
the full $5.7 million  reformulation  cost  savings,  less any
costs of administering the fee.
     One obvious policy-relevant  question is  how much will the
selection of the fee option  undercut VOC  emission reduction
goals.  Since products will  be  allowed to exceed the regula-
tory standard if they  pay  the  fee, the associated excess VOCs
from these products can  be viewed as the  foregone emissions
    £Note that 1,153 products represent two-thirds of the total number
exceeding the limits because  the other one-third were assumed to
reformulate without incurring the "major" reformulation cost.

                              2-20

-------









CO
H
CO
>H
iJ
<

*3,


EH
CO
O
U
1
r .
^
CO
< ^
S§
EFORMULATION : :
SURVEY POPULATI
: VERSUS R
FOR THE
w
W
h


,

^
i
CN
W
J
PQ
<;
£











01
c
§1 5
*> Ul~
^ c
° § w
^ •-, c
rn W 0
^ CO -H
2 •" jj
m 6 U
p 
tO E — -
3 ^
c 3 -u
c ,R w
<^o
l-l


0)
AJ ,
^ *>
10 (Tl
^S
^si
cui2:
fc




ro co r~
CO O 0
co ro CN
CN CN










in o CN
CN CN CN
in U5 U3
CN CO O
r~ r- ro
cyi o en
m m a~\

LD i^D U3
rH m CN
OJ CN O^
in rH CN
r- in vn
O CTi ^3"
UD -^ «ai
CN ^ ^r
m in CN
CN UD in
m in in








CN CN "S1
r- r^ -si1
r- r~ in
r- r~ in
r-H rH n





in CM in
•^ CN ^r
0 O 0
o o o
.
o o o












JJ
-H
E
•H
rH

w

r .
fcH
Q)
£
JJ
lH
Q)
>
O
ie excess VOCs (g/L
r^
JJ
>1
£1
T3
OJ
-H
, — <
pi
>r^
f^
^3


0)
JJ
ro
^
i
0) >H
w o
a, jj
ro
rH
• 3
>, 01
CD m
> a
i-j
3 S
W H
<
>i
V-i 0)
0 .C
^-
CD £.
>ji
-1— '
C -H
M 3
ings VOC Emissions
ion in cooperation
1993.
urface Coat
gs Associat
raft Report
co C Q
•r-t
2 AJ rH
M (0 K3
< O C
O -H
T-
UM
• T3
U) C
JJ (0
X! W
Cn JJ 3
•H C O
U) -H 3
C ID rtJ
M O. U
>irH X
!H ro M
JJ C AJ
WOW
3 -H 3
T3 JJ T3
C (0 C
M S M


CD
tj
M
3
O
W
2-21

-------
reduction resulting  from  the  fee.  This quantity is estimated
as follows:

                         Ej = eve, • Qi                    (2-4)

where evcL is excess VOC content per liter for stratum I and qi
is stratum I product volume in liters.  EiF is  summed  across
all strata selecting the  fee option to compute aggregate
foregone emissions reduction as a result of the fee selection.
     In Table 2-4 we see  that foregone emissions reduction in
1997 amounts to roughly 883 Mg per year for the survey popula-
tion.  This is approximately one-quarter of a percent of base-
line emissions, so the aggregate effect is quite small.
Dividing the avoided reformulation costs  ($5.7 million) by the
foregone emissions (883 Mg) indicates that the fee option
precludes reformulation efforts that cost approximately $6,500
per Mg of emissions reduction.  This is more than 20 times the
average reformulation cost per Mg indicated in the "reformula-
tion-only" scenario in Table 2-3.  These results suggest that
the fee option provides a cost-effectiveness "backstop" by
avoiding reformulations that generate relatively little in the
way of emissions reduction benefits.   By their existence,
however,  these fees do provide an incentive for marginal
reductions in VOC content down to the limit,  since the fee is
paid on excess VOC content per unit volume.
     To analyze sensitivity of this decision to the fee rate,
we also examined the effect of halving the fee from the
proposed rate of $0.0045/L to $0.0022/L (1991 dollars).  The
lower rate leads to a significant increase in the number of
products selecting the fee in 1997:   565 compared to 323 under
the higher rate.  Interestingly,  total fee revenue rises with
a rate reduction to $5.1 million.  Still,  only a small portion
of coating volume is affected and the foregone emissions
reduction is now over 2,308 Mg,  which is almost three times as
large as with the higher fee rate,  but still  just over
                             2-22

-------
one-half  percent of total baseline emissions.   Avoided
reformulation costs are approximately $10.0 million,
indicating  that  the cost per Mg of these 2,308 Mg of emissions
foregone  by paying the fee is about $4,350 per Mg.
     Table  2-4 also examines the sensitivity of the results to
the  cost  of reformulation by doubling the assumed costs  (to
$35,544 per year).9  The  results are  similar to  those  derived
when halving the fee in terms of the number of products now
opting for  the fee option and the  foregone emissions
reduction,  but the fee revenue effect under the double-
reformulation-cost scenario is higher since the fee rate is
not  reduced.

2.4  MARKET ANALYSIS

     In this section,  market effects of the regulatory action
are  analyzed by  presenting a model of the firm's decision to
reformulate,  pay an exceedance fee,  or withdraw the product
from the  market,  followed by a model of how the outcome of
this decision collectively affects aggregate  supply conditions
and  market  outcomes.   Then,  operationalizing  the model using
baseline  market  data and regulatory costs is discussed to
analyze the welfare effects of these market outcomes.
2.4.1  Potential Firm  Responses and Market Effects  with
       Product Reformulation
     In this  section,  the current  "least-cost"  model of the
reformulation/fee decision is extended to include the
possibility of product withdrawal  if the cost  of the least-
cost option exceeds  the profits generated by the product.
Since the exit option  completes the set of producer options
considered  here,  this  discussion can be viewed as evolving
    9The increase in annualized costs could reflect either an increase in
the  initial cost  of reformulation or a reduction in the time  horizon of
benefits (from infinite to finite).
                              2-23

-------
from a  "least-cost" to a "best-response" criterion for
decisionmaking.
     A  simple model is presented of a firm's decision to
either  reformulate or pay the exceedance fee and remain in the
market  or to do neither and exit the market.  The potential
for some firms to withdraw products from the market and for
the effect of the unit fees on the marginal cost of fee-paying
producers is incorporated into the discussion of potential
market  effects.
     2.4.1.1  Firm-Level Model Extension.  Up to this point,
the analysis has focused on firms responding to the regulation
by choosing the least-cost regulatory option.  However,  this
view of a producer's likely response is incomplete because the
cost of the regulatory response must be weighed against the
benefits of the action to the firm.  Here the analysis equates
regulatory compliance with the decision to pay the costs and
remain  in the market.   Thus the benefits of the compliance
action are the net returns (revenues minus variable costs)
obtained from continuing to produce the product.  The net
payoff of compliance for a particular AIM coating exceeding
the limit can be expressed as follows:

                   nR = P • q  - c(q)  - r*  .               (2-5)

To ease the notational burden, all terms are expressed in
their annualized form:  P is  product price,  q is annual
output,  c(q)  is the cost function with respect to annual
output,  and r*  is  the  annualized  cost of  the  least-cost  option
among regulatory responses (i.e., reformulation or fee).  In
other words,  r* gives the cost of the solution to the least-
cost decision discussed in the previous  section.
     Assuming that the market for the AIM coatings product is
competitive,  the price-taking firm maximizes profits  by
                             2-24

-------
maximizing function nR with respect to  its choice of q.   This
is  indicated by the first-order condition

                dnVdq = 0 :  P = dc/dq  +  3r*/6q   .           (2-6)

Here  the profit-maximizing  solution equates  product price with
marginal production cost  (dc/dq) plus the marginal compliance
costs (dr*/dq) .  Note that,  if  reformulation is chosen as the
least-cost compliance  option,  then dr*/5q = 0,  because the
cost  of reformulation  is  independent  of product output level.
However, if the exceedance  fee  is the least-cost  option,  then
<3r*/6q = F, where F is the per-unit fee rate.
      The solution  to Equation  (2-6) generates  (first-order)
optimal output level and  profits, qR*  and nR",  respectively.
However, the firm  will only operate in  this market if it can
cover its  production costs and  compliance costs;  that is, if
the following condition is met:

                       nR" (qR*  r') > 0   .                   (2-7)

If the condition in Equation  (2-7) is not met,  then the  firm's
best  response is to withdraw the product, produce no output
(qR* = 0) , and generate zero profits for the product  (nR* = 0) .
      The absolute  change  in the output  of the  regulated
producer can be approximated as follows:

  Aq = qR"  -  q° = (6q/3p)  (AP - dr*/dq*)  (remains  in the market)
    = -q°   (leaves the market)                               (2-8)

where  (6q/dp) is the firm's  usual supply response with respect
to a price change  (i.e.,  the inverse  of  the marginal
production cost function) and  (AP - 8r*/dq*)  is  the change in
"net" price  for the firm, reflecting a  change  in  the market
price after  market  adjustments  take place less  the marginal
compliance cost.  Again the marginal compliance cost term is
                              2-25

-------
zero when  reformulation is chosen and equals the per-unit  fee
under  the  fee  option.  For products with VOC content below the
proposed limits  (i.e., unconstrained by the regulation), the
supply response  can be characterized as a special case of
Equation  (2-8) with the marginal compliance costs  (<3r*/dq")
equal  to zero.   Net changes for each producer remaining  in the
market depend  on whether the price increase exceeds the
marginal compliance cost.  When it does, as is the case  for
unconstrained  producers, the output effect is positive.  When
the price  increase is less than the marginal compliance  cost
(e.g., a relatively high per-unit fee),  the  output effect is
negative.
     The change  in market price depends on the aggregate
effects of the supply responses of the individual producers.
Product exits  will shift the aggregate supply function inward,
and marginal cost effects,  such as the per-unit fee,  will
shift  the  function upward.   This change can be expected  to
raise  the post-regulatory market price as the new equilibrium
is attained.
     Appendix  B  describes the methodology for incorporating
the reformulation/fee/withdrawal effects into a linked
multiple-market  model framework.  Appendix B also presents the
methodology for  measuring the social welfare effects (e.g.,
producer and consumer surplus)  of the changes in market
equilibrium, which is effected by the proposed regulation.
2.4.2  Model Execution and Results
     To estimate the effect of VOC content limits on AIM
coatings markets, a baseline characterization of affected
markets is constructed,  empirically estimated the shifts in
market supply  and demand as a result of  the regulations are
computed,  and  the market equilibrium model is applied to the
data to generate changes in prices and quantities in each
market.
     2.4.2.1   Baseline.   The coatings categories are grouped
into market segments,  as defined in Table 2-5.92'93  The price
                             2-26

-------
   TABLE 2-5.   AIM COATINGS MARKET SEGMENTS BASELINE DATA
No.
1
2
3

4

5
6
7
8
9
10
11
12
13

Market segment"
Exterior & high performance
solventborne coatings
Exterior & high performance
waterborne coatings
Interior solventborne
coatings
Interior waterborne
coatings
Solventborne primers &
undercoaters
Waterborne primers &
undercoaters
Solventborne clear
coatings, sealers, & stains
Waterborne clear coatings &
stains
Architectural lacquers
Wood preservativesc
Traffic marking paints
Special purpose coatings
Industrial maintenance
coatings
Totals /averages
Quantity
produced
(kL)b
162,
468,
94,

833,

61,
75,
134,
120,
40,
27,
91,
34,
231,
2,375,
937
345
935

434

298
212
678
738
Oil
449
067
568
261
933
Value
($103)
540,
1,046,
302,

1,747,

171,
160,
412,
266,
83,
493,
132,
141,
797,
6,296,
511
383
264

341

583
960
743
174
320
965
358
633
006
241
Average
price
($/L)
3
2
3

2

2
2
3
2
2
I
1
4
3
2
.32
.23
.18

.10

.80
.14
.06
.20
.08
.45
.45
.10
.45
.65
a  See Appendix A for an explanation of  products  included  in each
  market segment.

b  The quantities and values are taken from Census  data except  the
  quantity for wood preservatives,  which is taken  from the survey.

c  For wood preservatives the quantity is taken  from  the survey, but
  the price is taken from the Census data.

Sources: U.S. Department of Commerce.  Current  Industrial Reports:
         Paints  and Allied Products, 1991.  Washington, DC, U.S.
         Government Printing Office.  1992.

         Industry  Insights.  Architectural and  Industrial Maintenance
         Surface Coatings VOC Emissions Inventory Survey.  Prepared
         for  the National Paint and Coatings Association  in
         Cooperation with the AIM Regulatory Negotiation  Industry
         Caucus.   Final draft report.  1993.
                                 2-27

-------
and quantity data necessary to analyze market effects are not
provided in the survey conducted for this study but are
available from the U.S. Census Bureau Current Industrial
Reports publications.94  Because the Census Bureau categorizes
AIM coatings products differently than they are classified for
this study, the market segments were constructed so that data
can be used from both sources and provide the necessary level
of resolution for market analysis.  This process resulted in
the 13 market segments presented in Table 2-5.  Appendix A
provides the details of this product/market cross-referencing
scheme.
     Table 2-5 lists quantities and value of shipments for
each market segment.  From these data, the average price for
each market is imputed.  Because the market segment price is
an average value, it may obscure heterogeneity of products
within each group.  Although the model aggregates different
products together to construct individual market segments, the
objective in aggregating to the market segments in Table 2-5
is to provide a level of resolution that both highlights
differences in the end use of the product  (e.g., exterior
coatings versus interior coatings) and distinguishes between
groups that will be affected differently by the VOC content
regulation  (e.g., solventborne versus waterborne).  Eight of
the 13 segments consist of four pairs of related product
groups; one in each pair represents solventborne products and
the other represents waterborne products  (e.g., interior
coatings) .  Although the products in each of the paired market
segments possess different attributes, they perform similar
functions,  thereby  suggesting a high degree of product
substitutability in demand.  Demand elasticities are estimated
using  procedures outlined in Appendix A.  Supply elasticities
could  not be  econometrically estimated because of data
limitations;  therefore, the aggregate supply elasticity for
each market segment was assumed to be unitary  (1.0) .
                              2-28

-------
     2.4.2.2  Quantifying Market Shocks.   To simulate the
reformulation/fee/exit decision, per-unit profits are
estimated to compare with unit costs for each stratum,  and
computed as follows:
                         nu = P • m   ,                     (2-9)

where, P is output price and m is the profit margin.  For each
product category analyzed,  the average market price in Table
2-5 for the market in which the product category belongs was
used.  The model derives the returns-to-fixed-factors (RFF)
profit margin as follows:

              m = 1 - (variable cost/revenues)   .         (2-10)

The ratio of variable cost to revenue can be computed using
values provided by the NPCA.95   The variable cost component  in
the numerator includes cost of goods sold plus variable
selling and storage costs.   These variable costs comprise 81.7
percent of revenues for the mean producer surveyed by NPCA, so
our estimate of the RFF profit margin is 0.183.
     The least-cost regulatory option for each stratum in the
survey exceeding the TOS limits is computed in the previous
section.  For the market analysis,  the least-cost solution
obtained previously is compared to an estimate of per-unit
profits.  If the cost term exceeds the profit term, that
stratum is identified as a "withdrawal" stratum.  If the
profit term exceeds the cost term and the least-cost option is
reformulation, the stratum is identified as a "reformulation"
stratum.  If the profit term exceeds the cost term and the
least-cost option is the fee, the stratum is identified as a
"fee" stratum.  The model computes the total quantity share of
the withdrawal strata by summing the total quantity from these
strata  (Qsx) and dividing by  the total baseline quantity from
all strata for that market segment in the survey (QST) -  This
share was then multiplied by two-thirds to compute the market.
                              2-29

-------
quantity  subject to the withdrawal option,  which is denoted  as
the term  Rx.h

                     Rx  =  (QSX/QST) • (2/3)  .               (2-11)

Similarly,  the model computes the total quantity shares for
the reformulation (R superscript) and  the  fee strata  (F
superscript),  respectively:

                      RR =  (QSR/QST)  •  (2/3)                 (2-12)

                     RF  =  (QSF/QST) * (2/3)  .               (2-13)

Finally,  all quantities not allocated  to  the exit,  reformula-
tion, or  fee actions can be viewed as  the  unconstrained share:

                     Ru  =  1 - Rx  -  RR  -  RF  .               (2-14)

     To perform the market and welfare effects calculations,
the initial baseline market-level values  for the exiting,
reformulating, fee-paying, and unconstrained sectors  are
obtained  for reasons explained  in the  methodology description
in Appendix B.  The model derives baseline  quantities by
multiplying the quantity shares derived from the survey data
by the  initial baseline market quantity,  Q0.

                          Qx = Rx • Q0                      (2-15)

                           QR = RR • Q0                     (2-16)

                          QF = RF • Q0                      (2-17)
     "Multiplication by two-thirds incorporates  the previously discussed
assumption that one-third of all products exceeding the limit can be
costlessly reformulated  (and thus would not be withdrawn).
                               2-30

-------
                         Qu = Ru • Q0  .                    (2-18)

     To quantify the supply effects of  the per-unit fee from
the fee-paying sector,  as indicated in  the equilibrium model
discussion in Appendix B, the model computes a value for the
unit fee.  Since the fee actually depends on the VOC content
of the product, the model computes an average fee for the fee-
paying producers in each market segment by taking a quantity-
weighted average of the stratum-specific fee for each fee-
paying stratum per market segment.
               NF
           F = £  F. •  (QsiF/QsF)                          (2-19)
where Fx is the fee for fee-paying  stratum I,  QsiF is stratum
I's quantity, and N is the number of fee strata in the market.
     Finally, note that the measure of producer surplus losses
requires an estimate of market-wide reformulation costs.  The
model estimates this cost by first taking the number of
products in the reformulating strata from the survey and
multiplying it by the annualized cost of reformulation.  This
calculation gives an estimate of annualized reformulation
costs for  the survey population, which is a subset of total
market quantity.  The model then computes the market-wide
estimate by multiplying the survey population reformulation
cost estimate by the ratio of survey quantity for that market
segment by the total market quantity indicated by the census
data.  This calculation may provide an underestimate of
market-wide costs if the ratio of the number of total products
to survey  products exceeds the ratio of total product volume
to survey  volume.  For instance, in some markets many of the
nonsurveyed products are produced by small producers and are
low volume products.  In these cases, the ratio of total
products to surveyed products will likely exceed the ratio of
total volume to survey volume.  The magnitude of the effect on.
                             2-31

-------
the cost estimates is unknown,  since there are no data on the
nonsurveyed products (such as the number of products over the
limit, VOC content, and other data necessary to determine how
many products face reformulation, how many will opt for the
fee, how many will be withdrawn,  and how many will be
unaffected).   However,  the possibility that these cost
estimates are biased downward should be considered along with
the other downward and upward biases discussed previously.
2.4.3  Market and Welfare Effects Results
     To assess the potential effects of the proposed regula-
tions we considered four separate scenarios:
     1. "Standard":  reformulation  (at $17,772 product/year
        reformulation cost) or exit
     2. Reformulation plus fee option:
           reformulation cost = $17,772/product/year
           fee rate = $0.0045/L  (1991 dollars, equivalent to
           $0.005 in 1995 dollars)
     3. Low fee rate (one-half):   $0.0025/L  (with standard
        reformulation cost)
     4. High reformulation cost  (double):
        $35,554/product/year (with  standard fee rate).
     The first scenario reflects the impacts of the TOS and
reformulation costs, while the fee  rate considered in the
regulation.  The third scenario considers the impact of
setting a  lower fee rate.  The fourth scenario is generated to
see whether a substantial  increase  in the cost estimate would
have a large impact on the estimated market effects.  This
scenario was modeled because, as will be seen below, the
market effects are quite small under the other scenarios.
Each of the four scenarios was modeled for the proposed TOS
limits, leading to a total of four  alternatives.  Tables  2-6
through 2-9 report the associated market and welfare results.
A  separate estimate of welfare costs that increases the number
                              2-32

-------















Q
OS
j*
<•
Q
2
EH
CO


rH
O
M
OS
g
w
u
CO
1
CO
EH
.1
t— t
*3
*—>
CO
w
OS

i-3
H
D
O
s
EH
W
^M
PS
s




VD
1
CN
w
j
m
 t/j.
(i) —
z
c "*•' •"""
.is c w o
g 3 -
S § a-
§ * ss
ri O w
u 01 —
•S|SS
0) 5 r*
£ | S£
5 g g S
u u 
•n
il 2 ~
o -S ° o
c 2 Si «
D a $ ~
w
*T w
£ g = » -
. V- -,H U o
« 0^3;;
;2 «w * TJ v>
0) -H O
g « = &
"3. en „
5 .5 2
" i?§-o
n ^ 3 ,— I
m 7" "0 
S S 2 ~
•o u a
o
a „£
C 0) *~"
.5 •- So
*J 3 rH
* X 0 W
? w °
re a
u

fl) "o
J-1 rH
O  -H -v
Z >-" v>
Q. —
>,
ti j
S.fH -1
„ XJ
0) 7- "
z g o
" rH
&~
c
•.H — .
QJ t— (
Oj U cr\
DI -H en
C ^ ^
re a v>
6 ~
>i
« "3
W -rH «-J
ecu
<0 -H C o
u 1 =

Jj
c
0)
1
01
tn

ij
o>
x
re
£
O
Z
o
0

p"
m
CC
m
1



O
O
o


o\
T^
in
r~


0
m
in

CT>
CN
O
CC
v£i
n-i


0
O
O


ri
p-
U3
P-
1

CN
in
_,
U3
U3
m
i
•^T
CO
m

m
v>
n
U3
CO
pj"
U3
1-1
^
0
o
o
v>

m
p-

X
tn

!c i o>
o> e
•Jl O M
c o
^ re ja
O £ *J
-H W C
WOO)
0* «u >
X 01 O
U O. «

*-*
i^ tn ^-* ui tn co
1-1 (N m — f o o

^ fN V3 rg o co
'-O T-H f^i CC tN O
i (J\ 1 CC 1 CO
mm f-i




o o o o o o
o o o o o o
o ooo o o


\o m CN vo I-H o
vo r- P- m *» r-
o OD TT ^n cr^ o\
p- m co •& vc ^
i i i m i pj


CO P- O P- PJ 00
O i-t P- m ro cr>
in o -w 04 cr\ ^»
fN iH
rH CO O -^ •-* MS
cr\ p*  1 U3
mm m


O OOO O O
O OOO O 0
o ooo o o


»J ff\ C\ i-H "O- CTi
\o oo pj m vo xo
in m pj m p- o
1 rH (N ^T 1 O


m o o co p- co
in in CN r- m m
^o m oo o 0*1 co
r- «* *? ^ m
CD m in
mm m
•o1 m r- p- o m
*» ^ \D "^ «-H 1C
m oo 01 o ^ ^D
CM rH O 00 i-H O

(N m fN CN CN m
V> t/> V> V> V> V>
(N *? r- o m o
U5 PJ rH ^ (N CD
m o-> ^T rH fN m
CD •"» m rH m ^i
u) o~i m vo p- m
•^T CO rH
fN ^ rH \O ON CT\
o o o in o -M
0 OOO O O
o ooo o o
o ooo o o
W V> V> V> Vr tO-

CO (N oo oo m OD
rH rH t-t in T-t 0^
1 1 rH 1
t
X m
w - c
• o)0> >EO> 01 E  ro -H
xure co cm one -
UD.3 •-<« w3 Uia3 3O.3 wuw

CN m ^ m vo p-
r-> p-
•W T

rs; ^?
r-i ys
( ^-





O O
O O
o o


a\ o
m o
rj o
m


a\ o
in o
PJ 0

O P-
o *»
o *»
^D
-w



o o
o o
o o


m O
T 0
rvj O
rH
'

VO P-

O *»
•^ ^D
<»

O *J
m fN
O CD
(N O

fN fN
Vr V>
rH  v>

P- O
1


to
CT <-*
c re
-H U
0) iJ 3
C (0 *-"
u o n u n
50 c a> w
•H 4J 01
Ul r4 (C -H 3
01 * *J X &
m r-* n a
S u J> < •-<

CO (T>
CO CN P- P-
rH ** r- *r

P. in mm
p- •<* rH p-
m m p- in
--H \£>




O ,-, O O
0 g 0 0
o 0' oo


in PJ * co
^- CO rH 0»
m O m «-*
VD fN CO O
rH rH m m
rH

in
p. rH PJ rH
O Cft 00
rH
(N P- ID ON
IH m -
CO *» CM CO
en m m PJ
o en PJ «)
p- m p- o
(N P) in O
rH \O
1 1


0 0 O O
o o o o
o o o o


O) \O iH \D
CO rH rH O
^" m m p-
^ m o m
i i rH m

m O rH CT\
p- ^ ^D rH
m ^ o fN
rH fN m p*
P4 PJ m PJ
i •• i
rH m
i i
*» P- mo
fN ^" CO (N,
m m o in
*f *f rH ^

rH rH T m
v> v> w v>
fN U"l rH PJ
en PI oj r-
m o mo
P- rH ^ rH
PJ cn mm
PJ
O m I-H \o
\£> rH rH IO
O O rH O
0 O O O
O O O O
w w 
in •*» ^f co
1 1 1 r-t
t
Oi Oi
C tn
« -H O
0) M Q.
> w u
•* «J 3 rM
we a c
A -H
> o »H n
M -H co re AJ
Q) l*-( AJ -H 0)
•O n «w c o 3
I §, S I S I
O »H fN m
rH -H »-( rH































































«
(J
c
<0
c
0)
JJ
C
re
£


2-33

-------




w
H
Cn
CO
J
P-i



0
H
EH
*3<
t_3



O
fc,
w
ex;
CN

O
H
OS


1^
u
CJ
CO

1

CO
EH


*""j
CO
w


w
Q
0
S

EH
W

&
s



•
r-
i
CN

W
m
j^
EH






| B °
CD 0) ^
at ^
2
5 £ 2 2
£
c £ & »
5 |S?
c ~
•~ jj w o
Q) § r^
Ci 5 D. M
c J2 u <*


, •g 
C u 2 ~
5 a
u



(0 o
o «
e-> ^-


aj 	
3 u J
S1S


>^
j .t< J
2 C "o

&~
a) "*


*° *H 'u 2
o a £

^,
01 u ~
O> -H J
C C U „
« -< c ^>
cr
4J
C
0
0
CO
iJ
0
V.
£
o
Z
VC
T4
s
7

m
oi
o

*~4
O
CTs


'

O
«-•
o

rH


O
CN
n
in
o
ft


m
o




r-
VO
7


CD


*-«
^
m
1
cJ
^
ro
v>

O
r-


"
^^
fN
O
O
o


r-

r-<
£

•H
O


X
u

r-(
eo S^cn S """ ° r-oorMvocN
1-1 cT\ino r- oo o "»vommm
^> OOJ.-I --( CO 1 VOVO'JVOoj
7 ^ ^ ' ' ^ ' „•
'
p- vomrM vo m m otrvo^oo
o in 1-1 f4 CD r-t m o^om^
r-i in m in r-J CN rn O WocNr-i
S 5 S S 5
r-t 04
CN m^tm m co <*-i otrifN^
o cNoor- CN m ty> or-cooo
vo i^rncn m o vo omovooo


r-t fs)
1 1

vo ^r-o m m CN ovo«HoOf-i

vo r-mtn ^ (N vo ocor-co *AV>V> W V> V> V>

CN mmm cri ^ m r-immmm
m CO^»GD nj m r- o-^oinoo


^> CD r-l r-t CN
m cNrHrH n in n ovoonvo
O rlOr- (N CN O O^r-(COr-t
0 O 0 r-l O O O OOOO.H
o ooo o o o ooooo
o ooo o o o ooooo
v> vt vt vt- v> v* v> v> v> v* «• w

oo mooin oo n in o*»
01 - c « « -S o
cue 01 C CrJ i^c-HjJ -H X4 '"m 3^-iu
co cc o coutJJCt«jJO °m'~H U£


aw was ^w i-t s was sa3 wu« su^i «t--i sa E-Q. w we
O «H CN m
fN f-i-wm vo r- oo c^rnr-i-Hr-t
p*
en
ac
7
o
f-1
r-
Cfi
"
^
CN
p,'
CC

in

^
m
CC
VO
*H
*
^^
VO

r-
m
r-


i

CN
DC
m
r-
1

VD

in
m
•^
CO
*•*

















r-
r*
r^






-------















w
EH
3

w
w
fa
1
S:
0


ro
O
H



n
3
H
3
CO
lucer
u
o
a
c
o>
01
\
u




































v _—
*"" O
to n ._,
^ U r-t
01 0) a,

•*-> (U t/>
0) v
2
.5 c w o
1 3 ^
o> 5 .-H
o> r a •"•
C r. u °^
* > 3 2
u g, w */>

c ~~
« §|:
c !2 u 

, 7 w
c c g ~
c 2 jjj 2
M w
u)
E C 01 _~
S - 3 2
«-< (0 T) 
0) ^ O _
a: 3 v.
a

c w
••I *"* — .
0 | °
?••§ «
s £ -


w
Cn w
C 0| -
'U 3 °
-^ TJ rx
X O _
a

11 "o

o Z>
H _


01 ~
S u J
CD -H ^
2 W v>
C. —
3 £ 3

2 C "o
s. ~
&



g .C .H en
— ^ t-<
u a JJ
C ^
* *J


?•?.—•
u °


c
s
i
«
in
jj
01
i
0
2
r- rv rffMfM m ,- r^ om/*i^in
in VD ^- TT a> r- en o Tfcn^r-4cri
CD oo vom— i CN i*i«-^fn •^ co *"^ rjcocof^f^

CO ^»H in_ *** C»ff»
*-« (N


2 S 5§l; S S S SSSSS

^J1 i-l OJOJCO i-l ro 1 rMI«Hi-tO
I 1 1 1 1 1 1 1 II-
T*

^
m S SS2 2 S S 3?^S^
ro ro O) ^ «-< ro IH 9 CO in ^O in y-j
^ N

fN i« OJ_tn -, O r-) OO1C°i-\^
r- CD ojinoj ^f r~ co minorjro

VD OJ(NVDr-»CTi .HOJiHi-HVO
1(11 II

o in incor- oo ^ o\ t^-r^'Hoin
f-> OJ ,H 0 CO r^ 0  o ONr^tH^ro
VD VD ^S^ ^ ^ ^J roojchmro


oj o oaOro ^-t 01 o r-riiHinm
O O OOt-l O O O OOOOrH
O O OOO O O O OOOOO

O O OOO O O O OOOOO

OJ VO COVD^ CPl «H CD OO^fNi-4
oo vo en co  «*^ 01-HO
j: i o> x; i a> mm tnaiCffi c t) at ^ a
OIC 0) C C^ ki c -^ *J -1 *-< >*-• ^ (U
co cc o coji-ucuj-uonJ c * u *J e D,QC

o e 4J o £ o o^j oo jjwo ono a.) u w o u c a> i-« > u <-< ^ c
•HtiC-HUjQ-HC-Hja CWUJ3V4U C VJJ3 -H4JCI W-HOT (0 4-* (H
u w o t-< k< o> wt-i a) Q) u< Lifuv^ ai^ai s-( u *o -H 3 u "4-< iJ -H tn AJ
0>*4-i> Q)vwQ) (D> 4)4) >6dl £QJ > (0 •-* flJftjAJ XCT'OW <0 CO C
-------







o
H
EH
d
^*j
S
OH
o

w

EH
O
CJ
1
ffi
o
H
ffi


f t
"HI
o


ss
w
u
CO

1


CO
EH
^
D
CO
w
K

i_5
W
Q
O
S
EH
s
K
^C




•
CTi
1
CN

W
CQ

EH







*-* "o

c ^
• H JJf U) O
§3 ^
v "o, i~l
C « ti "*
£ o 5 S
(_> 0 V>
n
'  r- CM *» CD
^ "" or^rsj r- m -H fN(N^i-HrM
o o rn o\ -q> o -T o fNoor-or-
^ r< or^iro ri oc i fia\oco^
in rH O'O i in insc^rHi^
Tf ' kO " " ' ' " '
''ii *? T


** r^ ^mrg \o o m ooomtN
O «H mr^ico »H CN m o *r Q *r \o

- -" «

o in ccmr* p- in QD pocfiin^in
V£> rH COVOO O O O COr-rHCOfN
cr^ *o r** m r^* in  CN Of-r^rHO
1^1 ft oi^-rN m «> rn o% r-iinm
rH rH r-


rH m oonm (N£ CN fN ° ° ^ M P P


fl in rH ^ fN CJ
1 1 1 1 II

S 2 * 15 £ 0 ro ^ O^gSS

O 1 fN 1 fN 1 O 1 1 rH O>
rH ^| frj

rH eo o-iinc-, m co o a\(NfNcor-
in CN OO^HT, o m o ujvoi^cnCD
ri f-i rnr^tN in CD o rn'riorHr-'
«-< i i m ro m i «-H i r-


i

O r- fjcNin o os oo *cr^f-^^
^ « "^^^ °! °. ^ "^^^"^
S ^ *SS o £ N ®^£St2
in VO (N in rH CN
1 1 1 1 1
•W 0 CDrHO^ in fN ^D OOflMJO
(*! (N rHOCO rH O fN rH^^rHtf
f-t fN r-ifNfN fN rn pj oirHrH-^r^
 V> V>V>V> « V)- V> V)-V>V>V>VV
2 S «^^; S « s 5SSis
fN CD W m" o' in W o (* r-" o V m
 v> v>*/>v> v> v> v> ww>v>v>v>-


r- «> p. r» fH oi o in r-^McftJJ
^ rH rH «H O*l  OO JJ01O OWO u O tfl O U C 0)h > O «-^ rM> 0)*WOI 0)> OlOi >EO) OtgOl >«r-t (UfljAJ *£"r'SS *B^H GJ -O^
Xft»O XOlfO CO C<0 OlnC «J*HC Or-fO) (0<-< IntO Oti tlS ^ '"c
O "-< (N ft
rH (N m*»m «3 t- CO <^rHrHrHt-t
2-36

-------
of produces projected to face reformulation is provided in
Appendix D.
     In general, the price,  quantity,  and welfare effects are
quite small relative to baseline values.   Under each scenario,
the average price increase and quantity reduction across all
market segments are less than one-tenth of a percent of the
industry average baseline price and total baseline industry
output.  Estimated quantity reductions,  across all AIM
coatings markets range from approximately 670,000 L/yr under
scenario 1: Standard,  to over 3.6 million L/yr under the
"high-cost" reformulation scenario.  Still,  this latter figure
is less than two-tenths of a percent of the industry baseline
quantity.  Price increases are typically well below 1 percent
of baseline price, with the exception of the solventborne
primers and undercoaters market segment,  where the projected
price increase is as high as $0.034/L (1.1 percent) under the
high-cost scenario.
     In some of the waterborne market segments there is a
projected increase in market quantities,  as consumers
substitute away from the solventborne counterparts because of
the regulation-induced supply contraction and price increases
in those segments.  While noteworthy,  these increases are
quite small empirically.
     The method for estimating changes in consumer and
producer welfare effects are demonstrated in Appendix B.  To
summarize, the net welfare effect  (social cost) of the
regulation equals the sum of consumer surplus and producer
surplus effects.  Changes in consumer surplus measure losses
to consumers from higher prices and foregone consumption.  The
total change in producer surplus for each scenario equals the
sum of the change in producer surplus for the exiting
producers, fee-paying producers, reformulating producers, and
unconstrained producers.  Losses to exiting producers reflect
the foregone profits they would have received had they stayed
on the market.  Losses to fee-paying producers measures the
                             2-37

-------
net effect of fee payments plus the partial offset of these
losses by the rise in price caused by the regulation.  Thus,
the producer surplus loss estimates for fee-paying producers
in Table 2-6 through 2-9 are not as large as the fee payments
these producers will make.  For instance, the producer surplus
loss to fee-paying producers in Table 2-7 is approximately
-$4.7 million.  Actual fee payments under this scenario are
estimated at $5.2 million.  The difference between these
numbers ($0.5 million) reflects the offsetting gains to these
producers through the rise in market prices.  In a similar
vein, losses to reformulating producers measures the net
effect of the reformulation cost outlays and the offsetting
gains from price increases.  In the standard scenario modeled
in Table 2-6, the producer losses for reformulating producers
is estimated at -$23.4 million.  Actual projected
reformulation costs are $24.1 million; the balance
($0.7 million) is due to offsetting price gains accruing to
the reformulating producers.
     Note that in all cases presented in Tables 2-6 through
2-9, the producer surplus effects for unconstrained producers
is positive, reflecting the fact that these producers gain the
benefits of the regulation-induced rise in price, without any
change in their cost structure caused by the regulation.
Please note, however, that the welfare gains accruing to the
unconstrained producers are transfers from coating consumers
and, as such, should not be viewed as a net welfare gain of
the regulation.
     Costs  tend to be distributed across parties in such a way
that reformulating, fee-paying, and exiting producers
experience welfare losses by incurring the regulatory costs
(or ceasing operation) and consumers bear welfare costs
through higher prices.  Net welfare cost estimates range from
approximately $13 million under the low-fee scenario to $28
million under the high-cost scenario.  Welfare gains accrue to
unconstrained producers  (through higher prices) and the
                              2-38

-------
recipient  of  exceedance fee revenues, identified here as the
government sector.   However,  the government  may redistribute
these revenues  back to any of the parties  affected directly by
the regulations or  back to the citizenry via the Federal
Treasury.   From society's perspective, the net welfare effects
of the current  transfer method (AIM producers to the
government) or  alternative distributions  (e.g.,  back to AIM
producers)  are  zero.
     Approximately  1.4 percent of all products are projected
to withdraw under scenario 2:  reformulated  plus fee; these
products account for less than one-tenth of  1 percent of
baseline industry output.1 Not  surprisingly,  the  very small
products are  the ones most likely to exit  in light of the
relatively higher unit costs of reformulation.  When a fee is
introduced, about 5.8 percent of products  choose to pay the
fee instead of  incurring reformulation costs or exiting, but
these products  only account for about 1 percent of industry
volume.  Again,  the attractiveness of the  fee option seems
restricted to the smaller volume products.   Approximately 16.4
percent of products are projected for reformulation and the
remaining  76.4  percent of products bear no costs with the
regulation.
     To focus more  on the relative merits  of the fee option,
model results under the first scenario are compared with the
results under the second scenario.  The fee  reduces the net
welfare costs of the regulation by about $6.4 million (25
percent),  primarily through avoided reformulation costs and
product withdrawals.   As indicated in the  previous section,
these cost savings  are attained with very  little effect on the
     'Note that the analysis is performed under the assumption that any
product exceeding the category standard whose annual profits exceed
annualized reformulation costs will reformulate when the Table of Standards
takes effect.   It is possible, however,  that some products may satisfy this
condition but are not (immediately) reformulated because of technological,
product quality, or capital  constraints.  The product withdrawals estimated
here may therefore be viewed as low-bound estimate of true short-run
product withdrawals in response to the regulation.
                              2-39

-------
attainment of aggregate emissions reduction targets.  However,
a greater portion of the total welfare costs (29 percent) is
borne by consumers when the fee is allowed than when it is not
(12.2 percent) because of the fee's upward effect on product
prices.

2.5  SOCIAL COST-EFFECTIVENESS

     The results of the market analysis can be used to compute
measures of the social cost-effectiveness of the proposed
regulation.  The distinction of "social" cost-effectiveness is
made to illuminate the fact that the costs that are evaluated
are the net costs imposed on society (i.e., the net welfare
costs estimated using the market model).  Earlier,  when cost
per Mg of emissions reduction was evaluated, a static view was
taken, one with no product withdrawals, aggregate supply
shifts,  and market price and quantity effects to consider.  By
using the net welfare costs of regulatory scenarios, these
expected market effects are incorporated into the assessment
of costs in a manner consistent with microeconomic theory.
     The measure of social cost-effectiveness is computed as
follows:

                        SCE = AWF/AE  .                  (2-20)

AWF is the aggregate net change in welfare  (i.e., social
welfare costs), summed across all markets, and AE is the
change in aggregate emissions.  The social welfare cost
estimate is produced by the market model.  The change in
emissions estimate needs some elaboration here.  In Table 2-3,
an estimate is presented for the emissions reduction target,
but this estimate was for the survey population, a subset of
the industry.  To correspond with the welfare cost estimates
generated by  the market model, a national estimate of
emissions reduction must be used.  The baseline estimate of
                              2-40

-------
national VOC  emissions  from regulated AIM coatings products  is
approximately 481,000 Mg.3  Given  the  targeted reduction of
20 percent  in 1997,  the aggregate  emissions reduction  target
is approximately  96,200 Mg,  which  is AET.   However,  the
emissions target  must be adjusted  by two factors:  foregone
emissions reduction  due to  selecting the fee option and
changes in  emissions due to regulation-induced changes in
industry output.
     The first adjustment is computed by taking the ratio  of
foregone emissions reduction to emissions reduction target in
the least-cost analysis performed  above.  This ratio is
(AESFR/AEST) , with  the numerator indicating the foregone
emissions reduction  from the survey and the denominator  is the
emissions reduction  target  from the survey.  This ratio  is
based on the  survey  population but is then multiplied by the
national emission reduction target to provide a proxy of
foregone emissions of the different fee options at the
national level.

                     AEFR = (AESFR/AEST) • AET.               (2-21)

     The second adjustment  is computed by taking the ratio of
change in industry output to baseline industry output and
multiplying by baseline industry emissions:

                      AEQ = (AQ/Q0)  • E0  .                (2-22)

AQ is the change  in  industry output, which is the sum of
market-level  changes, Q is  baseline industry output  (2.375
billion liters),  and E0  is baseline emissions (480,816 Mg
indicated above).
     3This estimates is based on a national baseline emissions estimate
provided by Radian Corporation96  of 530,000 tons, which we convert to Mg by
multiplying by the ratio of tons/Mg = 0.9072.  The result is a national
estimate of 480,816 Mg.
                              2-41

-------
     Thus, the net change in emissions reduction is computed
as follows:
                    AE = AET +  AEFR + AEQ   .               (2-23)

The net change equals the sum of the targeted change,  foregone
emissions change  (due to fee) ,  and emission changes due to
changes in industry output via regulation-induced market
interactions .
     The analysis focuses on computing social cost-
effectiveness measures for the four market model scenarios
outlined above.  Table 2-10 presents the results.97  Social
cost-effectiveness estimates range from $139 /Mg for the low-
fee scenario to $299 /Mg for the high-cost scenario.  Social
cost-effectiveness in the standard scenario is $260/Mg, while
offering a fee lowers the cost-effectiveness by $64/Mg,
reinforcing  the notion previously discussed in the least-cost
analysis that introducing the fee provides significant
improvement  in cost-effectiveness, yet, further improvements
are possible with a reduction in the fee rate (Scenario 3) .
     As a normative policy issue, the socially preferred
outcome cannot be inferred without knowledge of the social
benefits of marginal reductions in emissions.  Moving from the
suggested fee rate to the lower fee rate reduces social costs
by about $5.7 million but forgoes about 2,000 Mg of emissions
reduction.   Dividing the cost savings by foregone reductions
gives the marginal social cost of actually achieving the
foregone reductions.  This figure is about $2,900/Mg.
Although this is much higher than the average social cost-
effectiveness numbers reported in Table 2-10, the normative
issue is whether  this action generates marginal social
benefits that exceed $2,900.  If  so, then society loses by
forgoing these reductions by allowing the fee option;  if the
benefits are lower, society receives a net gain by forgoing
the reductions.
                              2-42

-------


















CO
W
EH
S
^.TJ
II
EH
CO
W
CO
CO
W
£2
W
>
H
r .
U
W
fl4
fe
W
1
EH
CO
o
CJ

-]
v~j
^4
H
U
O
CO
O
rH
1
CN


W

PQ
rtj
EH





































a
Z




T3
0)

E
•H
m
W




T3
d)
4J
1
•H
4J
to
W





National
































1J
W
0


r— (
(0
''H
U
O
C/1





£
o
-H
w
en
•rH
£
0)





j-J
CJ
•£
»*D
E






(D
C
O
Dl
rH
O
14-H






targeted







01
rH
fO
14-1
rH
0)
4J
CU
2








0)
v
j_i
rrf
2




m
w
0)
C

CD en
> -S
•H \
U •—
CU

,
**H
CU




•D
CU
DI c;
c |P
£ —
o




„
Dl
4-1 S
D ^
a
4J 0
D '
0 -o
to




W Ol
C 2
o ~"
1/3 Dl
E its
(1) _£•
u




emissions
hanged (Mgi
u






^^
o *-'

1 ^
u t2







o

r-H '["J
o> ::
•o 2
0 £

CJ
en






o «3 a* en
i£> ci m c^
CN rH rH CN










CT* VO CTt rH
CN rH CM CO
vc in ^o ro
CTl CTt CT* O*\
1 1 1 1






l£> OO CO O
rH CN CN f~
1 1 1 1








O CM CM O
V « *
rH m n






p") po m ro
VO VD VO VC
CTi CTl CTi CTt
1 1 1 1







0 O O O
o f- r-- VD
CTI M1 in oo
kD CJ^ rH «3
«a- vo r- m
O VD CTi O
in CD CM CD
CM rH rH CM
1 1 1 1










rH CN rn ^J1








CD

^r
o
CM
O3
^^
0
O
o
CM


Dl
S

O
4->
JJ
C
01
4J
in
3

•51
10
03
CU
•H
4-J

in
C
o
"
o
o
o
o
m
in
in
C
O
• H
in
in
• H
E
0)

0)
of baselin
01
4J
10
E
• H
4J
01
01
rH
18
C
O
•H
4-1
IB
C
^^
in
o\
o\
r-t
	

in

*c
CU
W


CO

3
,







01
c
o
-H
in
in
-rl
E
W

U
O
T3
18
01
01
rH
10
W
0)
c
0)
en
10
n

in
D!
C
-H
jj
O
CJ
rH
18
3
4J
CJ
CU
-H
• jr,
— u
t~- M
Ol <
c«
uction in 1
mination of
'O IM
CO 0)
r-l 4-1
01
01 Q
C
o
•H
01 C
en o
•H -rH
E 4J
01 10
IH
T3 O
0) Q
4J rJ
ifl O
e cj
•H
•U C
01 10
El3 >rH
— T3

o p£

,
o ••
01
oi u
01 ,4
e 3
•r| 0
OJ W







01
c
0
•H
en
01
-rH
E


~
CO
Cb
o
o

o.


CO
D

c
0)
rH
rH
U]
*
^
0)
u
c
3
Q
O
4J
C
o
• H
(8
r4
O in
r-l CTl
O rH
CJ

Rob, Radian
November 22

* 'D
n eu
0 4J
en 18
•n Q
^
ifl C
tc o
-H
E 
UH -H
Q
3 01
rD 'O
C rH
Ifl Ifl
M -D
o c
E ifl
0> 4J
S M











C
0
-H
4-J
Q
3
TD
01
tH

01
C
0
•H
en
en
-H
E
Or*
TJ
0)
01
Dl
rH
18
4J

O
4J
C
0
-H
4J
U
3
TD
0)
rH

in
C
0
>rH
01
01
• H
E
(1)
01
C
O
Ol
01
rH
o
1-f
o

0
times rati
4J
0)
D)
rH
IB
4J

0
-H
4-1
U
•o
01
V4
en
C
0
•H
01
01
-rH
E
0)

rH
Ifl
C
o

4J
18
z
JD













































10
10
•o

>,
01
>
rH
3
en

O
01
-H
en
>,
rH
18
ifl

4J
01
O
U
1
u
01
10
0>
rH

E
O
V4
14-1














4J
3
O.
4J
O
4J
en
3
C
-rH

01
C
-H
rH
0)
in
18
Q
0
4J

0
•H
4J
U
3
•d
0)
u
4J
c
18
D1

4J
eu

ndustry mar
•H

14H
O

0
.rH
4J
Ifl
)H
01
01
-H
4J
01
C
0
•H
01
01
-H
E
0)

0)
c
•H
rH
01
01
18

u
























4J
C
4J
in
3
•r-t
•a
Ifl
^
4J
-H
4J
c
Ifl
cr

01
Dl
c
10
0

01
c
0
Dl
01
rH
0
H-
CU
Dl
18
j:
CJ

TD
0)
4J
0)
Dl
M
10
4J
II
01
Dl
ifl
rC
u

01
C
O
-H
m
m
• H
g
Ol

u
0)
2
•D
2-43

-------
2.6   EMPLOYMENT  IMPACTS

      Regulation-induced reductions in  industry  output may lead
to corresponding reductions  in AIM coatings employment.
Employment impacts are estimated by multiplying the baseline
industry employment level  (L0)  by the proportional  change in
industry output  from its baseline level:

                        AL =  (AQ/Q0)  • L0   .                   (2-24)

This  assumes a fixed relationship between output and
employment, at least for the marginal  changes considered here.
                                                              Q Q Q O
      Table 2-11  presents the employment impacts results.  • '
Total employment for SIC 2581 is 51,100 employees.100  The AIM
coatings sector  is a subset  of SIC 2581,  so the AIM coatings
            TABLE 2-11.   ESTIMATED EMPLOYMENT EFFECTS
Regulatory
scenario
1
2
3
4
AIM coatings
output change
(103 L)
-670
-1,177
-1,225
-3,659
AIM coatings
output change3
(% of baseline)
-0.028
-0.050
-0.052
-0.154
AIM imputed
employment change3
(no. employees)
-7.4
-12.9
-13.4
-40.2
    Baseline quantity and employment computations are as follows:
                Industry output   Industry output
                  from Census       from Census
      Sector        (103 gal)          (103 L)       Industry employment
SIC 2581
AIM model
1,229,800
627,723
4,654,793
2,375,932
51,100 from Census
26,083 imputed from
output share
  Sources: U.S. Department of Commerce.  Current Industrial Reports:  Paints and
          Allied Products, 1991. Washington, DC,  U.S. Government Printing
          Office.  1992.
          U.S. Department of Commerce.  1991 Annual Survey of Manufactures:
          Statistics for Industry Groups and Industries.  Washington, DC, U.S.
          Government Printing Office.  1992.
                                 2-44

-------
employment is computed by taking the ratio of AIM coatings
output to SIC 2581 output (same source as above)  and
multiplying this by SIC 2581 employment.  This produced an
estimate of approximately 26,100 employed in the AIM coatings
sector.
     The proportional change in AIM coatings output is
computed by taking the ratio of the change in output from the
market model  (summed across all market segments)  over baseline
AIM coatings output.  This computation is performed for all
four scenarios of the market model.
     Given that the output change estimates in the market
model are relatively small,  it follows that the estimated
employment impacts are also small.  Under the standard
scenario, there is an estimated loss of approximately 7 jobs
nationwide, a 0.03 percent reduction.  Employment losses range
from approximately 7 to 40 jobs under the different scenarios.
The largest effects are associated with the high-cost
scenarios, which has the largest corresponding reduction in
output.

2.7  SUMMARY

     The proposed regulations impose a set of standards for
VOC  content for individual AIM coatings products.  Products
that exceed the limits imposed by these standards must either
be brought into compliance with the limits, have an exceedance
fee assessed on the product's VOC content above the limit, or
be withdrawn from the market.  These compliance actions must
be taken by the producers of the violating products.  This
leads  to a reallocation of resources toward these efforts,
which  imposes opportunity costs directly on the producers and
indirectly on other members of society as producers act,
markets respond, and prices and output change.  The purpose of
the preceding section of this report is to characterize the
reallocation of resources and quantify them in dollar-
                              2-45

-------
denominated terms to provide an assessment of costs and
economic impacts of the proposed regulations.
     Initially, the regulatory impacts are viewed in a very
restrictive light, assuming that reformulation down to the
proposed standards is the only option available to producers.
The aggregate costs of this restrictive option are then
computed for a sizable subset of all AIM producers comprising
the survey for this study to provide a sense of the empirical
magnitude of the proposed regulations.
     The analysis is expanded by progressively shedding the
restrictive assumptions of forced reformulation.  First,  the
exceedance fee option is incorporated, taking into account
that producers may choose to pay an exceedance fee rather than
reformulate if this is a less costly alternative for them.
Then, the least-cost compliance option (fee or reformulation)
is compared with benefit streams (net revenues) to determine
if the least-cost option is also profitable.  If the value of
the benefit stream is less than the cost of compliance, firms
are assumed to remove the products from the market as a best-
response strategy.  The collective effect of some producers
removing unprofitable products and some producers bearing a
per-unit fee on output will contract the aggregate supply
function and lead to changes in market prices and quantities.
The optimal best-response actions and resulting market
outcomes will determine how the welfare costs of the policy
are distributed across producer groups, consumers, and the
government sector.
     Several scenarios are modeled for the proposed standards.
In general, market model results indicate very small change  in
baseline market conditions as a result of the proposed
regulations.   This derives from the empirical expectation that
aggregate costs of the regulation are a small share of
aggregate industry costs.  However, because there is a high
degree of producer heterogeneity within the AIM coatings
sector,  the  costs for  some producers may be empirically large.
                              2-46

-------
Of particular concern is the potential impact of the
regulations on small producers,  which is the topic of
Section 3.
     The analysis does demonstrate the potential for substan-
tial cost savings due to adopting the fee alternative and how
this cost savings is likely to accrue especially to producers
of small volume products.  Moreover,  this cost savings is not
expected to have a significant impact on undercutting aggre-
gate emission reduction targets.  Two fee options were
considered, $0.0045 (1991 dollars) per g/L over the limit and
a rate half that amount  ($0.0022).  The lower fee is projected
to cause a substantial increase in the adoption of the fee
rate alternative with significant cost savings to the affected
producers and a negligible increase in foregone emissions
reduction.   The relatively large cost savings with small
attendant effect on emissions reduction targets may make the
lower fee rate warrant further consideration by EPA.
                             2-47

-------
                            SECTION 3
                 REGULATORY FLEXIBILITY ANALYSIS
3.1  BACKGROUND, AFFECTED ENTITIES,  AND REGULATORY REQUIREMENTS

     The Regulatory Flexibility Act (RFA)  of 1980 requires
special consideration of the effect of federal regulations on
small entities.  The RFA requires that federal agencies consider
whether the regulations they develop will affect small entities
including small governmental jurisdictions,  small businesses, and
small nonprofit organizations.101  Under the  1992 revised EPA
guidelines for implementing the RFA, an initial regulatory
flexibility analysis (IRFA) and a final regulatory flexibility
analysis (FRFA) must be performed for every rule subject to the
Act that will have any economic impact, however small, on any
small entities that are subject to the rule, however few, even
though EPA may not be legally required to do so.  The severity of
the rule on small entities may be measured once the small
entities are defined.
     Small entities may be defined using the criteria prescribed
in the RFA or some other criteria identified by EPA.  The SBA's
general size standard definitions for Standard Industrial
Classification  (SIC) codes is one way to define small businesses.
These size standards are presented either by number of employees
or by annual receipt levels, depending on the SIC code.  For SIC
2851, Paint and Allied Products, the SBA defines small business
as fewer than 500 employees.  Because the coatings manufacturing
                               3-1

-------
industry is not labor intensive,  use of this SBA definition would
result in,  almost all firms in the AIM coatings industry being
classified as small.  Alternatively,  based on input from the
regulatory negotiation process,  EPA has defined small businesses
as having less than $10 million in annual AIM coating sales and
less than $50 million in total annual sales of all products.
     U.S. industries are composed of businesses ranging in size
from small independently owned single-facility firms to large
corporations.  Environmental regulations affect all businesses,
large and small, but small businesses may have special problems
in complying with such regulations.  Therefore, this analysis
specifically addresses the RFA requirements by measuring the
impacts on small entities of regulating AIM coatings
manufacturers.  This analysis focuses on small company impacts.
3.1.1   Potentially Affected Entities
     A regulatory action to reduce VOC emissions from AIM
coatings products will potentially affect the business entities
that own the regulated facilities.  Facilities, or
establishments, comprise a site of land with plant and equipment
that combine inputs  (raw materials, energy, and labor) to produce
outputs  (AIM coatings).  Firms,  or companies, that own these
facilities are legal business entities that have the capacity to
conduct business transactions and make business decisions that
affect the facility.  Figure 3-1 shows the chain of ownership may
be as simple as one facility owned by one company or as complex
as multiple facilities owned by subsidiary companies.
     Potentially affected firms include entities that own
facilities that manufacture AIM coatings.  Determining the total
number of facilities and firms that will be affected by the
regulation is difficult because most of the available Census data
are reported at the  four-digit SIC code, and AIM coatings
manufacturers,  for whom this regulation applies, are a subset of
the entire coatings  industry represented by SIC 2851.  The 1987
Census of Manufactures, Industry Series:   Paint and Allied

                               3-2

-------
      Parent Company
Parent Company
           i
Parent Company
 (Direct Owner)
     Other Companies
      or Legal Entities
           I
        Subsidiary
        Company
       (Direct Owner)
           1
  Subsidiary
  Company
(Direct Owner)
         Facility
    I
   Facility
   Facility
            ABC

                    Figure 3-1.  Chain of ownership.

Products identified 530 companies with shipments of  $100,000 or

more  that manufacture architectural and special purpose

coatings.9'102  Data from the Architectural  and Industrial Surface

Coatings VOC Emissions Inventory Survey  (the  survey)  conducted

for this study were provided  by 116 firms,  36 of which identified

themselves as having under $10  million in  annual net  sales.b'103

While small businesses represent about 31  percent of  the firms in

the survey, a larger share of nonsurveyed  firms appear to fall in
      "These are the two Census categories within SIC 2851 where most of the
AIM coatings products are  represented, and this  figure  includes companies that
produce AIM products, whether  or not it is their primary product.

      bTwelve survey respondents did not  indicate company size.
                                  3-3

-------
the small business  category.0
3.1.2   Regulatory  Requirements
     The proposed regulatory structure imposes a TOS establishing
VOC content limits  for  AIM coating product categories.  The
standards must be met by all products produced after 1997 .
     As discussed in Section 2,  the proposed regulation
constrains firms that produce AIM coatings products over  the
content limits in one of three ways:

         forces them to produce  products with VOC content under
         the established set of  limits,
     •   forces them to withdraw the product from the market,  or
     •   may provide an option to pay a fee on each unit  of
         product they produce that exceeds the limits established
         in the regulation.

Thus, firms with a  heavy (baseline)  concentration of products
above the limit for their respective product categories are more
tightly constrained by  the regulation than those with a lighter
concentration of above-limit products, all else equal.

3.2  ANALYSIS
                                      *
3.2.1   Baseline Market Presence of Small AIM Coatings Producers

     Small business presence in  specific coatings markets
indicates one dimension of how small firms may be affected by  the
regulation.  For certain product markets, small businesses
predominate and thus may be disproportionately affected if limits
are particularly restrictive on those categories.  Table  3-1
     cThe 116 survey respondents comprise about one-fifth of the firms making
AIM coatings products but account for about three-fourths of industry output.
Thus the nonsurveyed firms are relatively numerous but produce relatively
little volume.

                                3-4

-------














Ifl

w
a!
ff.
*?
**
cn
CD
s
H
P_4
fi
O
U
s
h- i



X
EH

s
H
U
s
w
cn

cu

tn
W
3
M
cn

CQ



,J

^•<
^•^
CO


•
J—^

1
ro

W
•J
§
£•*



























-
ij

QJ
"D
C
O
QJ
1-*

n
—
cn
U]
0)
c
• »H
W
J3
_,
2
0"j










01
E ^j
.E O
n 3 --
2 T3 J
? 0
11 0,"°

HI j_, —
< a
° rH ^
(D ^ ~ —

i« O X — '
nj ^j ^

u) a
a "
^3 ifc-i -^
^ ^ O
2 ti
a
"*" *
*s s ^
4J ?• —1
3 ? v>
0- 5,
M £"0
"^
c
e
O _j "
- c
 ^

ir- C r' ' — '

5 i
C
tr.
IL-,
^ 0
tfi —




c
0
(C W
o c
a  a


^ ^
3

rH
i-H rH
(0 —
O
E-1







&i o3
10 5 >J 3
h 3 (D Ti-
ll 7! a o o
> £ ^ "!
< •* c. —

Ul
U XJ
(D U
6 O "D
2 *-<
a
o
T3 w

j-i 3 7-, ~*
-3 C c^
Q, 11 C ^
E > -1
M 1* *^
U

J
U]
11^
r-1 O
IB —I
t/3 —




0 >,
jj n
(0 M
rH d)

QJ (J




•*-> u
Q) ID _^


ID D 2
S W ~






<
Z




0

0
0
rH



rH




rt;
-*.
2



0
o
o






OS
2



2


rH




,«£

2









a
2





U)
i-l
-V
fO
0)
,Q

-o
c
o
ca


CN
rH





O OCN rHf-- r^i^ cn

O fNCN U3O On UD
mm m rn fN ro cn
CT>





o r-^1 r-^r >x>o o

O OV£> <>DrH rHCN in
CM rH CN m CN rH CN


0

o rHt"- fNin r^co o
co co cn m m p- CN



o mcN ^ro CNO vo
rH cn n ^D co co ^
m X> CO rH UD CO O
rH in

o cn co rH in co m
*C i— (O r-* ro fNfN CN
"~" fN





,—, ^CN oin On m
rH o rN UD co o cn
£-. CNCO ror^ r^'.D
^ * cn
<3< CN CO ^ CN CN rH


*^> co r~ rH m r- cn
^ ^co cccn r^irH co
« CN o^ ^r o r~- ^ *-^
<~lin LTlm rHrH f"1

rH rH
fN rHCn ^3*^ ^*fN O
m in rn ry H3 v£)



< fNr- cno ^ro m
~^. ocn cncn rnrN co
2 mo r~»cN oco ^

no n n co co ^
r^ o r^ co vo I-H CN
cn r- T n fN ro (N

co VD ** in o
cn co ^o o in CN in
*3- r- r- r- co <-H n
^r tn
Di O^D COCN Ch- ^
2 ^rH COLO OCN cn
^T n rH rH «H Cn CO


V)
Di V-* ^
coo n
•H -H -H t>
JJ V-i VH (L)C
03 0) (U rH U -H
Q) O AJ AJ CO C AJ
AJ(JC -MX ->H -H(OWfO
• H -H AJO 
in






o
in
^"
*3<


0
in
CN
rn

fN
n




CO

**

CO
cn
rH


rH
in
O

cn
I**




-o
U] IQ
D
O W
C Dl W
-H C U
E -H -«H
3 AJ AJ
iJ (0 W
•H O CO
co o e


fN

rH




^r in

ri n
n






CO CO

CO CO





CO rH
rH rH



rH n
0 O
\& rH
^


rH rH
rH O







o °^
\£> ^


co n.
co r^
in r^
tN n


in in
0 CN
CN rH



r-i ^O

rH <3"

rH l/l
^ CN
rH rH


r- co

0 rH

rn r-
U"l ^f



0)
1
d a
m O
P
- (TJ
W V«| Ul
c AJ q
-H .H -H
•3 g rO
JJ QJ AJ


X CO

r- r*




o r- rN

Cn fN U3
\£J CN M3
fN





c** r^ n

^ cn CN
CM rH CN




rH ^T CD
fN rH ^T



tN ^O fN
cn ^» in
rH CO ^3-
CO CO


O Cn rH
fN O CN
rH rH





cn r-
^ r-
\O CO rH
» i-H »
in n n


vo r** CD
t-H VD rH
in CN CN
in in rH


in rH in
CO r- rH
CN



in ^ cn
co ^o ty
cn ^ vo

r** cn cn
vo cn ^o



\D n : cai
JH -H rH
(D UH O
E O W
O - 


^ CO CO





—
T,.
C
j^
c

0-
C
O
o


























































.
j*
*»
s
^4
Q

•o
c


4J
Q
n
Q
o
C

&
HI
CO
3-5

-------








Q
W

J3

EH
O
U

«
W

si
W
o
H
EH
0
u
s
M

^t

W
K
EH

a
H
W
O
£5
W
W
0,

C/}
C/}

r^
M
r«
D
m
Jq
*i


cn
rH
i
PO

U
j
m
EH











w
^

•c
c
o
HI
W

VC
"~*
U)
H
41
C
W
,-,
10
£
t/j-














c
0
•rl i
Jj —
10 M
rH *J
3 c
a  -I
QJ W
> U

3
0) ^>
'H
rH »~<
TO -"
U
o















u
£|*3
10 g O
v "o a°
* " £-
a
i" in
O ,_, u
« u
S * ^*
o £ o ~
£ 1-
w a

o u
J2 >*j -
e o -o
3 0
2 ti
a
•a ™ £

If*
M P 0
W rH
0)
IfcJ w
0 In v
E
<" 3 «
^ (—1
O
^
~
vn *~^
 a —

^i jj
HI O
"e 'o -o
3 O
2 ^
a

0 •"*
•o w S
d) 11) l— ,

3 c v>
Q, 0)
e >-
tH 0) 0

—
W ^

rH O
10 rH



c
o >
u O
10 Dl
rH 0)
3 J->
Di (0
01 U
a.
jj r" ^
1 if
tc QJ r:
S w G

co m (N en o
r~ v^> [•-- co o
T rn rn




rH "31 O O O
CT\ CN o en o
rn r-* CN



in (N IH



CO TJ CO CN O
en VD rH r-i
[^ m rH en
en c^


co CO m m o
o o in CN o
r-* CN



O ^j
CTi CO
m in m o o
«g" -in
CN iH ^T m


«H ^" cn co r~\
r*~ o vo in o
rH rH fN rH O
rH


co VD in o in
CN CO ^D O rH
rH rH rH


CO ^" O CD f-i
^D r^ cn en ^3"
CO O3 O CO CO

O CO VD O -H
O"i ^ n ^3" OO

m f*i rH in ^
vo ro in r- ^>
rH (N m m CD

CN cn r^- tn in



to
0)
U TJ-H
T) C O J->
U Q) W 03 Ul O CQ
-HiJDi £ v-t w 3>
rH C C M 0) l-f ^
rHClt-H |Q 3  AJ O
W r-l - CO O -
o c ^ u a v-i
y-i -H ,y a) v-i MO)
AJ U 6 Q> OJrH
>, CC -H -H T3 AJ (0
Ui O 3 H C CD 0)
Q U O Q43 3 W

« * "
1-1 in r-


o in r^ o
o cn m o
CN




o m co in
0 ^ m vli
rH IN rH


O
O t^ O T
rH rH


O
O CN ^ O
U) ff\ c^
rH O 00
ai


O CTi C^ 'J
O O 0 T



o r-

O VO UD O
V£> - 00
(N CM


CvJ O ^3" IN
co in ri in
CN rH rH



(N m fN \Tt



&\ ro O rn
O rH ro fN
cn m PJ m

rv] c^ rs) o
rH rH OJ IN


IN \o m o
o cn *r a\

-


V
m ^
ft) W 3
> M Jj
a -H ai x
ui u u m >
c ra 10 w ui ^
10 > o ai •o «
H t4 U O C rH
U 0) IH -H .H .* QJ
•HT30J 01 4JJJ UE
EOO) T3 wiO -nra
OJOM c 100 3c
W3ft S SO OH)

0 "1 m ro
1-1 in ^


o o
o o





o o
0 0



0 0
o o


CD O
o o




0 0
0 0



o o

0 0




c^ r-
t~ rH
u-i r-



rH CO
r-H


r- r-j

o m

OJ *!
rH


r~ ^J

CA O\
-

r4
10
HI 01
rH 11) 0)
0 C T! >
U O'rt
O O 4J
M ma s ro
O 3 JJ >
10 crc ai u
rH 10 H> 3 i1
rH a > CT01
0) OrH 10 01
ff. O Q< ^
cn j; w o a

o

—
QJ
C
TJ
0
u


















































V
n
a

UJ
Q
•o
c
10
10
^J
o
c
(U
01
U)
3-6

-------






____ ^
Q
W
^
2
M
EH
O
0
tH
W
H^



^g^

CO
0
n


in
to
0>
c

U)
3
JJ

,
e













c
O
1C w

"3 c
0 T3

O
>, O.
01 jjl
5) *o
'"'
10 —
o
E-






QJ - J-* —
tT> = U J
t i-i D
^ ,£ CJ T3-
O - C. O °
> ^ U «~<
< a
'— CO
c ^ t*
^ /i 3 -j
S ° 0
r- J-1 ^
OQ a
U)
3 u
ja iu 3
£ 0 Tl
3 0
2 ^
Q.
T3 W cr!
4) 2 m
^ C *H
E "•
M " 0



Ul
t! -S
m £ 	 .

r^ " O
M 0


fl ""*
^ o
1C rt
0) —

& ^ U J
1C § ^ 3
k< _- 41 tJ-
0) "^ D. 0 0
> ^ 1-4 *"*
< a -"
W
VJ 4J
QJ U
n y_j "j
1 O 73
3 O
2 *-<
DJ


" rt
0) 4| ^
3 C '^
cx ^
M 41*0
^1 »H


tn ^
OJ-
r-( 0
CO rH
CO —



C
0 >,
•u o
1C Dl
.-< OJ
W 10
0) U
en

u
III
oD fM i— ( O O i-Hin
rH rH u3fNLn cNr-r^


o o o o o o o
coor—i OO^^TO o\ mvc
CN T-I

oo^c^o ^ "^
CN 2 o in <^> *H







< „ ^ ^ OO Or*")r'"iCN >c -i 01 o i- us r- u; r~ rt




*-DmvjD *& *f CQ \D r^ m in CM
?3fN S^SSS S CM"^

CN rH rH rH


rH W 01
0) W iJWOTJO) O-H
W iJ CV^O C> 01 OJ->
O WCDWtO EE O'*-1 3 (rjtw
rH W Q> *D rH *O U} J-) - *-iC
-H OJ -H -H V4 C O O £ -H H E -H J 0) 0)-H t-f O»*H
t3rH iJjJ U'Hfi E£X Q*'*-' AJ£ C*-" O'OW .C Q< J-* 3 *«H ^J
CfO rHfQ C^nE *H£ CO(D Qjy) -
1— t
CQ
i| j
o
•D
C
0>
1C
m
41
O
C
01
01
en
3-7

-------
Q
W
O
U
H
«
OS
 CO

 §
 O
 u

 g
 H
 H
 EC
 EH
 U
 a
 w
 &

 CO
 CO
  J
  •J
  CO
  m
  W
  J

  I
c
10
X!
JJ
tn
in
0)
rH

M
O

W
JJ
o
3
T3
o
JH
ft

in
Dl
c
•H
JJ
(0
o
u

52
M
<

c
-rj

c
O
• H
t — 1
rH
•H
E

O
rH


c
<0
£
JJ

in
tn
0)
rH

DI
g
•rH
O
3
T>
O
u
a

tn
(0

73
0>
c
•rH
4H
0
•0
4)
In
(0
tn
(U
in
tn
0)
c
•H
in
3
.5

rH
rH

C
W

ro







































































U
0)
rH
10
tn
rH
to
jj
O
jj
c
•H

c
o
•H
rH
rH
•H


O
in




C
•H
C
o
•rH
rH
rH
•rH
6

0
rH


M
0)
13

3

Dl
C
•H
>
(0
XI

tn
(0

tn 01
O) N
> -H
rH t/1
01
in >,
E c
01 (0
xi a
jj e
o
13 U
0)
•H JJ
MH tH
•rl O
•u a
C 0
0) tH
T)
•H JJ
o
Oi C
in
O 13
rG -H
JJ 15

m in
O JJ
C
vo 0)
m "D
C
T) O
c a
iO tn
0>
tn tn
JJ
C >i
01 0)
•0 >
C M
O 3
ft in
in
0) 0)
tH >
rH
vo 01
rH %
rH &
tn
(0 •
XI tn
0)
>rH
01 (0
> tn
>H
3 rH
in 10
3
0) C
XI C
EH (0

.a

01
XI
JJ
c
4)
x:
3

T3
0)
tn
3

tn
•rH

0)
Dl
10
tH
0)
>
(0

•a
01
jj
Xi
D)
•H
0)
3

<



CN

C
O
• H
JJ
U
0)
w

E
0
tH
V4H

c
Q)
J4
10
JJ

in
0)
u
-rH
rH
a

O)
Dl
(0
tH
0)
>
(0

Dl
C
•rl
in
3
13
01
JJ
3
I
•rH

Q)
tH
(0

in
0)
3

0)
>
0)
Oi

u


















































in
jj
C
0)

O>
0)
in

jj
QJ
^
tH
(0
E

O
3
jj

C
•rH

in
D)
c
o
rH
4)
X)
rT
O
Dl
41
JJ
<0
U

jj
U
3
13
O
tH
ft



0)
XI
01 JJ
in tn =
•H (0 C C •
•H 3 tn
- XI O -H
j u i; c in
•H O) .* >
r- x; tn c rH
r- 3 3 3 m
vo = C
- - > 10
in in jJ in
m rH -n O JJ
O) JJ 01
- E C = X
01 10 10 JJ M
IH c 3 a 10
4) 0) D1 E B
XI 0)
>, >, X 41
0) >H O) 01 X!
U 13 > = JJ
C in
0) ,X 3 in tH
tH U W 0) O
01 -H XI tn
H 0) C
m C D) 3 "->
in -H
00 JJ >i JJ JJ
- (0 rH (0 O
rH O JJ XI C
u x; jj
in DI a>
•H JJ -H >i >H
C rH QJ 0)
>i 0) in > 3
0) E tH
> 01 in 3 ~
tH U -H W 0)
3 C
tn 01 0) 0) tn
jj tn x: o
oi -H o) jj xi
x: in x: M
JJ 0) CO)
C "O -H JJ
C Dl 0) iO
•rH (0 jj tn 3
E -H 01
T! O -H tn
0 13 ft JJ O
JJ C 0) --H
tH (0 tH JJ 4)
o c c
a tn >, 
rH Xi tT 10 rH
o u o
> 13 4) 0 W
c xi xi
rH O EH O
iO XI in JJ
JJ -H
o o • tn JJ
JJ JJ T! >i U
0 rH 0
TH TJ jj 
•H
T!
c
•H

<4~l
o

0
tH
3
in
o
rH
u
in
• H
TJ

o
JJ

0
T3
T3
0
jj
tH
O
a
0
tH

jj
o
z

II

OS
s














































































0
rH
XI
(0
rH
•rH
10
>
10

JJ
0
z

ii

<
\
s
3
0
rG
JJ
W rC
C JJ
O -H
-H 3
in
in C
•rH O
E -H
W JJ
(0
U tH
O 0
> a
0
in o
Dl U
C
•H C
JJ - rH
C (0
rH (0 C
(0 -H
• rH JJ Itr
tH C
Jj -H
in (0 •
3 Cu in
13 3
C rH U
H (0 3
C 10
13 0 U
C -H
(0 JJ >i
(0 tH
rH 2 JJ
(0 in
tH tH 3
3 O t?
JJ MH C
U H
0 T)
JJ 0 C
•rH tH O
rG (0 -H
U ft JJ
tH 0 (0
< tH -H
cm JJ
o
Dl
in • 0
JJ >i 2
j- 0)
01 > >i
•H M tH
W P O
c cn jj
H (0
>T-I
>1 tH 3
tH O Ol
JJ JJ 0
w C OS
3 0
13 > S
C C H
H H <


0
U
tH
3
o
W
-8

-------
lists the coatings product categories provided in the survey.104
The survey data represent producers that account for
approximately three-quarters total industry product volume.d
     Small producers produce more than 20 percent of the products
in the survey, but these products account for just 3.6 percent of
total coatings volume and 3.7 percent of total revenue.  The
average price per product in the small business segment is
$2.52/L, compared to $2.44/L for the industry.  The largest
volume category for small producers is roof coatings, at 19.9
million L/yr.  Small producers comprise just over 22 percent of
the volume in that category.  Small businesses produce over 95
percent of the total volume of antigraffiti coatings, but the
volume is quite low, with six products totaling about 40,060 L.
     Other categories in which small producers comprise more than
20 percent of the market volume are lacquers, mastic texture
coatings, graphic arts coatings, bond breakers, and
appurtenances.  In addition to roof coatings, small producers
collectively produce over 4 million L in the following
categories:  traffic marking paints, exterior nonflats,
bituminous coatings, lacquers, and interior flats.
     The extent to which small businesses are affected by the
regulation to reduce VOC emissions from AIM coatings will depend
partly on the average VOC content of small business products
relative to the industry average.  Table 3-2 presents the average
baseline VOC content for products manufactured by small busi-
nesses as compared with those manufactured by the industry as a
whole.105  Small business products generate approximately 6.2
percent of total VOC emissions in the survey, which is substan-
tially greater than their output share.  The average VOC content
for small business products, 325 g/L, is almost 75 percent higher
          is  based on the ratio of Census product volume  (part of the total
SIC 2851 volume)  to the survey product volume.
                                3-9

-------
                 TABLE 3-2.   BASELINE VOC CONTENT
Size
category*
All products
Small business
products
VOC emissions
(Mg)
344,059
21,431
Sales
(kL)
1,853, 623
65,914
Average VOC
content
(g/L)
186
325
  a  The survey has 116  respondents and 36 of those identified themselves as
    having under $10 million in annual sales.  Twelve  survey respondents
    did not report company size.

  Source: Industry Insights.  Architectural and Industrial Maintenance
         Surface Coatings VOC Emissions Inventory Survey.  Prepared for
         National Paint and Coatings Association in cooperation with the
         AIM Regulatory Negotiation Industry Caucus.  Final Draft Report.
         1993.
than the  average VOC content for  the industry, 186 g/L.   This
suggests  the  potential for disproportionate impacts on  small
businesses  unless the regulatory  approach were designed to
mitigate  these impacts.
     Small  business products have a higher VOC content  than the
industry  average for two possible reasons.  First, small
businesses  specialize in products that tend to be higher in VOCs
because of  fundamental performance requirements of the  products.
Second, small businesses tend  to  produce higher VOC-content
products  regardless of the product category.  The first reason
can be  termed a specialization effect and the second  reason a
.technology  effect.
     The  observed difference in average VOC content of  small
businesses  and all products as separated into the specialization
and technology effects using a simple procedure.  First, a
measure of  the projected average  VOC content of small business
products  is computed based entirely on the distribution of small
                                3-10

-------
business products among product  groups  with different average VOC
contents.  This  is  a measure  of  its  specialization-based VOC
content:
                    N
              Vs =  £ V^ • SiB .                               (3-D
                   i=l

Here, V^  is  the  average  VOC content  for all products in product
category I,  StB  is  the  share of total small business product
quantity attributable  to product category I,  and N is the total
number  of product categories.6   The  separation of  the average VOC
content difference  into  the two  component effects derives from
the following equation:

      (VB - V1)     =     (VB - Vs)    +    (Vs - V1)                (3-2)
     Difference   =     Technology +  Specialization
     in Average        Effect            Effect
     Content

Vs and V1  are,  respectively, the small business and industry-wide
VOC content  averages.  The specialization term  quantifies the
difference between  the specialization-adjusted average for small
businesses and the  overall industry  average;  the technology term
quantifies the difference  between the specialization-adjusted
average and  the  actual average VOC content for small businesses.
     Table 3-3 yields  the  computation of the Vs  measure for the
small business products  in the survey.106  The computed Vs  value
is 261, meaning  that one would expect an average VOC content of
261 g/L for  the  small  business sector,  based purely on the way
their products are  distributed among product groups  (i.e., their
specialization).  Placing  this value into Equation (3-2), along
with the values  for VB and  V1 given above  (325 and 186), the
breakdown  is computed  as  follows:
     estB is not the small business share of total production in category I,
but rather the contribution of category I to total small business production.
                               3-11

-------
TABLE 3-3.
SPECIALIZATION-BASED AVERAGE VOC CONTENT:
   SMALL  BUSINESS PRODUCTS'"
Market
segment
number
12
12
1,2
11
1,2
1,2
9
3,4
1,2
7,8
3,4
5,6
13
13
12
7,8
7,8
7,8
3
12
7
13
1,2

Regulation
category
Bond breakers
Concrete curing
compounds
Roof coatings
Traffic marking
paints
Nonf lat,
exterior
Bituminous
coatings and
mastics
Lacquers
Flat, interior
Flat, exterior
Varnishes
Nonf lat ,
interior
Primers
Mastic texture
coatings
Industrial
maintenance
coatings
Metallic
pigmented
coatings
Stains ,
semi transparent
Sealers
Waterproofing
sealers, clear
Quick dry
enamels
Graphic arts
coatings
Shellacs, clear
Sc opaque
solventborne
Apurtenances
High performance

All
products
average VOC
(g/L)
N/A
621
239
369
173
23
657
52
79
474
134
172
146
374
459
475
312
632
461
366
539
411
335

Share of total
small business
volume
N/A
N/A
0.3025
0.0857
0.0723
0.0675
0.0665
0.0639
0.0504
0.0482
0.0425
0.0422
0.0400
0.0395
0.0363
0.0091
0.0053
0.0048
0.0042
0.0038
0.0032
0.0030
0.0022

Share -weighted
content factor
(q/L)
N/A
N/A
72.20
31.66
12.49
1.54
43.72
3.30
3.99
22.84
5.71
7.23
5.85
14.78
16.66
4.34
1.66
3.05
1.96
1.40
1.72
1.25
0.74
(continued)
                         3-12

-------
      TABLE  3-3.    SPECIALIZATION-BASED  AVERAGE VOC  CONTENT:
                 SMALL BUSINESS  PRODUCTS3  (CONTINUED)
Market
segment
number
12
13
5,6
12
12
7,8
7,8
12
13
10

10

10
10

12
12

13

13
5,6

Regulation
category
Swimming pool
coatings
Sanding sealers
Undercoaters
Dry fog coatings
Antigraf f iti
coatings
Stains, opaque
Waterproofing
sealers , opaque
Pre treatment
wash primers
High-
temperature
coatings
Eelow ground
wood
preservatives
Clear wood
preservatives
Opaque wood
preservatives
Semi transparent
wood
preservatives
Form release
compounds
Multicolor
coatings
Fire-res is tent/r
etardant
coatings
Magnesite cement
coatings
Quick dry
primers ,
undercoaters
Sums / averages
All
products
average VOC
(g/L)
552
525
206
300
397
257
239
706
561
541

419

362
548

599
321

16

N/A
439

Share of total
small business
volume
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

0.

0.
0.

0.
0.

0.

0.
0.
1.
0019
0012
0010
0010
0006
0006
0003
0002
0001
0000

0000

0000
0000

0000
0000

0000

0000
0000
0000
Share -weigh ted
content factor
(g/L)
1.06
0.64
0.21
0.29
0.24
0.15
0.06
0.12
0.04
0.00

0.00

OfOO
0.00

0.00
0.00

0.00

N/A
0.00
260. 87b
° Small businesses are defined as producing less  than $10 million in AIM coatings
  products or less than $50 million in total sales.
  Specialized average VOC content equals the sum  of share-weighted content
  factors.
N/A = Not available
Source:   Industry Insights.  Architectural and Industrial Maintenance Surface
          Coatings VOC Emissions Inventory Survey.  Prepared for National Paint
          and Coatings Association in cooperation with the AIM Regulatory
          Negotiation Industry Caucus.   Final Draft Report.  1993.
                                      3-13

-------
               (VB - V1)   =  (VB  -  Vs)  +  (Vs - V1)
               (325-186)  =  (325-261)  +  (261-186)
                139      =64      +75

     Approximately 54 percent of the 139 g/L difference between
the small business sector VOC content average and the industry-
wide average can be attributed to greater specialization in high
VOC product categories (specialization effect),  and the remaining
46 percent can be attributed to the disproportionate presence of
small business producers in the higher-VOC products of the
respective product categories  (technology effect).
     This finding has implications for the feasibility of
designing a TOS to minimize small business impacts.  Since small
business producers are somewhat concentrated in the higher VOC
categories, as indicated by the empirically sizable
specialization effect, the TOS can be designed to be somewhat
less restrictive in categories with high small business presence.
However, the effectiveness of such an approach in mitigating
small business impacts will be limited by the fact that small
business producers are also concentrated in the high-VOC range of
each product category.
     In 1993, the National Paint and Coatings Association
performed an analysis of the then-current 1997 Table of Standards
(TOS) and found that  the projected emissions reduction from the
small business sector would be 19.65 percent of baseline
emissions, compared to a projected 25 percent reduction for the
industry.107  This estimate provides some evidence of relief for
small business products under  the standards proposed at the time.
Unfortunately, data were not available to recompute these
estimates based on the current proposed TOS to see whether the
proportional reduction from the  small business sector is still
less than the overall  target of  20 percent.

                               3-14

-------
3.2.2   Costs Associated With Regulatory Compliance
     As discussed in Section 2, compliance options include
product reformulation and the optional payment of an exceedance
fee.  Depending on the technical difficulty to reformulate  the
current formula, the reformulation may be considered "major"  or
"minor."  A major reformulation requires the following level  of
effort:
         2 to 3 scientist years and
         1.5 to 5.5 years elapsed time
at an estimated cost of $250,000 per reformulation over 3
years.108  This level of effort was converted to an annualized
cost of $17,772 (in 1991 dollars)/  Fee payments made for
products that exceed the limit is an alternative compliance
mechanism under consideration for the proposed rule.  The per-
unit fee is computed as follows:

              fee =  (VOC content  - VOC limit) • rate.         (3-3)

VOC content is measured in grams per liter, and the  fee rate  is
paid on the grams per liter in excess of the limit.   The proposed
fee rate is $0.005 per excess g/L  (1995 dollars).  Total fee
payment per product simply equals the per-liter fee  multiplied by
total liters  of production.
3.2.3   Reformulation Cost Impact Estimates
     Given the data from the survey and the 1997 TOS limits,  the
number of products produced by small businesses that exceed the
proposed VOC  limits are identified.  The number of potential
reformulations is estimated by applying the TOS limits from
Section 2 to  the number of products reported by category and  VOC
content in the survey to determine the number exceeding the limit
for each category.  Results are  reported in Table 3-4. An
     fDetails  of the derivation of these estimates are presented in Section 2
of this report.

                               3-15

-------














It!
P-


rH
-
CO
o
o

s
o
H
EH
J
B
2
o
W
a
CO
CO
u

M
CO


m

t "i


f^i
BE!
CO


*tf
1
m

W
J
f*4
<
EH













-- M
o e c "
~ -H C "^
U3 "4-1 ^ ^ ~—
O (U J~ ?
O l-< -j ^,

U-l 1 ~-.
o E c
^i O '~~
AJ O-H tO
O 01 ra
U U--H
3
«>" i CTC 2
N o E C-H 7;
•H 5 .H ra Si ~
rH j_) Q jj ^| •* vy
nj (o IM fl JJ o *~*
3 O 0>rH W g
i°*3§a
< 0
"S^o
'ra^'o ~
M ™ n*
In O ^^
C-O '
5 ° *
O,^
•0
O C ^j
* ra 3
^ ij'D
E in 2
28
O -U -H
U 3 ^-H
0) T3 > f-l
"E ^t ® o
3 a o




0 3
3 jr v>


£ ">
Kl

£
Oi CD -
"-H ^
0) W V>
> a~-
<


CO
01 —
ra ^"
CO



0 >
•H It
*J O
(8 Oi
r-l CD
3 4J
CD O


4J aJ lj
JSp
11 =
Z yj C

CO
o
O O





m
o on
CO
CN
rH
in
CN
o o o

in


0 0 m
0 0 rn
^*





0 00
o o cn




o o rn
rH



O

O
**• ** \o
*7 7 ^*
*• ^0

0
in

0 O rH
rH rH 1/1
•* ** CN


cn
vo
^. -^ co
Z 2m

cn
rH
Cn
•H Ul
U] tH O>
kn 3 C
0) U -H
X. W aJ
(^ 0*0 *fl
01 J-J C O
w a; 3 u
£) ^ O
-o u a •"
C C E 0
O 0 O O
m o u ct

rN (N M



cn TT M
CN r^ o
rH






&
u
fl


cn



r-
m







0
in

^
in
CN
O

in
rH


[^
O
rH





O
cn




m
rH



cn

o
m
rH
m

cn


rH
CN
CN


in
0
**~
rH
CN
^P

^
O
•H
i-l
0)
4J
C
•H
4J
(0
rH
[lj

*»
m


rH m OP*
CN CN rH m
rH r-H ^H p-i





O *»
r*- n O «-H
in rH \£> £~*
O O -
PO - - ^D
PO in cn rH
in r- o co
m co in cn
oo 0*1 o cn

CN o co in
o cn o m


»x> co cn TT
«3" m cn co
CN ^T rH m





o o o
n CN t**- ro
(N CN rH in



^ rn vo o\
mm fN r-



r«- vD CTl ^

CO U? VD O
m rH rH CN
r-i in cn co
m ^" rH f"-

CO CO  VJ C tJ) -H jJ O>
CJC UOJC rnlCC
•H -H OJ JJ -H i— ( Q) -«H
jj jj 3 C J-> rtl E *->
W(0 *D'H(0 jJ&iHJ
(DO CtOO (U-HO
zu M£U SE&U

mm (N
rH rH (H



^D CO
m <£>
fN






P"t
CO O
m
rH

r- cn


(N pn
^J CN
rH


9 CO
«7 -?
 in
CN rH
rH CN
o m
v0 cn

r-,


vO VD
^O VO
CN CN


co *r
r- m
cn ^
-H 0
o in
vo m


jj
C
ft)
Wi
a
01
c
- (0 01
W rH lH
C JJ 0)

-------










^^
Q
P
2

E-
2
0
Si

10
r-

cn

1 '


**
CO
EH
CO
0
CJ

2
O
H


<
cJ
5
§
o
(X

CO
CO
w
*2
H
CO
D
PQ

1
CO


•
^^*
1
m
W

«

^H









° M 2~
jJ O "— ^ dP

0 £ ~ ^
— ^i

y-i . ~-
DEC
w, O ~
JJ O -H w I Oi C 2

H •*•* C « *—•
•H jj Q jj ^j J ^/J.
3 o "a) ^ 1r O
C U U 3 C £
< 8
*I? w
*C , AJ
 -H
M 3 bl -H
0) TJ S^-H
|So8


"81
AJ 3 —
E ^*
1-1 2


'"o
g1 00
P~


a
•H J
« —




0>

xJ O
<0 Dl

3 JJ
DI (0
o o
IX



» §1
S * C

vo o « ^.HC ^^^(N ^^ tn o o
(N M cri cocTio ^r.-ir-t^ cnm CN o o
i— i i— i n~i r^jtN ^ino
•H


^ co co ocoo in *? r*i o ^CN o o o
f*"l C*~l f*l r*lO~iO i-HlTliHrO CO^> CN O O

O O O OOO OOrHO rHrH ^" O O



ro m in or* co^coco cove oj

*<& ^ r^ (Ni-i_ co m o co OVD m _ o
vo VD ^ cr\oi »H in «— t «-^ rH m r*-
o o cr* in^* «Hrnr*«H r*-c^ ^*
»H r-< rH




co r*~ o r*coo rovDrHr-t ^ r** r*" o o
CN \XJ 0 ^0*3-0 C0cdr-*t-i VOVO U3 0 0











'•O VD LO OrnCC Or-^rN)^1 r-t^" i— ( PI O







en en co ompj o»Hf*ivo I-HVC c^ *d* o

CN O \D OO^* Ot^^^D CO VD rH ^ •— I
O ID O Or"l*? r^-COCOVD *f lT\ Cn t-* (N
o o in i-HVDco oinoo (NCO ^r (N ^H
»x> o o m t-t n co^ojcn ^"(N ^ *X) in
CO CD O v£)^f"i inr>J t— ( r*~ in in r-* (N co vo VD c^ r^ CD
o oo n co «3" en m rH **J «H o VD p* CN PO
oo en rH oo^1 vDo*iD m o co VD »-H ^
rH r^ in t-Ho^1 r^covD vo ^ rn *H t-n
m rg fN CN (N rH .-H

0) m V
In U W O> , 10 -DC**-- WJJU*M OO EEa

hW *OO) UDl U3a-> C OJ CD) DI O Ol (OD> •» i-iW RJ^-t 
a^ rH-HC focrc CD a-nc c u o v^c w 0.1-1 aja-uc
1) ^H U E a •iJ rH G< J> V-< JZ E*J t3 U -r^ •*"* "*^ I' "^ *^ *G £ ^J
J-J (0 *H nj (0 <0 QJ O •— < 3 D^ -H (Q C *O ^i A-1 (0 fO JJ (0 QJ tfl &m
too 3c MO x: o a-H 30 to c M co jj (0
-------
II
•Sic 8
0 £ 0 3
11 W , — • — •
"•251-
S2-*
U " 3 Ul
II
11 J
"o EC:
II Ui O '~~

H Oi >AJ AJ —
II o 01 ro
| O UlrH
Q =
W Mr, "°
r^ T3 0) „,
t£ oiu-i i oic rj
\2 1 N O E C -rl
fj 11 ~-" U-rl «J " —
~ [I rHAJOAJUl^;
F-i  ™
10 u ^ u
f- ?• o S,
^ -3-1
2 3s2*
rH M °
£3
II T3
- 5 ° m
S aw
E-1
CO „ -o
O •" <£ 01
u °5£
^2 =
2 -o^o
0 eg 2
H 2§D
P II f I
E-1 u
<
J M-l 0) AJ
M O AJ -n
B II u ui E

§ U •§ £ rt
ffi jg o S
o E r. ° u
^ 3 a o
ClH Z >
N
&
•o S
CO v%
_ -. (1 i) J ^^
c/) 3 crx
S l>~
* *£
M ^
cn
r-. I)
1—* II «
m 8," 3
, * u J
^ »'£«
iJ > a-
rf! [1 ^

W
11
l| W
*3" tn ~
II
ro
w
"J
9 1 §^
<3 || -H Ul
^ 1 *&
r-l 01
1 3 ^
I Ol ID
1 «• °
a

AJ e ^
II *•* C (1
0) OJ ,
It Ci '
(0 01
S W

oooo ooo o o
oooo ooo o o



oooo ooo o o
oooo ooo o o

oooo ooo o o





oooo ooo o o



oooo ooo o o
oooo ooo o o








oooo oo oo o





oooo oo oo o






oooo oo oo o




mi/i intn oo mm ^
TJ^ ry^r fH r-i ^" ^ <*







oooo oo oo o


AJ
jj ^ C
C AJ QJ
W W W (U y) C E M
•COJOJOJ^OJat tO Q) W
C>>T3>*D>W AJ U -0
3 -H *D -H O -H a -H (0 ^ W C
OAJ OAJ OAJ COAJ OJU) O -HAJ (U >i3
^ m om 3m c ic --co «-• w wcw AJW u^
& > S> > S > OJC 0 0» 0) at v^ai 3d) AJ 
!-H

CJ
in
(N



CD
C\)
^

^«
rH
a\
m
V£>







C
O
•r-*
(0
^H 3
ro E
o AJ o
4J (1) «4-t
^. OJ
C Ul Ul

E Ui Ui
c o o
O O •!-> -n
-n 3 10 IB
rH AJ E 6
•n C O -O
E 'rl Ol 0)
Ul N
O 01 01 -H
in Ol 'O rH
 c c ro
O 3 3
C rH C
ro oi >t c
£ J2 rH 10
AJ r-l
>* ID 01
01 ui -H s:
Ol O AJ AJ
01 Oi C
rH 01 01 >*
AJ AJ X3
Ui ID O
o u a TJ
01
Ul AJ AJ -H
AJ U 10 rH
u 3 JH a
3 -O AJ -H
•DO AJ
O Ul C rH
ui a o 3
a -r. e
QJ AJ
01 £ Ul C
Oi AJ O O
c a -r<
-rH C AJ
AJ 11 0) 10
TO £ JS rH
O 3 AJ 3
0 E
•O AJ Ul
E 11 C O
M 01 01 y-i
< 3 03 1)
0) Ui
C 0) Ul
•H -H P< CT
QJ C
C QJ lu -H
O 01 U
•T-< (Q 0 <3
r-l !M xJ «AJ •
•— i o; —
•H > U) W fN
E ro *0 AJ
U U C
O *O -«-t 3 O
rH QJ X! *O'-»
 AJ AJ O AJ
.C i JH 0
C Di O Q-OJ
10 -H ? W
J^ QJ U H-t
AJ 3 0 C
>i -^
W fQ JJ *-<
W • Q) W
QJ *O *O *O J2 •"*
rH C 01 01 E-H
to -H T? 3 ro
Ol rH C C AJ
c - as 01
•H OJ -rH O 01 *D
3 C rH AJ
•O O 3 Ol M
O -H 6 Ui ui r^
Ui AJ (0 -H r-
a o AJ
0) -n oi c r-
W 03 6rH OrH
ID -H ja -H v>
E r-l ID AJ
•O O AJ (0 >*-!
01 Ul 01 rH O
C IAJ J3 C 3
-H *J -H E AJ
U-l C K U
o> oi ri 01 o g
businesses are d
ge prices are tak
nts.
r of products ove
mulation. Number
lized cost of ref
estimate per proc
rH 01 ID 01 0) M (0
r-l 01 Ul g J3 O 3 AJ
(OrHOlOlE'A-'COl
ElO>OI30ICO
WOlriiOlSZ^I^U

, c u -a

•o
UrH
ui ro
ro c
01 U.
Ul
0.
03
• 3
>iO
01 3
> 10
W C_J
tn >,
Ul
>.AJ
Ui 0}
O 3
AJT3
c c
01 rH
>
c c
M O
• rl
0) AJ
C ifl
O-H
•rl AJ

W Ol
•H 01
gz
u
>,
U Ui
o o
> AJ
iO
0] rH
Ol 3
c 01
•rl 01
AJ 0£
ro
O .E
CJ UH
<
01
U D
ro z.
IM AJ
Ui
3 C.
W AJ

01 3
u
c c
ro o
C *ri
01 AJ
AJ 10
C Ul
•H 01
S §•
0
rH O
ro
•rl C
Ul-H
4J
01 C
g o
rrj .,_(
C AJ
H (0
T3 U
C O
ro 01
U)
r-l «3*
f—1 «,
ID
Ul 0]
3 Ol
AJ C
CJ— 1
0) AJ
AJ (TJ
-H O
•C U
L)

< c
ID •
' AJ CV
05 C fj,
AJ-H rH
£ a
>t available.
Industry Insig
for National P
Draft Report.
W
Z
CJ
II U
tH
< 3
^ 0
Z t/3
3-18

-------
estimated 421 small business products  in  the  survey  (42  percent)
exceed the 1997 TOS.   This figure is slightly higher  than  the
proportion of all survey products that exceed the  limit  (36
percent).  As established in Section 2, approximately one-third
of products over the proposed limit can costlessly comply  with
the regulation because of their similarity  to the  remaining  over-
the-limit products that are being reformulated.  The  remaining
over-the-limit products are referred to as  "constrained" by  the
regulation and the sum of the costless compliance  products and
under-the-limit products as "unconstrained" by the regulation.
     Less than 10 percent of the small business products in  the
sanding sealers,  mastic texture coatings, and bituminous
categories will be constrained by the  regulation.  Swimming  pool
coatings, shellacs, and high-temperature  coatings  produced by the
small business sector will require no  reformulations.  Traffic
paints, roof coatings, and varnishes are  all  relatively  high-
volume categories in which over 40 percent  of the  surveyed small
business products are constrained by the  1997 TOS  limits.
     3.2.3.1  Small Business Costs Under  "Reformulation-Only"
Option.  In this section, total and per-unit  annualized
compliance costs for small producers in each  product  category
with reformulation as the only compliance option is estimated.
The annualized $17,772 estimate of the cost per reformulation (in
1991 dollars) is multiplied by the number of  products constrained
by the regulation  (i.e., all products  over  the limit  less  the
one-third that can costlessly comply).  Table 3-4  provides the
cost estimates.  To contrast these costs  with product revenue,
the analysis uses average price per liter for each category  from
Section 2 for the market segment in which the category is
classified.9  The cost of reformulation as a percentage of
revenues is computed using the estimated  cost of reformulation
     'Where a coating category could not be separated into waterborne and
solventborne market segments  (categories in market segments  1 through 8)  a
weighted average of the two prices was used.
                               3-19

-------
divided by the imputed revenues for each product category.
     The data presented in Tables 3-1 and 3-4109 illustrate a
number of scenarios pertaining to potential small business
impacts of the regulation under a reformulation-only response
scenario.  Some of the phenomena indicated by the data are
examined below.
     Roof coatings is the largest quantity and highest revenue
category for small businesses.  For small business roof coatings,
43 percent of the individual products will be constrained;
however, the cost of reformulation as a percentage of sales is
relatively small, less than 1 percent.
     In the opaque waterproofing sealers category, small
producers comprise a very small share of the market--less than 1
percent-both relative to all producers in this market and
relative to small business shares in other markets.  However,
only two products are manufactured by small businesses in this
segment, and the data indicate that both of them exceed the TOS
limit.  Reformulation costs would exceed 50 percent of total
small business sales for the category.  This increased cost may
well lead to a compliance strategy other than reformulation
(i.e., fee or withdrawal).
     Antigraffiti coatings present quite a different small
business impact scenario.  Small businesses represent almost the
entire market but produce small quantities in relation to other
coating categories and generate lower revenues.  Only one product
requires reformulation under the TOS limits, but the cost
represents over 7 percent of revenues in the category.
     For the small business segment of the AIM coatings market,
42 percent of  their products are over the TOS limits, and 28
percent are expected to undergo a major reformulation, pay a fee,
or exit.  The  total annualized cost for the sample of small
businesses in  the survey under the reformulation-only option is
$5.0 million.  Table 3-5 presents small firm and  industry
                               3-20

-------
        TABLE 3-5.   AVERAGE  REGULATORY IMPACT BY FIRM SIZE3


                                         Industry            Small firm
        	average	average	

                                          38,990,002           4,614,010
   ($1991)

   Number  of products'1                           42.4                27.5

   Number  of products facing                      9.9                7.8
     major reformulationc

   Annualized  reformulation cost"3             176,648             138,720
   ($1991)

   Ratio of annualized reformulation             0.45                3.01
   cost  to revenues (percent)	

  "  The survey has 116 respondents and 36  of those identified  themselves  as
    having under $10 million  in  annual sales.  Twelve survey respondents
    did not  report company size.
  b  Data for  revenues and products per firm were based on data reported in
    Table  3-1.  The number of products per firm is based on the total
    number of  products for which quantity  data are available.
  c  This number represents two-thirds of the products over the 1997 TOS.
    Industry  experts estimate that approximately two-thirds of the products
    with VOC  contents exceeding  the TOS limits will incur the  estimated
    $250,000  per product "major" reformulation cost.
    Annualized cost of reformulation is the number of major reformulations
    multiplied by the annualized reformulation cost estimate per product  of
    $17,772  (1991 dollars).

  Source:  Industry Insights.  Architectural and Industrial Maintenance
          Surface Coatings VOC Emissions Inventory Survey.  Prepared for
          National Paint and  Coatings Association in cooperation with the
          AIM  Regulatory Negotiation Industry Caucus.  Final Draft Report.
          1993.


averages  for revenues,  number of products, and reformulation

costs.110   Small  businesses  on average manufacture approximately

one-third fewer products than the industry average.   On average,

small firms  have fewer constrained  products  than the industry

average,  but they comprise  a  slightly larger percentage of  total

number of products,  28 percent,  as  compared  to 23 percent for the

industry.   Similarly,  small business reformulation costs as a

percentage  of  revenues are  higher at 3 percent than  the industry

at roughly  0.5  percent.
                                  3-21

-------
3.2.4   Potential Mitigating Effects.  Fee  Option  and Small
        Product Exemption
     Fee Option.  As discussed in Section  2,  a product's output
level affects the choice between reformulating the product and
paying an exceedance fee.  Since the cost  of  reformulation is a
fixed cost (i.e., it is independent of output level)  the average
reformulation cost per unit of output falls as output levels
increase.  However, the exceedance fee per unit of output is
constant with respect to the output levels.  Thus, the fee is
more likely to be chosen by small volume producers, all else
equal.  Because the fee will be more cost-effective only for
lower volume products and lower excess VOC categories, allowing
the fee option should have a relatively small impact on variation
from the aggregate emissions reduction targets as long as the fee
assessment rate is not set at an inappropriately low level.
Therefore, the fee option provides increased flexibility for
small businesses by placing an upper limit on the per-unit costs
of complying with the regulation, without  significantly
jeopardizing VOC emission reduction targets.
     The model is unable to conduct a least-cost analysis of the
fee/reformulation decisions  (see Section 2) for the small
business segment of the  survey because of  insufficient VOC
stratum-specific data on small businesses.  Therefore, the
analysis cannot directly estimate the mitigating effect of the
fee option on  the costs borne by small businesses.  However, the
results  of the least-cost analysis in Section 2 can shed some
light on small business  impacts.
      In  Section  2,  the proposed  fee rate of $0.005 per g/L over
the  limit  (1995  dollars, adjusted to  $0.0045 in 1991 dollars)
leads to selecting  the  fee  over  reformulation for  323 of the
1,153 survey products  constrained by  the proposed content  limits.
Note  that  the  average  sales  quantity  for the 323  fee-selected
products is  approximately  the  same as the  average  product volume
for  all  small  business  products  in the  survey, suggesting  a

                               3-22

-------
strong correlation between small business producers and selecting
the fee.  To provide a measure of the maximum potential benefit
of the fee option for small producers,  assume all of the products
opting for the fee are produced by small producers.  Recall from
Section 2 that the fee option saves surveyed producers from
spending over $5.7 million in reformulation costs in return for
fee payments of just under $4.0 million, for a net savings to
surveyed producers of $1.8 million.   Deducting this cost savings
from the total reformulation cost for small business producers in
the survey (Table 3-4) leads to a 36 percent cost reduction if
these savings accrue entirely to small business producers.
     Low Volume Exemption.  As discussed in the previous section,
the fixed cost of reformulation places a substantial burden on
low volume producers that exceed the applicable VOC limit.  As an
alternative to a fee option which also can be costly for low
volume producers, the EPA can establish a low volume exemption.
This exemption would be designed to relieve low volume producers
from reformulation costs that can be difficult to recover from
the small amount of revenue generated by a low volume product.
Both the exceedance fee alternative and the low volume exemption
are compliance options aimed at addressing the potential issue of
"niche markets" in which low volume products exist for which it
may not be cost-effective for either the manufacturer or resin
supplier to develop a lower VOC formulation.
     The EPA lacks data to evaluate an appropriate sales volume
cut-off for a low volume exemption.  In the architectural coating
rule proposal, the EPA requests comment on a low-volume exemption
in the range between 1,000 and 5,000 gallons per year.  Through
proposal of the architectural coating rule, the EPA is seeking
detailed information from manufacturers who produce low volume
specialty niche products which cannot be cost-effectively
reformulated.  Specifically, the preamble requests the following
information for each identified product: sales volume, VOC
content, a detailed cost estimate for reformulation, and whether

                               3-23

-------
alternative products exist in the market which are able to meet
the table 1. VOC content standards.   Once this data is gathered,
the EPA can evaluate factors such as reformulation cost versus
affordability to small manufacturers.  In order to determine
affordability, one could compare the estimated reformulation
cost/unit with the current production cost/unit.   Further, the
size of the small volume cut-off would affect the overall
cost/effectiveness of the rule.   In addition,  the EPA plans to
work with the Small Business Administration Office of Advocacy to
evaluate available data on lower volume sales which was collected
in a survey conducted by the National Paint and Coatings
Association in 1992.  Under either the exceedance fee approach or
the low volume exemption, the EPA would consider the expected
magnitude of foregone emission reductions as compared to the
expected cost to achieve these reductions.
     While seeking ways to mitigate the impacts of the regulation
for small manufacturers, EPA also recognizes that the two
different approaches discussed here, the fee option and small
product exemption, have different implications for the marginal
incentives for VOC reductions.  Although the fee option continues
to provide incentive to reformulate the small niche products
because marginal reductions in VOC content will reduce the per
unit fee paid, a low volume exemption would provide no such
incentive.  Thus, when evaluating these options,   EPA must
consider the tradeoff between the level of relief for small
products versus the incentives to achieve VOC reductions.
3.2.5   Market Effects
     Data were not available to estimate market-level consumer
and producer surplus effects for the small business segment of
the AIM coatings  industry.  Table 3-6 summarizes the small
business share of total volume and  industry-wide producer surplus
effects for the 13 market segments  defined in the market model
described in  Section 2.111  The survey data did not allow
separation  into waterborne and solventborne segments for  the

                               3-24

-------
small business  data;  therefore, market  segments 1 through  8
combine  the  waterborne and solventborne portions.  In the  eight
categories  that can be divided into  solventborne and waterborne
market segments small businesses produce more products  in  the
solventborne segment than in the waterborne segment, except  for
interior coatings.   Although it is not  reflected in Table  3-6,
the majority of reformulations in each  sector occur with
solventborne products.
     In  the  architectural lacquers market segment,  small
businesses  produce over 25 percent of total volume, but that
market has  relatively small adverse  impacts on producer surplus,
a $464,000  reduction.  Small businesses produce 12 percent of the
total quantity  in the traffic marking paints market, where there
is a $1.2 million decrease in producer  surplus due to the
regulation.   Industrial maintenance  coatings experience the
largest  producer surplus impacts, a  $3.7 million decrease,  and
small businesses produce slightly more  than 5 percent of  total
volume.
     The data in Table 3-6 can be used  to compute a rough  proxy
for small business producer surplus  effects that is a share-
weighted average of total change in  producer surplus in each
market  (APSJ :

                         APSSB = 2 PS1 • Si58                    (3-4)

where  s^8 is the small business market  share in market  I.h  The
standard case scenario for the 1997  limits produces an  estimate
for APSSB of approximately $840,000  in producer surplus  losses
for small business producers.  This  is  roughly 4.7 percent of
     This estimate is a very rough measure of producer surplus effects  for
small businesses since  it merely apportions the market-wide producer surplus
based on small business market share, rather than on direct tracking of small
business withdrawals,  fee payments,  and gains by unconstrained producers, for
example.  However,  it  does underscore the notion that the  static losses
(aggregate reformulation costs) without accounting for price effects and
withdrawals will overstate the regulatory burden on producers. This holds for
all producing sectors,  not just small businesses.
                                3-25

-------
               TABLE  3-6.  MARKET-LEVEL  IMPACTS AND
                      SMALL BUSINESS PRESENCE3'D
Market
segment
number
1

3
5

7











& 2

& 4
& 6

& 8

9

10
11

12

13


Small All
business businesses
share of producer
total surplus
sales volume effects
Market segment
Exterior & high-
performance coatings
Interior coatings
Primers &
under coaters
Clear coatings,
sealers, & stains
Architectural
lacquers
Wood preservatives
Traffic marking
paints
Special purpose
coatings
Industrial
maintenance coatings

(
6

0
2

2

25

0
12

6

5

3
%)
.16

.96
.18

.44

.27

.00
.05

.78

.39

.56
(103 $1991)
-3,279

-3,468
-2,257

-3,336

-464

-245
-1,225

-182

-3,699

-18,156
Share-weighted
small business
producer
surplus
effects
(103 $1991)
-202

-33
-49

-81

-117

0
-147

-12

-199

-842
.1

.1
.2

.4

.4

.0
.6

.4

.2

.2
a  The emissions survey data did not  allow  separation into waterborne and
  solventborne segments for the small business data; therefore, segments
  1-8 combine the waterborne and solventborne portions.

b  The survey has 116 respondents and 36 of those identified themselves as
  having under $10 million in annual sales.  Twelve survey respondents
  did not report company size.


Source:   Industry Insights.  Architectural  and  Industrial Maintenance
         Surface Coatings VOC Emissions  Inventory Survey.   Prepared  for
         National Paint and Coatings  Association in cooperation with the
         AIM Regulatory Negotiation Industry Caucus.  Final Draft  Report.
         1993.
         See Appendix A for construction of the market  segments.
                                  3-26

-------
total producer surplus losses,  a proportion that is higher than
small business share of industry output (3.6 percent)  but lower
than small business share of industry emissions (6.2 percent).

3.3  SUMMARY

     The potential for significant impacts on small businesses of
the proposed regulation arise from two primary sources:

     •   Products made by small producers, on average, have
         higher VOC content than the industry average.
         The costs of reformulating products to comply with the
         regulation are fixed and thereby impose higher average
         costs on small volume coatings.

     The first problem is related to small producers'  tendency to
specialize in coating categories that are naturally higher in VOC
content and to their tendency to concentrate in the "high-VOC"
end of the distribution of products within a given category.
Thus the potential for disproportionate impacts of VOC reduction
regulation on small businesses follows partly from the fact that
small businesses contribute a disproportionate amount of the
aggregate VOC emissions that are targeted for reduction.
     The second problem follows from the nature of reformulation
costs.  A coating's formula is the product of an intellectual
capital investment, much like the development of a drug or a
computer software product.  The cost of the investment follows
directly from the level of effort necessary to revise the formula
to meet both the VOC standards imposed by the regulation and
performance standards imposed by the marketplace.  This level of
effort is essentially independent of the quantity of the product
that is eventually sold.  Therefore, the relative impacts on
smaller volume products is, by definition, greater.
     The data used in this analysis suggest that these two
primary factors are empirically relevant in the case of small AIM

                               3-27

-------
coatings producers.  The average VOC content  of the products made
by the small business producers in the survey is 75 percent
higher than the average VOC content of all  products combined.  A
little over half of the difference in the averages is attributed
to the specialization of small producers in high-VOC content
product categories, with the remainder attributed to the tendency
for small businesses to produce higher VOC  products within each
product group.  Moreover, the average product volume of products
made by small businesses is less than 20 percent of the average
product volume for the entire survey population, implying much
larger average reformulation costs.  Thus,  without mitigating
factors, the impacts on small businesses are  potentially
significant.
     The proposed regulation has been designed to mitigate small
business impacts.  Despite their inherently higher VOC content,
the proportion of small businesses products exceeding the
regulatory standards proposed for 1997 is not much higher than
the corresponding proportion for the survey population at large
(42 percent vs. 36 percent).  This suggests content limits have
been assigned among categories in the TOS with sensitivity to
small business presence.  In addition, the  availability of the
exceedance fee option is beneficial to small  business producers
because it places an upper bound on the per-unit costs of
compliance.  Data analyzed in this study indicate that producer
costs are reduced by nearly 40 percent in the cases where the fee
option  is chosen.  It is likely that much of  this savings in cost
.will accrue to small producers.
                               3-28

-------
                           SECTION  4

                          REFERENCES
1.   Stigliani, William M.  Chemical Emissions from the
     Processing and Use of Materials:   the Need for an
     Integrated Emissions Accounting System.   Ecological
     Economics 2.(4) :325-341.  1990.   (Adapted from Figure 2).

2.   Industry Insights.  Architectural and Industrial
     Maintenance Surface Coatings VOC Emissions Inventory
     Survey.  Prepared for the National Paint and Coatings
     Association in cooperation with the AIM Regulatory
     Negotiation Industry Caucus.  Final Draft Report.  1993.

3.   U.S. Department of Commerce.  1987 Census of
     Manufactures, Industry Series:   Paints and Allied
     Products.  Washington,  DC,  U.S.  Government Printing
     Office.  1990.

4.   Ref. 2.

5.   Ref. 2.

6.   Johnson, Duane.   The Best Paint for the Job.  Family
     Handyman. 42.: 18.  June 1992.

7.   U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products,  1991.   Washington,  DC,  U.S.
     Government Printing Office.   1992.

8.   Industrial Economics, Inc.   Use of Economic Incentives to
     Reduce Consumer and Industrial  Consumption of Solvents.
     Draft Report.  1989.
                              4-1

-------
9.



10,


11,


12


13,


14


15,

16

17

18
19


20,

21

22

23

24

25
Radian Corporation.  Regulation of Architectural
Coatings.  (Appendix E--U.S. Environmental Protection
Agency Non-flat Architectural Coatings).  June 26, 1987.

Consumer Reports Magazine.  House Paints and Stains.
52{6) -.365-374.

Consumer Reports Magazine.  High-Gloss Enamels.  p. 173.
March 1988.

Consumer Reports Magazine.  Interior Semigloss Paints.
p. 317.  May 1989.

Consumer Reports Magazine.  Paints for Finishing Touches
p. 619.  September 1990.

Consumer Reports Magazine.  Interior Latex Paints,
p. 333.  May 1991.

Ref. 10.

Ref. 13.

Ref. 14.

Brand, Benson G., and Guy E. Weismantel.  Exterior
Coatings for Wood.  In:  Paint Handbook, Guy E.
Weismantel  (ed.).  New York, McGraw-Hill.  Pp. 9-1 to
9-35.  1981.

Sunset Magazine.  What You Really Need to Know About
House Paint and Painting, 179:80.  September 1987.

Ref. 19.

Ref. 10.

Ref. 11.

Ref. 12.
Ref. 13.

Consumer Reports Magazine,
p. 568.  September 1988.
Interior Latex Paints,
                              4-2

-------
26.  Ref.  17.

27.  National  Paint  Coatings Association.  The Household  Paint
     Selector.  New  York,  Barnes and Noble Books.   1975.

28.  Bakke, Timothy  0.   Clean Air  Paints.  Popular  Science.
     211:85.   August 1990.

29.  Ref.  9.

30.  Ref.  6.

31.  SRI International.  U.S. Paint Industry Data Base.   Menlo
     Park, CA.  1990.

32.  U.S.  Department of  Commerce.  The 1982 Benchmark Input-
     Output Accounts of  the United States.  Washington, DC,
     U.S.  Government Printing Office.  1991.

33.  Whittington, Trevellyan V.  Paint Fundamentals. In Paint
     Handbook, Guy E. Weismantel (ed.).  New York,  McGraw-
     Hill.  Pp. 1-1  to 1-23.  1981.

34.  Beno, J., W. Brown, and P.P.  Obst.  Formulating and  Using
     Water Based Thermoplastic Resins for Wood Finishing.  In
     Proceedings of  the  Nineteenth Water-Borne, Higher-Solids,
     and Powder Coatings Symposium.  Robson F. Storey and
     Shelby F. Thames  (eds.).  University of Southern
     Mississippi, Department of Polymer Science.  Pp. 626-638.
     1992.

35.  Singer, Elias.   Raw Materials.  In Paint Handbook,  Guy E.
     Weismantel (ed.).  New York,  McGraw-Hill.  Pp. 3-1 to
     3-22. 1981.

36.  Rauch Associates, Inc.  The Rauch Guide to the U.S.  Paint
     Industry.  Bridgewater, NJ, Rauch Associates,  Inc.    1990.

37.  Ref.  35.

38.  Kemezis,  Paul.  Wait-and-See  Stance Taken on Zero-VOC
     Architectural Paints.  Chemical Week.  Pp. 52-53.
     October 1992.

39.  Ref.  34.


                              4-3

-------
40.  Ref. 36.

41.  Ref. 33.

42.  Ref. 33.   (Adapted  from Figure 1.4).

43.  Ref. 36.

44.  Ref. 28.

45.  D'Amico,  Esther.  Waterborne  Systems Gaining Nitche By
     Nitche.   Chemical Marketing Reporter.  238 (18) :SR20-SR28.
     1990.

46.  Ref. 28.

47.  Ref.  6.

48.  Ref.  28.

49.  Ref.  28.

 50.  Ref.  28.

 51.   Ref.  38.

 52.   Ref.  38.

 53.   Mullin, Rick.  Reformulation Continues as the VOC Target
      Broadens.  Chemical Week, p.  37.  October 1992.

 54.   U.S. Department of Commerce.   Current Industrial Reports:
      Paints and Allied  Products,  1982.  Washington, DC, U.S.
      Government Printing Office.  1983.

 55.  U.S. Department of Commerce.  Current Industrial Reports
      Paints and Allied  Products,  1983.  Washington, DC, U.S.
      Government Printing Office.  1984.

 56.  U.S. Department of Commerce.  Current Industrial Reports
      Paints and Allied  Products,  1984.  Washington, DC, U.S.
      Government  Printing Office.  1985.
                                4-4

-------
57.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1985.  Washington, DC, U.S.
     Government Printing Office.  1986.

58.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1986.  Washington, DC, U.S.
     Government Printing Office.  1987.

59.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1987.  Washington, DC, U.S.
     Government Printing Office.  1988.

60.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1988.  Washington, DC, U.S.
     Government Printing Office.  1989.

61.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1989.  Washington, DC, U.S.
     Government Printing Office.  1990.

62.  U.S. Department of Commerce.  Current Industrial Reports:
     Paints and Allied Products, 1990.  Washington, DC, U.S.
     Government Printing Office.  1991.

63.  Ref. 7.

64.  O'Reilly, Richard.  Product Development Drives Market
     Growth.  In:  Standard and Poor's Industry Surveys:
     Chemicals Current Analysis.  Pp. C40-C43.  New York.
     Standard and Poor's Corporation.  1991.

65.  Ref. 64.

66.  Ref. 64.

67.  Ref. 6.

68.  Ref. 14.

69.  Ref. 27.

70.  Ref. 3.

71.  Ref. 32.
                              4-5

-------
72.  U.S. Department of Commerce.  Census of Manufactures,
     Subject Series:  Type of Organization.  Washington, DC,
     U.S. Government Printing Office, February 1991.

73.  Source for Percentage Accounted for Data:  U.S.
     Department of Commerce.  1987 Census of Manufactures,
     Concentration Ratios of Manufacturers.  Washington, DC,
     U.S. Government Printing Office.  1992.

74.  Telecon.  Lindsay, Sam.  Small Business Administration,
     with Dempsey, Jenny.  Research Triangle Institute, July
     6,  1993.

75.  U.S. Department of Commerce.  1987 Economic Censuses.
     Volume 1, Report Series, Release ID.  Census of
     Manufactures:  Location of Manufacturing Plants.  file
     MC87LMCO.  1991.

76.  Ref. 36.

77.  Ref. 3.

78.  Loesel, Andrew.  Coatings Industry Faces New Mix.  In
     Chemical Marketing Reporter.  231(18):SR3-SR8.  1990.

79.  Ref. 78.

80.  Finishers' Management.  The U.S. Paint and Coatings
     Industry.  Pp. 23-25.  April 1991.

81.  U.S. Department of Commerce.  U.S. Industrial Outlook
      '92, Business Forecasts for 350 Industries.  Washington,
     DC, U.S. Government Printing Office.  1992.

82.  AIM Coatings Regulatory Negotiation Committee meeting,
     Washington, DC.  Meeting Summary.  July 28-30, 1993.

83.  Ref. 82.

84.  Ref. 82.

85.  Brady, Catherine.  VOC Regs Put a Strain on Structural
     Coatings.  Chemical Week.  Pp. 47-48.  October 1989
                              4-6

-------
86.  ICF Consulting Associates, Incorporated.  Small Business
     Economic Impact Study.  Prepared for the South Coast Air
     Quality Management District.   Final Report.  June 17,
     1988.

87.  Ref. 2.

88.  Ref. 2

89.  Fax from Sarsomy, Chris, Radian Corporation, to Murray,
     Brian, Research Triangle Institute.  January 16, 1995.

90.  Ref. 2.

91.  Ref. 2.

92.  Ref. 7.

93.  Ref. 2.

94.  Ref. 7.

95.  Letter from Nelson, Robert N,  National Paint and Coatings
     Association, to Ducey, Ellen,  EPA/QAQPS.
     April 2, 1991.

96.  Radian Corporation.  Determination of Architectural
     Coatings Baseline Sales and VOC Emissions.   Memorandum
     from Harrison, Rob, Radian Corporation,  to Duncey,  Ellen,
     U.S. EPA, OAQPS, Emissions Standards Division,  Dated
     November 22, 1995.

97.  Ref. 89.

98.  Ref. 7.

99.  U.S. Department of Commerce.   1991 Annual Survey of
     Manufactures:  Statistics for Industry Groups and
     Industries.  Washington, DC,  U.S.  Government Printing
     Office.  1992.

100. Ref. 99.

101. U.S. Small Business Administration.  The Regulatory
     Flexibility Act.  October 1982.

                              4-7

-------
102. Ref. 61.

103. Ref. 2.

104. Ref. 2.

105. Ref. 2.

106. Ref. 2.

107. Memorandum from Nelson, Robert,   National Paint &
     Coatings Association, to Madariaga, Bruce, EPA/OAQPS
     October 14, 1993.

108. Ref. 82.

109. Ref. 2.

110. Ref. 2.

111. Ref. 2.
                              4-8

-------
         APPENDIX A

     MARKET DEFINITION,
DEMAND ESTIMATION, AND DATA

-------
A.I  PRODUCT/MARKET CROSS-REFERENCE METHOD
     Data on coating prices, quantities, average VOC contents,
and proposed VOC content limits are necessary to estimate the
effect of VOC content limits on AIM coatings products.  Price
and quantity data were taken from the 1991 Current Industrial
Reports:  Paint and Allied Products.1   The Architectural and
Industrial Maintenance Surface Coatings Survey  (the survey)2
provided the sales-weighted average VOC emissions, which
represent VOC content.  VOC content limits were from the 1997
and 2000 TOS developed by EPA.
     Census data are organized according to product codes,
which define product categories; however, these Census product
categories differ from the product categories in the survey.
Furthermore, the TOS  (see Table 2-2) gives VOC content limits
for product categories that differ slightly from those
categories for which data are provided in the survey.  Data
from all three sources are necessary to conduct the economic
impact analysis.  Therefore, a fourth product categorization
was constructed, which is called market segments,  that
aggregates the categories so that data may be used from all
three sources to provide the necessary level of resolution for
market analysis.  Table A-l3'4 illustrates the individual
product categories represented by each data source and how
they map into the market segments used in the analysis.
                              A-l

-------
W
tt
 i
EH
O

B
o
cr;
n ki
o c
0 0
u A

u
n
•o B


•O E
C -H
a -,

* 1
X! C
a o>
|M kl



X
inventor
tsgory
_ «
0 °
•^ >
2*
e -
® n
u
O



" »
3 y 01
S| o
01 P U
u 0




X
kl
0


AJ
0
u
B
3
B
C

U

AJ
kl
O
£

a
ki
AJ
m
|
AJ
•
3






e



«
B

AJ

ki

JE




0>
ki T) k-
O C AJ U
O 1C W C
- 30
u m ki o
01 .
c - - K
01 •»* 0) 31 O
• U AJ u U ~«
• ki c a c c ki
B o n o ic • « • oi
Bi ••«. E u E oi e » AJ
C kl k. n > A. > X
04 01 OB a • o ~< o -* oi
a em ti o ^
«j AJ £0 3 * £ k, .c i. IA«
o a DIU AJ E en a. a o, c
a; u. i B x z z
. AJ
S £ 2
no > e
0 XI rt «
2 £ . g S
C 01 £ I 0
S> u i a
-1 kl (J 01 1
-H O « C Dl i
O B * C kl
B • > -« 0
1 • U ** AJ -H
1 kl C O IB kv
B o « B o «
C C k- i X
~< 41 O B W 01

« X K C 0
O 01 Oi •*« C AJ
U 0. AJ -i 1C


O  B
a « o> 3 -* AJ
c o c
AJ C AJ « • •«
g S S U V, S
> XI > 0 0
*-» iH O •* «
O Di O *H ki 3
0 C 0 *4 V O
kl T3 kl C V ^4
0 3 0 -1 « E
-, — _!, 3
01 C 01 X « -H
AJ -^ AJ . £ XI
K X AJ
U U O
01
c
kl
3
AJ
c
V


o
n
u
o


V
AJ

"
a>
k, T, £
O C AJ u
o « m c
-< 3 O
IM tn u u
> Oi
k, . c

• -H U AJ U U
* • ki C * C C
o ma IQ o IQ«IO«
•^ OlAJ E O EVEOI
kl CX k. « kl>kl>
01 -1 01 OBB< O -i O -I
••H « E B 01 1
E K z 3 3 x x

9 h
C « «
II -5
ki a £, i
9 ki O *
AJ 9 ( kl 0
o AJ i a DV
2 C k. k, ^C

U 1 ki C 41 10
o B oi n k o
••4 Ol AJ 01 E 1 O
kl C X C ki 1
01 ~« 0) kl o B «
AJ AJ O »w Ol 3
X 10 - XI kl C O
0) O AJ M 01 ~* C
0 « 01 Q AJ -g


a o c * DI o AJ
^ o o ~« -<
U. K Z X «
*-i (N in
w ^ in

in tn in
as as as
(N CN n
ki
O
O



T>
C

n
01 X
B U
a a>
XI -O


C B 0
^ AJ -4
AJ C kl
C -H 01 B
•H 10 AJ Q>
AJ a c c
? oT 3
« O ki O
kt O 0
0 -H
AJ "O kl kl
C C 9 9
iH « AJ AJ
Oi C 9 )
ki « kt ki
9 £* 9 O
5 c 5 S
* 5 * 0 S
O 3 O *9 9
ki U ki q ki
9 C 9 C 9
*J -rt AJ 9 £

u u o


9
C
o
A
kl
9

«

ki
O

ki
9
AJ

"






m
kl 0
O -^
• — « 9
k. kj £
O 4) H)
k. C 9
AJ >t
C - kl
AJ *^ U
5 1 g
9
C
U

O i
0 kl 0
i O —

o » c e
ki COD
4) -H J3
AJ AJ >^
C - C ki
C >

AJ VI O U
« C « -H
£20
ra in ^

^3 ^ H

OD 00 OD
r< n o*
-o
c
c

C 0

AJ V C
C E Of
-H « >
AJ C ^
« 0

« k> C

0 AJ
AJ J£ V
a u AJ B
-« — c
« 3 -£ ^-
— < *o *0 (5 «
^H C C -C 0)
c c « n a

0 * Oi J2
AJ B AJ 9
C j> C Oi
> »'•«"•«
— a 0 AJ
O AJ 0 C
0 C AJ ~* t-»
9 C 01 AJ
AJ > • -H
a — > E TJ
- 0 -• 9 C
0
O 0 O ki O AJ
£55^22.
C C C

9
c

jg
*J
c
9


O
0
b
O

Ll
V

c
t-H







kl
o

kl kt
0 9
ki C
AJ
C »
a
CO C
*H O
u. z

9
c
waterbori
r--
i 0

O V
ki C C
01 •* k

5 J "!
«o <
• rH AJ
AJ *J *
10 C *

•-< m
OD 00

in tn
00 00
(N (N



ki
0 9
9 £
0 AJ



C «
^H
AJ C
C -H
-H AJ
AJ 10
0
•o
c -
AJ £
C n «

ifl O* C
a.,,
9 0 *0
AJ 0 C
« O *
3g5
a
35 &

C C
M M



C


kl
9
AJ

J
kl
O

kl
9
AJ
C

0
« kl
m 9
9 

9
c
w
9 AJ 0 <
C C U >
W 9 9 — i
X) ^ "3 •
AJ O 9
C 0 0 '
5 . .S
O kl k. AJ
0 9 9 «
AJ E O
- « -* 0
IR O kl kl
w u a 9
9 ki *C
E 9 • (

ki C •
Ox O O
tn •-(

rH r*
in tn
eo ac
Ol OI







0 a

9 9
E E
ki W
a a



* 0

0 0
kl kl
9 ffl
AJ AJ
* c
O O
0 0
9 9
11
AJ AJ
C C
9 9
<-H .-I
O O

o o
-H «4
AJ AJ



B

01 B
E kl


a •
o
« u
C kl
kl «
.81
AJ 3 a
C 3
0) *O O
> C k,
~! « 0
0
M
                                                                                                                  4J
                                                                                                                  C
                                                                                                                  o
                                                                                                                  o
                                                                                                                XI
                                                                                                                 IB
                                                                                                                TJ

                                                                                                                 a>
                                                                                                                 w
                                                                                                                 (1)
                                                                                                                 4J
                                                                                                                 0
                                                                                                                 0)
                                                                                                                 4)
                                                        A-2

-------
O
U
u
En
W
cn
 I
CO
CO
o
tf
o
EH
0

g
o
OS
a
3





O
•o
Oi >
W ki
o o
C. O>
0 01
kl JJ
a 0
u
V)
TJ n
kl 4J
10 -«H
•a E
c •••«
0 *H
JJ
O J.
«
3 i
H !



><
O
4J H.
c £7
i|
c •

s »

« i
§

J^J
3 *
"8 £


kl
o
S1
•J:
u

«
3

C
a>
o

Jj
kl
o

a>





kl
JJ
to

*g


4J
s
kt
kl
3




jj



&
V


JJ
01
kl

X






n

"w o>
o> m
JJ 0)
0 w
0
u
k< n
o> ki
TJ 0> •
C E Q
3 1-1 h
kt V
*0 k\ JJ
C 0
0 NO
u u
n T) t
E U i
£ I3




c

O * ki
J3 k< 4
0) kl 01 4J
C 01 » 0
i t B 1
ki a
01 - - k.
jj a a o
0 W kl JJ
> 01 01 0
jj E O
- 0 -rt 0
a o k< fc*
w o a o
E 8 • }
•** Tl Q I
ki C •
O. D O
9 «>
9 CO
•H «-*
r* ft
in in
00 CD
(N (N



m a

0> 0)
E E


a a
TJ TJ
C B


n a

01 0)

0 0
o o
O 0

0) 01
TJ TJ
C C
3 3
kl kl
0) Ot
4J 4J
§ S

o o
kl kl
01 01

X C
bl M

B a
kl kl
01 01
E <•>
-5 0
ki O
a u
ki
01 01
£ c


ki TJ 0
01 C kl
JJ 0 OI
0
X



01
£ &
0) «
- a
u o


n M kl
ai a> oi
U) U) U






0>
c
kl
c
a>
*o
m
• tn
n o
u a
01 ~r
0) JZ
W M
m m r- r-
n r*- r*\ r-
r4 rl r+ *+
.H *-4 f* >-l
in in tn m
CD CD CD CD


to a
a oi
0 0
0) 0>
n 09

T) TJ
c c
0 c
n n
a> a*

n »

c c


n n
u u c c

* 01 0 "0

u u a a

jj jj jj AJ
£c c c
V « 9


o o o o
n « • n

O O O 0
"u *u "u "u
v v a* v

X C X C
w M « £

kl -
a kt
Ol 01

u <
a>
a) ta
c _c

jS c 2
jj ^ ta
C JJ
Ol « TJ
> 0 C
^ U *3
o
w
•
10


u

n m

a, ai

"« 1o TJ ^
01 01 C 1
m m "Q * !
» & 3 kl I
c c cr * a
*j *>-i « a -< c
§•0 n o u a
0) O 0) ki
W 3 k! £ - - JJ
& a & - c c i
X 3 > W W
O
c
kt

I O JJ JJ
I C C C
n os u • oi
oi c a> ja -* *
** v * jj o a
0 > *s c n n
0) ^ 0) JJ U C
n o ta c > - *
n ai -•< « ki a>
a 01 > o 3 jj c
c -c^no»-^ki
-H JJ — O « E 0
IM c *j m -aoijQ
o 01 o « o a jj
o E o • » c
Cli*^Q'BB01W>
kt&k.4J-^CC^
2 3 > « CO



























































ki


i-i
U

n n

0) 0>
•0 « ^ -
c JJ « n
a c o> 01
• ai n m
0) - ki
&U fi Ot 01
« Q. C c
a -i c *j iu
O U 0 O . O
. ki o 01 O
- a > tj u 3 k*
5^:^ § s as
w w co X 2


01
c
O JJ i

i- oi tn n
jj 9 n v oi
« c Q, ^H n ^
> b a 0 V 0
O C « JJ 0)
- ,Q 0 n 0 n
0) ki kt S
& JJ — 0) C - C
0 0 E C -H JJ —

0 « O 0 Oi O
j2 o e o
m ki m at Q, ^ t^
caiCJ-'kTQikiOi
w w w 2 2
Ch 00
^ 00
*-» <-4
in m
00 OD


0 n
Oi pi
n C

TJ JJ
C (0
0 0
o
n
c u
•«H Q)
0 JJ

n 2

a> o

> 0)
JJ «
kf e ki

-H <-H
01 01 01
JJ £ Q
w O
n
c
0

ki 01

0) TJ
** C
U 0

01 01
ki C
o -H a
A JJ 3
ki 0 O
0) O ki
JJ U Ol

2









0)
T

(
01
• o
tn — i
0)
«c .
•si
5S






01
kl
o
01
*J
;

n
n
c
>






































































0)
kl
<
i
i














n
ki
0
I

J
s
IT
CO
CJ




















kl

&
U

rH


10
kl
3
jj
U
0)
4J


U
kl








JJ
kl
0)

&*
u

i - Ol
0 <0 &
-» 0) 0
* -H a
ja u o
tn n C tn
SOi 01 0>
_^ _> ki >
jj JJ U jJ
0 0 B 0
> > C >
kl kl 0 kl
0) 01 k) 0>
H} a jj n
a> T) a -« oi
ki C U E M
a s a « o.
1 I'S^ 1
i i . £
i a>
,2 J g1
oi > a
J3 - 0
u •a
a o
a > o n
0) U > 01
> 01 >
-H B U 01 -.
u 0) C 01 u
V kl O t» 0
> a ki -H >
kl 0 4J kl
oi T a 0 «
» o a > a
* TJ O C ki 01
kl C 3 0 V kl
as ki a a
0 kl *J 01
TJ ki 0 -H ki 13
o 01 01 E a o
o — * » o
X u in 3
CT*
tn
en
(M
a
TJ TJ
0 C
- O 3
K J C
jj ^
3 k. O
TJ 3 O
O a
ki a oi
a oi c
u a N
ceo
•« O -"
0 C Ol
a
- T)
TJ B C
01 kl 0
~* 01
-i C -
-< 0 >•
0 01 U
-t U
m u 3
3 a
O £
0) a -
c 3 a
« k, 01
-i .0 _>

1 CT u
u j: 0
-~ TJ k.
65 S .
kl (J 01 U
01 C ki u
c •* a 01
o


a
j;
^j
«
4)
0)
h. a
a 3
Q
•O k,
0 Oi



tn
c
I
c


u
I

U
•^
•u
ki
I-









n

c

i

u

lu
ki
t-
S
tH
in
CO




















n

C

a


c


i

u
IW

0
ki
t-



0>
c
Jt
w

a
•i! c

IW «]
kl
•o
0)
c
• H
XJ
O
u
*"*"






























































aJ
i— i

(0
u

«-i
o
T3
C

10

V)
(U
4J
0
c


-------




Q
W
P

^H
FERENCE (CONTI
W
K

t
W
W
o
ry
H*
U

EH
CO
t*S
K

JEJ
^^
EH
O
»-^
i— >
Q
O
OS
0,



	 i
T"H

«:
u
ij
CQ
g




















•o
Table of standards propose
regulation limits category




>
VOC emission inventor
survey category



Hi
f 0 ?,
o |°




X
^
o
0.
4)
u
«
u

B)
a
i
u

w
o
d
«i
14
r«
*

1,
4.
n
—

c
u
c
•
IH
x







4,
C
i
i
i
4
4
A
I
l
a




in w
tji D>
c n c n
•^ -O -H V-
jj C jJ 41
RJ W 3 W iQ En
O W UOT3«O -HOI
U 0) Ol 01 d C Dl O IMC
pi C u£3C n D. -^
TJC--^ flOO-HCOi *-»
4) -H *J >MUQ.UOC£iQ
OT 4J *J * U E «•**—• « O
O\ C Q O 3DiOO«4J(BU
c o» o u tn c u u i-i c S
••-leu w 01 -^ *o ^*n
4J& -^bWM«a)EU4-*O&
« ^. Cn 4J flJ 3«Jjg COC
o cxc^tXTJU Qt-i-^ki at n -^
u -K *j * M oa o E *J
{j ^^ **-• fti « ® •"• *O -^ *J W «
&^3*UO^fU4>O«CO
O<-4O U £ 13 0))M-4iJU *» -H U
tM^ucn JC »- JS O -^ n E
iO ^ — « T3 -« U E & «J J-> *J E S
> *j *J i-> C * C >- « Ci — 4) -^ O
W«CCO£OOU£3U}^
Dr-2ia3uou.o-a:a.coti.


m
? n n
C 01 w)
•r* TJ ^*
4J C 0)
« 3 • E a
O 01 o t3 n — » o>
u cn a C en ^ c
C E 3 C 81 ft-H
TJ —^ O O -H O> *J
nu m e « -H B o
O>C O t?i O 0 AJ «J U
co) u c u u o a
^ E -H O -H
*j tji -^ ki m U4JO
(0 -^ 4J 3 AJ CO
O O. ••-< U kJ ^ 4> O.
u >w a o E
tl *M 0 "H -^ *J Ol
Cn-H ,AJ AJ C * -.It-H
tn 41 C OOV4 31x3
ax 3. u u. o z ix w



o
0
r>

s












J<
0)
c
0)
O)
c
-r4
AJ
 01
> c c *
-r* w< v4 Ct
*J iJ *J C
a a e -*
•* o o «J
: ° ° §
U « *J O
•-*. u C
AJ 3 0 «
C 4J E w
« a *> 3
Tl fc< U AJ
i* u ai x
c w a * *
W C E *J AJ
« •** 0) •**
h< AJ AJ n o
i « di -H
a o £ c AJ
u u a> o> m
••* -^ « «
b. XXX


in
o
n
in
OD
r<






•o
c
«
c
o
o
3
14
AJ
C
0
(J
J
0)
Q c n
AJ ^H AJ
C «J C

a b a
f± u Q

9 3 V
u -a u
£55
C C
« U 0)
WO*-1
C — t C
~H k« -«
S * S
E AJ E
X
u








V
u
c
(0
c
0)
u a
C 3
-H O
« W
1 &

n
ai
c
Sanding sealers
Thermoplastic rubber coati





Sanding sealers





















































and maitici



























































Industrial maintenance





Industrial maintenance





















































coatings
Heat reactive coat ings





coatings
Appurtenances





















































j=
AJ
o
£
u
ft)
TJ
C
3
TJ
O
JJ
to
n
•r4
u
u
3
TJ
O
l-i
a
A)
j=
^
c
o
JJ
fi
JJ
c
o>
n
ft)
u
U
W
o
•u
.,
£
>
Oi
n
^-t
n
a)
c
u
£

«
jj
c
»
•0
C
a
4J


>
p-t
0
n
£
U
O
J3

»4
«M
iJ
^4
^
ft)
C
o

>.
•-4
g
O
.
p
f*

ft)
£

*J
c
M
•



. Government Printing Office
u
Q
^
c
O
roducts, 1991 . Washingt
&
•o
0)

<
TJ
c
«
y
c
Kj
04

8)

0
a
V
{£
.-t
«

b4
iJ
n
3
T)
C
M
jJ
c
9
W
V4
3
U

0
U
iM
e
1
5
*0
jj
c
ft)
E
u
w
a
Q,
^
o
en
5


••

w
U
W
3
O
n



Prepared for the National
»ft report. 1993.
>. £
«TJ
W «^
3 a
CO C

VOC Emissions Inventory
tion Industry Caucus. F
n c
o> -^
c ^
— • o
AJ en
n o
o z
U
>,
0) U
o o
c u
*J C
w ^-*
3 3
V) O
o
o) o:
u
c x
•3 IH
C <

C £
•^ 4J
££
4J
rH — ^
« S

W C
U 0
n — •
3 *J
TJ «
C ^
»-< ft)
a
TJ O
C 0
« u

«-» c
« -H
1-
3 C
JJ O
O -~
a> *->
«j «
u o
tx B

»> C «
rH *-t Ow







A-4

-------
     The mapping  in  Table A-l proceeds  from  the most
aggregated  category  to  the  least aggregated  category.  In some
cases, however, the  survey  provides more detail than the TOS.
Where possible, the  market  segments were paired as solvent-
borne and waterborne coating categories.  Separate market
segments could not be created for  flat  and nonflat coatings in
the  interior  and  exterior segments because the Census data do
not  differentiate between exterior flats and nonflats.
     The necessary data were developed  for each of the 13
market segments using the mapping  scheme presented in Table
A-l.  Data  for individual Census product codes were summed
where necessary to compute  prices  and quantities.

A. 2  ESTIMATING DEMAND  ELASTICITIES FOR COATINGS

     To perform the  market  analysis, own- and cross-price
elasticities  of demand  were estimated for four broad coating
categories:   exterior solventborne and  interior solventborne
and  their two respective substitutes, exterior waterborne and
interior waterborne.  The variables used in  estimation are
domestic consumption quantity; real value of domestic
consumption;  real consumption price; national income; a
housing variable; and the real price of alkyd resins, acrylic
resins, and titanium dioxide.  Complete data for these
variables were collected for the years 1981  through 1991.
Justification of  these  variables and their data sources is
given below.
A.2.1  Estimation Procedure and Results
     Econometric  estimation of the interrelated demand system
for  interior  solventborne,  interior waterborne,  exterior
                              A-5

-------
solventborne, and exterior waterborne architectural coatings
generated estimates of own-price demand elasticities for each
of the four groups and cross-price demand elasticities between
the solventborne and waterborne segments of each interior
(exterior) pair.
     The quantity demanded of a commodity is a function of its
price, the price of any substitutes and other factors,  such as
income, that affect aggregate demand.  Estimating the demand
function, however, is more complicated than just running
regressions of observed market quantities on observed market
prices and other demand variables.  One must account for the
fact that the observed prices and quantities are equilibrium
values, which are simultaneously determined by both demand and
supply factors.
     Variables that are determined within a system (such as
prices and quantities in a market equilibrium system)  are
endogenous to that system, whereas those variables determined
outside of the particular system  (e.g., income,  housing
activity) are termed exogenous.  In simultaneous equations
models, endogenous variables are correlated with the error
terms through solution of the system.  As a result of  the
interdependence of the endogenous variables and the error
terms, the application of standard regression techniques is
modified to estimate the effect of an endogenous right-hand
side variable  (i.e., equilibrium price) on the endogenous
left-hand dependent variable (equilibrium quantity) .   In
general, ordinary least squares estimation of the individual
demand equations leads to biased and inconsistent parameter
estimates when a regressor is endogenous.
                              A-6

-------
     Endogeneity bias  is corrected by applying the two-stage
least squares  (2SLS) regression procedure for each estimated
equation  (see,  for example, Reference 5).  In the first stage
of the  2SLS method, the price observations were regressed
against all exogenous  demand and supply variables in the
system.   This  regression produced fitted  (predicted)  values
for the price  variables that are, by definition, highly
correlated with the true endogenous variable (the observed
equilibrium price) and uncorrelated with the error term.  In
the second stage, these fitted values were employed as
observations of the right-hand side price variables in the
demand  equations.  This procedure can also be used to estimate
the underlying structural supply equations; however,  because
of the poor performance of various specifications in the
supply  estimations, only demand estimates are reported here.
     The  2SLS  procedure was used to estimate the four demand
functions.  Both linear and double-log regressions were
estimated.  The double-log specifications are presented here
because of slightly better statistical fit and because the
parameter estimates are directly interpretable as point
elasticities.
     For  the two exterior categories, housing completions are
included  as an exogenous demand determinant.   Exogenous supply
factors incorporated into the first-stage regressions include
the prices of  various  raw material inputs and a price index
for substitute outputs, which captures the effect of non-
exterior  coatings prices on the supply of exterior coatings.
For the two interior categories,  U.S. domestic GNP is included
as a proxy for the exogenous effect of aggregate income on the
demand  for interior coatings.   Exogenous supply factors
                              A-7

-------
incorporated into the first-stage regressions also include the
prices of various raw material inputs and a price index for
substitute outputs, which in this case captures the effect of
noninterior coatings prices on the supply of interior
coatings.  The results of the demand estimations are shown in
Table A-2.
     Unfortunately, sufficient data to estimate the demand
parameters for the other market segments were unavailable.
For the other two solvent/water-paired segments—clear coatings
and primers/undercoaters—the mean of the respective own- and
cross-price elasticities from the interior and exterior
estimation process were used as proxies for the elasticities.
The other five segments—special purpose, industrial
maintenance group, traffic marking paints,  lacquers,  and wood
preservatives—are specialty groups whose demand is assumed to
be fairly inelastic and not dependent on prices in the other
segments.   Therefore, a value of -0.5 for the own-price
demand elasticity and zero for all cross-price elasticities
were assigned to each of these categories.   Table A-3 provides
the matrix of own- and cross-price elasticities for all 13
market segments.
A.2.2  Data Used in Demand Estimation
     Domestic consumption quantities and values were
calculated using data from U.S. Department of Commerce
publications Current Industrial Reports:  Paint and Allied
Products6"14 and U.S. Exports Schedule B Commodity by
Country.15"22  Domestic quantity and value of  shipments  figures
were used, which include exports.  Exports were then
subtracted to estimate domestic consumption (AIM coatings
imports are negligible and are not included in the
                              A-8

-------
               TABLE A-2.   DEMAND CURVE ESTIMATES
Variable
Exterior solventborne
demand
Log-housing
completions
Log exterior
solventborne price
Log exterior
waterborne price
Exterior waterborne
demand
Log-housing
completions
Adjustable Elasticity
R2 F-value estimate
0.94 50.52
0.17
-1.43
0.20
0.92 39.36
-0.05
t-statistic

3.30
-1.89
0.36

-0.62
   Log exterior                                    0.51         0.42
   solventborne price

   Log exterior                                   -1.89        -2.17
   waterborne price

Interior solventborne      0.69         8.49
demand
Log GNP
Log interior
solventborne price
Log interior
waterborne price
Interior waterborne 0.99
demand
Log GNP
Log interior
solventborne price
Log interior
waterborne price
1
-1
1
588.90
1
0
-1
.01
.50
.43

.00
.36
.39
1
-1
1

5
1
-3
.67
.74
.28

.07
.28
.80
                               A-9

-------








w
EH
U
O
W
W

W
a!
•?
g

c/3
0
s
M
EH
O
U

y
H

rf»

c:
o


XJ
H
E-J
s

^4
r .
H
o

r

rrt
Li-<
^
w
D
2
&
m
D

,
ro
^*<
TABLE J














4J
U
1
a

o
01
o
•H
a

c
•H
01
D)
«
J=
u

01
DI

-------
consumption variable).   Consumer price indexes from the U.S.
Department  of  Labor's  Handbook of Labor Statistics23  and the
U.S.  Department  of  Commerce's  Survey of Current Business24"26
are used to adjust  the current figures to real values.  Real
consumption price was  imputed  for each product by dividing
real  value  of  domestic consumption by the quantity of domestic
consumption.
      The GNP in  constant 1987  dollars from 1981 through 1991
was used as an aggregate income measure.a'27"29  Housing
completions for  1981 through 1991 were obtained from the U.S.
Department  of  Commerce's Current Construction Reports.30
Prices  for  alkyd and acrylic resins are obtained from the U.S.
International  Trade Commission publication Synthetic Organic
Chemicals,  U.S.  Production  and Sales.31"40   Prices  for titanium
dioxide were imputed using  quantity and value of shipment data
for U.S.  production from the Current Industrial Reports,
Inorganic Chemicals.41   Real prices  for these raw materials
were  calculated  by  deflating normal values using CPIs.  Alkyd
and acrylic resins  were  used to represent  raw materials for
the nonvolatile  vehicle  portion of the coatings,  which are
found mainly in  solventborne and waterborne coatings,
respectively.  Titanium  dioxide was used to represent a raw
material  in the  pigment  portion of the coating,  which is found
in both types  of coatings.  A  Laspeyres price index was
constructed to incorporate  the price of substitute outputs as
a supply-side  effect in  the first stage regressions of the
2SLS  procedure.  Let the price and quantity of  commodity n in
     aAll constant values were converted to 1982-1984 dollars for the
analysis to be consistent with the consumer price index  (CPI), which has
1982-1984 as a base.
                              A-ll

-------
period t be p^ and qnc, respectively  for n = 1,  ..., N and
t = 0, 1,  ...,T.  Then the Laspeyres price index of the N
commodities for period t  (relative  to the base period 0) is
defined as
             N        I  N
            V*   t 0    v^   0  0
        PL 5 2L,  Pn ^n  I JL  Pn ^n  •                       (A-l)
            N=l       I N=l
     Real domestic prices and quantities of nonexterior
coatings were used to construct the price index for the
exterior coatings equations and real domestic prices, and
quantities of noninterior coatings were used to construct the
index for the interior coatings equations.  Each index is
computed for the years 1981 through 1991, with 1981 serving as
the base year.

A.3  EVALUATION OF DATA QUALITY

     The Current Industrial Report series is generally
considered a reliable source for quantities and values of
products shipped.  Monthly and annual data were estimated from
a sample designed to measure activities of the entire paints
and allied products industries.  Each annual report provides
data for 2 years, and the most recent figures were used for
the coatings analyses.  In addition to the four representative
coatings products, the architectural coatings Census category
includes two other products:  architectural lacquers and
architectural coatings,  not elsewhere  classified.   These
categories were not included in the estimates because of
                             A-12

-------
 insufficient data.   However in 1991,  these two product
 categories combined  represented only 1.3 percent of  the  total
 value of shipments for  the architectural coatings market.4
 Statistics reported  in  the Current Industrial Reports at  the
 seven-digit SIC product level are based on Annual Surveys  of
 Manufactures and  represent about 95 percent of total shipments
;in the paint industry (SIC 2851)-43   To  produce  estimates for
 the entire industry,  the Census Bureau inflates the  quantity
 and value figures reported in the annual survey by a factor
 based on data reported  by all establishments in the  1987
 Census of Manufactures.13   The inflation factors  for
 architectural coating product categories are as follows:   1987
 through 1991, 1.00;  1982  through 1986,  1.004; and for 1981,
 1.04.c  Quantity  and value figures for  the four  product
 categories used in the  demand estimation are inflated using
 these factors.  Prior to 1981,  data were not collected at  the
 more specific seven-digit SIC level.   Using the longer time
 series would provide more data points but would also preclude
 analysis of the individual product categories,  and
 representativeness would be lost.
      The export data used are the best publicly available;
 however, combining export and domestic data to estimate
 domestic consumption poses some problems.   The classification
 systems used to gather  both types of data are different, and
 the corresponding product categories used cannot always be
     ''The inflation factor for 1981  is based on 1977 Census relationships
and for 1982 through 1986 on 1982 Census relationships.
     cThe 1991  quantities and values used in the model (values  to impute
price)  also include products in the  special purpose and miscellaneous
allied paint products categories. The special purpose inflator for 1991 is
1.06,  and the miscellaneous inflator in 1991 is 1.18.
                               A-13

-------
compared.  For  example, data from the U.S. Department of
Commerce publication U.S. Imports for Consumption and General
Imports, TSUSA  Commodity by Country of Origin were not used
because the  imported commodity classifications had no
comparable domestic output classification.  Exclusion of
imports from the estimate of domestic consumption does not
pose a problem  because in 1991 the value of imports for
architectural,  OEM, and special purpose coatings (SIC 28511,
28512, 28513) combined represented less than 0.9 percent of
the total domestic value of shipments.44   Data  from U.S.
Exports Schedule B Commodity by Country were available for
1981 through 1991, and the export categories correspond well
with the four domestic product categories except for 1989
through 1991.45~52  In 1989,  the export codes and categories
changed and  are no longer compatible with the domestic
categories.   In addition, quantities are reported in kilograms
rather than  gallons, as they were in previous years.   For
these reasons,  export data were not used to adjust domestic
consumption  after 1988.  The GNP data typically represent
income for the  entire nation including income generated from
American businesses located overseas.  The current price data
for the paint products and raw materials should be considered
reliable, though their accuracy may be affected by the
exclusion of imports for the coatings products and of exports
and imports  for the raw materials prices.  CPIs for all urban
consumers with  a base of 1982 through 1984 were used in
calculating  real prices.
     The raw material prices used are representative of the
entire U.S.  and export market for these products, rather than
just the U.S. supply to the paints and coatings industry.  The
                              A-14

-------
alkyd resins were used in this estimation to represent an

input found only in solventborne coatings and acrylic resins

to represent an input found only in waterborne coatings.

However, some acrylic resin materials are used in some

solventborne coatings and alkyd resins are used as modifiers

in waterbornes.  Exports and imports were not considered when

computing raw material supply prices because foreign trade

data were not available for alkyd and acrylic resins.  In

1991, exports of titanium dioxide represented 17.9 percent of

the total domestic value shipped and imports were 10.9

percent.53


A.4  REFERENCES


1.    U.S. Department of Commerce.   Current Industrial Reports:
     Paint and Allied Products,  1991.   Washington,  DC:   U.S.
     Government Printing Office.   1992.

2.    Industry Insights.   Architectural and Industrial
     Maintenance Surface Coatings  VOC Emissions  Inventory-
     Survey .   Prepared for the National Paint and Coatings
     Association in cooperation with the AIM Regulatory
     Negotiation Industry Caucus.   Final draft report.   1993.

3.    Ref. 1.

4.    Ref. 2.

5.    Pindyck,  Robert S.,  and Daniel L.  Rubinfield.
     Econometric Models and Economic Forecasts.   2nd Ed.   New
     York,  McGraw-Hill,  Inc.   1981.

6.    U.S. Department of Commerce.   Current Industrial Reports:
     Paints and Allied Products,  1982.   Washington,  DC:   U.S.
     Government Printing Office.   1983.
                             A-15

-------
7.   U.S. Department of Commerce.   Current Industrial Reports
     Paints and Allied Products,  1983.   Washington,  DC:   U.S.
     Government Printing Office.   1984.

8.   U.S. Department of Commerce.   Current Industrial Reports
     Paints and Allied Products,  1984.   Washington,  DC:   U.S.
     Government Printing Office.   1985.

9.   U.S. Department of Commerce.   Current Industrial Reports
     Paints and Allied Products,  1985.   Washington,  DC:   U.S.
     Government Printing Office.   1986.

10.  U.S. Department of Commerce.   Current Industrial Reports:
     Paints and Allied Products,  1986.   Washington,  DC:   U.S.
     Government Printing Office.   1987.

11.  U.S. Department of Commerce.   Current Industrial Reports;
     Paints and Allied Products,  1987.   Washington,  DC:   U.S.
     Government Printing Office.   1988.

12.  U.S. Department of Commerce.   Current Industrial Reports:
     Paints and Allied Products,  1988.   Washington,  DC:   U.S.
     Government Printing Office.   1989.

13.  U.S. Department of Commerce.   Current Industrial Reports:
     Paints and Allied Products,  1989.   Washington,  DC:   U.S.
     Government Printing Office.   1990.

14.  U.S. Department of Commerce.   Current Industrial Reports:
     Paints and Allied Products,  1990.   Washington,  DC:   U.S.
     Government Printing Office.   1991.

15.  U.S. Department of Commerce.   U.S.  Exports  Schedule B
     Commodity by Country.   FT446/Annual 1981.   1982.

16.  U.S. Department of Commerce.   U.S.  Exports  Schedule B
     Commodity by Country.   FT446/Annual 1982.   1983.

17.  U.S. Department of Commerce.   U.S.  Exports  Schedule B
     Commodity by Country.   FT446/Annual 1983.   1984.

18.  U.S. Department of Commerce.   U.S.  Exports  Schedule B
     Commodity by Country.   FT446/Annual 1984.   1985.
                             A-16

-------
19.  U.S. Department of Commerce.   U.S.  Exports Schedule E
     Commodity by Country.  FT446/Annual 1985.   1986.

20.  U.S. Department of Commerce.   U.S.  Exports Schedule B
     Commodity by Country.  FT446/Annual 1986.   1987.

21.  U.S. Department of Commerce.   U.S.  Exports Schedule E
     Commodity by Country.  FT446/Annual 1987.  1988.

22.  U.S. Department of Commerce.   U.S.  Exports Schedule E
     Commodity by Country.  FT446/Annual 1988.  1989.

23.  U.S. Department of Labor.   Handbook of Labor Statistics.
     Bulletin 2340.  1989.

24.  U.S. Department of Commerce.   Survey of Current  Business.
     v. 70,  no. 7.  July 1990.

25.  U.S. Department of Commerce.   Survey of Current  Business.
     v. 72,  no. 7.  July 1992.

26.  U.S. Department of Commerce.   Survey of Current  Business.
     v. 72,  no. 12.  December 1992.

27.  Ref. 24.

28.  Ref. 25.

29.  Ref. 26.

30.  U.S. Department of Commerce.   Current Construction
     Reports, Housing Completions,  November 1992.   Washington,
     DC:  U.S. Government Printing  Office.   1993.

31.  U.S. International Trade Commission.   Synthetic  Organic
     Chemicals, U.S. Production and Sales,  1981.   USITC
     Publication 2470.   1982.

32.  U.S. International Trade Commission.   Synthetic  Organic
     Chemicals, U.S. Production and Sales,  1982.   USITC
     Publication 2470.   1983.
                             A-17

-------
33.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1983.  USITC
     Publication 2470.   1984.

34.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1984.  USITC
     Publication 2470.   1985.

35.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1985.  USITC
     Publication 2470.   1986.

36.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1986.  USITC
     Publication 2470.   1987.

37.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1987.  USITC
     Publication 2470.   1988.

38.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1988.  USITC
     Publication 2470.   1989.

39.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1989.  USITC
     Publication 2470.   1990.

40.  U.S.  International Trade Commission.  Synthetic Organic
     Chemicals,  U.S.  Production and Sales,  1990.  USITC
     Publication 2470.   1991.

41.  U.S.  Department  of Commerce.   Current Industrial Reports
     Inorganic Chemicals,  1991.  Washington,  DC:  U.S.
     Government Printing Office.   1992.

42.  Ref.  1.

43.  Ref.  1.

44.  Ref.  1.

45.  Ref.  15
                             A-18

-------
46.  Ref.  16
47.  Ref.  17
48.  Ref.  18
49.  Ref.  19
50.  Ref. 20
51.  Ref. 21
52.  Ref. 22
53.  Ref. 41
                             A-19

-------
             APPENDIX B

METHODOLOGY FOR COMPUTING MARKET AND
         WELFARE ADJUSTMENTS

-------
B.I  METHODOLOGY FOR COMPUTING SUPPLY EFFECTS
     For  the purposes of modeling the regulatory  effects  in
each market, producers are separated into  four categories,
based on  their  response to the regulation:

      •    producers  of products slated for  withdrawal,
      •    producers  of products on which exceedance  fees are
          paid,
      •    producers  of products slated for  reformulation,  and
      •    producers  unconstrained by the regulation.

     The  baseline  (preregulatory) quantities from these groups
are denoted as  follows: Qx,  QF,  QR, and Qu  for groups  1, 2, 3,
and 4, respectively.  Total baseline market output  equals the
sum of the four components:
                    Q = QX +  QF + QR  +  QU.                (B-l)

     Figure B-l depicts the aggregation of these  subgroups
into a market supply function.  The regulation causes a shift
in the aggregate supply function depicted  in Figure B-l as a
result of two phenomena:  an inward supply shift  due  to
eliminating Group 1 through product withdrawals (e.g., the
shift from S° to S1) ,  and  an  upward supply shift due to
imposing per-unit fees on the products from Group 2  (the shift
from S1 to S1') .  There is no supply shift  emanating from  Group
3 because the reformulation is assumed not to affect marginal
production costs, and there is no shift from Group 4 because
the unconstrained producers experience no regulation-induced

                              B-l

-------
            S/Q
            P1
            P0
                                 Q'
 Figure B-l.  Single  market effects of VOC content regulation.
change in their  cost structure.    So  the full  regulation-
related shift  is from S° to S1',  which leads  to a new market
equilibrium.   At the new equilibrium,  price  rises to  P'  and
quantity falls to Q'.a
B.2   DEMAND EFFECTS
      Figure B-l  depicts a partial equilibrium view of  the
short-run effect of imposing  content  limits in  one market
           graphical analysis demonstrates that the post-regulatory market
effects  are uncertain if we were to consider the possibility that  the
reformulation process changes the marginal cost of producing the coating as
a result of changes  in material or labor costs, for example.  This
empirical issue can  be resolved given sufficient data on the effect of VOC
content  on production costs for all affected products.  Unfortunately,
these data were not  available for this study, and we cannot conduct the
appropriate empirical analysis to draw such conclusions.
                                 B-2

-------
One must also consider the role of substitute products in
determining the equilibrium adjustments,  which suggests a
multimarket perspective.  Figure B-2 depicts  the markets for
two products  (A and B) that are demand  substitutes.   The price
of product B  factors  into product A's demand  function and vice
versa:

                       DA  =   DA(PA, PB)                    (B-2)
                       DB  =   DB(PB, PA).                   (B-3)

Given that A and B are substitutes implies

                         6DA / 6PB > 0                      (B-4)
                         6DB /  6PA > 0  .                     (B-5)

     Suppose the supply of A is affected  by the content limits
in the manner described above, but that the supply of B is
unaffected.  This initiates a supply shift in market  A from SA°
to SAR.   Holding the  initial demand function constant, this
shift would generate  an equilibrium quantity  of QA" and  price
of PA".   However,  the associated price increase  in market  A
induces an outward shift in the demand  for product B,  which
raises the price of product B.  Likewise, the increase in B's
price leads to an outward shift in the demand for product A,
which raises its price and so on.  This interaction continues
until post-regulatory equilibrium is established at  (PAR, QAR) .
(pBR'  O •

B.3  COMPUTING CHANGES IN EQUILIBRIUM PRICES  AND QUANTITIES

     The change in equilibrium prices and quantities  for the
products affected by  the content limits and their substitutes
can be numerically computed by adjusting  the  equations  in the
multimarket supply and demand system to reflect  the imposition
                              B-3

-------
        $/Q
                                           ? (Pg)
                                             QA/t
        S/Q
                          Product A
           B
          B
                                         Dg (Pj)
                            QB QB

                          Product B
Figure B-2.  Multiple market effects  of VOC regulations
                           B-4

-------
of these limits.   For each market,  i,  the equilibrium change
in quantity supplied of  each product affected by  the
regulations equals the sum of  the  supply changes  from each of
the producer subgroups :

                                               .           (B-6)
The change  (from baseline)  in quantity supplied by  the
withdrawal  sector is simply the negative of the quantity
originally  supplied by  that group:
                              = - QiX.                     (B-7)

     The change in quantity  supplied from the fee-paying
sector is specified as  follows:

                   AQiF = eJ(QiF/Pi) (APr FJ              (B-8)

where e^ is the supply elasticity of the fee producers in
market i, APi is the change  in equilibrium market price, and
other terms are as previously defined  (without the
subscripts) .  APi- F^ is the change in "net price" for the
fee-paying producers  (i.e.,  the change in unit process  less
the unit fee) .
     The changes in quantity supplied from  the reformulating
group and unconstrained groups, respectively, are
                                                          (B-9)
                        iU =  elu(Qiu/P)APi.                (B-10)

     These producers respond to  the increase in price with no
counteracting effect on costs.   Given the higher price in the
post-regulatory equilibrium, output will increase from these
two groups of producers.
     The aggregate change in equilibrium supply quantity can
now be restated by combining the preceding five equations:
                              B-5

-------
                                  VFi> + eiR(QiR/Pi)APi
               +  eui(Qui/P)iAPi .                            (B-ll)

     The  change  in market  demand  for  each product  is  given by

            AQ,D  =   E,; (Q.  /  PJAP, + E,.j(Q,  /  PS)AP,       (B-12)
               1       J.A  1     11    1J   1    J   J

where Eii is the own-price demand elasticity for product, i and
EAj  is  the associated cross-price demand elasticity between
products  i  and j.  Consumer  demand theory supports  the
assertion that own-price elasticities are negative  and that
cross-price elasticities of  substitutes are positive.   To
attain equilibrium,  the change  in quantity  demanded must  equal
the change  in  quantity supplied in both markets:

                         AQ^ =  AC^3.                    (B-13)

     This provides a system  of  M*3 equations  in M*3 unknowns,
where M equals the number  of markets  affected by the
regulation.  This can be reduced  to an M*2  equation system,
simply by substituting AQiD=AQis=AQi.  This  system can be
solved simultaneously to compute  the  change in equilibrium
price and change  in  equilibrium quantity for  each market.   To
do this,  baseline market data,  model parameters  (supply and
demand elasticities),  and  an empirical characterization of  the
various supply shocks alluded to  above are  needed.

B.4  COMPUTING WELFARE EFFECTS

     Changes in  the  market equilibrium cause  changes  in
resource  allocation,  which,  when  quantified,  provide  measures
of how the  welfare costs of  the regulation  are distributed
across groups  affected by  the regulation.   The groups  focused
upon here are  AIM coatings producers  and consumers, because
the changes in prices and  quantities  directly affect  their
                              B-6

-------
welfare.  Since  fee payments are considered, the government
sector  is also included  in the welfare analysis because they
collect the  fee  revenues.   This study does not measure the
welfare benefits of reductions in VOC emissions, a value
against which these costs may be measured to determine the net
value to society of the  proposed regulatory structure.
B. 3.1  Effects on AIM Coatings Producers
     The profits earned  at the new equilibrium to the profits
earned at the old equilibrium can be compared as a measure of
effects of the regulation on the individual producer.  Forgone
baseline profits (7t°)  provide a measure of the loss to
producers that choose to exit rather than reformulate:

                   A*  =  *R*   -  n°  =  -ic°.             (B-14)

For the remaining producers, the change in profits is affected
by several factors, including the incurrence of the fixed
reformulation cost and any associated changes in price,
quantity, and marginal cost.
     The remaining firms' costs may be affected through either
the reformulation cost or the fee payment. The effect of the
content limit on producers is generally not uniform and thus
raises some  distributional considerations.  As indicated
above, shifts in the aggregate supply function will cause the
market price to  rise.  For some producers, the benefits of the
price increase may outweigh the net costs of compliance.  This
is certainly the case for producers of coatings with VOC
content below the regulatory standards, because they incur no
reformulation costs but  would gain from the rise in market
price sparked by the compliance costs and/or product
withdrawals  incurred by  their competitors.  Alternatively,
fixed reformulation costs may be substantial for some
producers, outweighing the positive price effect.   The profit
effect will  be negative  for those producers.  Other producers
                              B-7

-------
may fall  in  the midrange, where the price benefits and  cost
effects essentially offset each other.
     Changes in producer welfare are generally reported as
changes in producer surplus.  The aggregate change in producer
surplus for  the withdrawn-product producers equals the  sum of
forgone profits from all withdrawn products in market i:

                      N*
              APS,*  = -£ n.. .                              (B-15)
                      j-i

The j subscript indicates forgone profits from the j ' th
product in market i.  N/1  is  the number  of withdrawn  products
in market i .  The change in producer surplus from the
reformulating sector can be approximated as follows:
                       Q/ + 0.5*AQiR*APi-  (R^N) .      (B-16)

APi is the change in equilibrium price, AQ/ is  the change in
equilibrium quantity from  the reformulating producers, QiR is
the initial quantity of  the reformulating producers, Rac is the
annualized reformulation costs, and NiR is  the number of
products needing reformulation.
     The change in producer surplus for the fee-paying
producers is
         APS/  =  (APi-FJ* (Q/  + AQiF)-0.5*AQiF*(APi-Fi) .   (B-17)

The first term reflects  the net revenue effects of the price
rise less the  fee payment and the second term reflects changes
in deadweight  loss .
     Finally,  the change in producer surplus for unconstrained
producers is
                 APS/ = AP^Q/ + 0.5*AQiu*APi           (B-18)

with the Qiu reflecting the quantity supplied by these pro-
                              B-8

-------
ducers .  Total  (net) producer  surplus effects is simply the
sum of  the  terms above:

              ApSi  = APSiX + APS/ +  PS/ + APS/.         (B-19)

B.3.2   Effects  on AIM Coatings Consumers
     Changes  in consumer welfare  are measured by the change in
consumer surplus, which quantifies losses due to a combination
of the  higher price and reduced consumption quantity.   This
change  can  be approximated as  follows :
                 i =  -APi*(Qi + AQA)  + 0.5*APi*AQi.        (B-20)

B.3.3  Effects on the Government Sector
     The transfer of fees from the fee-paying producers  to  the
recipient of those fees must be considered.  For the purposes
of the welfare analysis, the government is identified as the
"recipient" of the fees.

                     AGSi =  Fi*(QiF +  AQ/) .               (B-21)

     Ultimately, the government may choose to redistribute
those fees back to affected producers or consumers or back  to
other members of society via the Treasury; however, for
purposes of quantifying these distributional flows, they are
assigned as gains to the government sector.
B.3.4  Net Welfare Effects
     The net welfare effects are computed by taking the sum of
producer, consumer, and government surplus:

                   Avrcv  = APSi + AcSi + AGSi.             (B-22)

This calculation nets out any transfers from one group to
another within society  (e.g., transfers from consumers to
producers through higher prices and transfers of fee revenues
from producers to the government)  because these transfers do
                              B-9

-------
not affect the total sum of resource costs,  just how they are
distributed within society.  AWFi provides an estimate of the
net social costs of the regulation.
                             B-10

-------
                    APPENDIX C.

VOLATILE ORGANIC COMPOUND CONTENT LEVELS AND EMISSION
     REDUCTIONS FOR SELECT ARCHITECTURAL COATINGS
               FROM THE 1990 SURVEY DATA'.

-------
r




Units are grams of VOC per liter
o

o
to
§'
.

5
c"
Q.
I
O.
ft
X
n
1
1

c
1
",-.
?
5
c.

o
ft
X
§"
c
3

5
5
(TO
S
6

tt
n
C-
_
v:
5-
a
c
f^
c
ft












cr




All emission reductions represent
3
a.
o
|"
i«
c
<
5
S"
jg
a
S









































B




Unless otherwise specified, units i
the manufacturer.
™

5"
*z
|

o
<
R
"S
ft
C
-*>
E
5
S"

B
5
ft

T3

O
0
1
O
c
•3
O.
V.
"
1
r.
o"
to
B
C.
C.
ft
o.
0
a
y
80
S
trt
CD
ft
i
X
§'
i
?
5
2
3
00
rt
c
3
3
a
i.
«•




03
£
i
to
O
i
?r
1'
£
a
n
n

n
3.
?
3

v:
c
c.
ft
•8
o
a
n
o.
c
3
*•<
C.
to
C

en
»
a*
W
^
Er
B"
rt
to
c
3-
8
ft
c
3
ft
S
ft"
<
R
ft
a.
o
c
cr
tr





















?- Crt

^ c.
:- B
; Sealers (other than Lacquer Sandi
TVvt.ir.. r.^tlo^
3

21
n
£•





^ ^



^

S o






£ §




5 0



3 —





i j















^^2DOOOt~"n<<'"'n

c «» ^ o ^j£ ~3 »?«""
^ -3 ^J'Ts ^S Pg.-3 "Z Q
!itl||||f 5!ll
^ifaKg^n^r y = m
 "• £. 5 "
c i.











0^^1^**UJ*.0^^0*.O^^^J

ocSooooooooooo






-s-----s--«--



— |O UJ U> U)
O LA LA OO ^O



J^ N) \f\ "LA "so yj 1*J ~C> "sO ^_ io V. "to

O



*"" ^ *o "~















zr is -Z. TO v> "Z.
£_ 2. o C to C
= 3 5 x 5 s
=. S s o T f
Coalings - Interior
Opaque
oatings
Coatings - Exterior
> and Undercoaters
al Maintenance Coatings









^ NJ tO




ii. Ui UJ tO U> U»

o o o o o o






UJ — UJ -~J K> O^



4* -t»- 4^ LA LA f>
V) LA Vj Lu "it*, "ifi.
UJ -J •— \C O LA



O ^s* 00 oo u> to
"oo u> 1y\ »- oo ~oo

-ti *O -f^ O^



-O OO Ui . — UJ
-o u* o ij to —
O* to u* ^^ -^j *.















W5
IT
5
is-
Clear and Semltransparent










LA



LA
LA
O






LA
U>



£
S



O
OO

ta—



OO
-o















H

^,
o
2
ft
rr
I
?
5
tn










u>





o






0,



to
0
8



to

LA



^



















0
c
5
re
o
tt
ft
re
C








HE >
^ • (r
iti

<
D
f

r
tc
€



s> C <^
1 s. a
Ui
n 5


^ a ^
j c. 5
r- C ^
"cr§ 3



O
A



m
? 1 D
o < £
D « «
—' D
ft
cr

























0

m F
D m
og
d o
0 >
2 -2.
•n "" 0
o o g
H S ?

m m c
- 3 z
o H o

w g o
c n 5
» £ H
< 3 m
m tn Z
D c fn
> i P
I [
8>
> 2
H C
Is
w
O
2











































^
-a
m

D
5?

































-------
             APPENDIX D:




Sensitivity Analysis of National Cost Estimate

-------
                                     APPENDIX D:
                       Sensitivity Analysis of National Cost Estimate
       In the Economic Impact Analysis, the total social cost is extrapolated from detailed
 information received from a survey of 116 companies in the industry. The method of
 extrapolation assumed the volume of the products not included in the survey and the companies
 that produce them have similar characteristics as those in the survey. Because this does not
 account for several of the companies in the industry that produce products with lower annual
 volume relative to those of the survey, the analysis provides a lower bound of the actual
 compliance costs that will result from the regulation. An alternative method of accounting for
 nonsurveyed products and their associated reformulation costs is presented below to provide and
 upper bound of total social cost of the regulation. The actual cost imposed on society is likely to
 be between the upper and lower bounds presented by the two methodologies.

 Current Method:

       National reformulation costs are currently estimated in the Economic Impact Analysis by
 computing reformulation cost for the survey populations and multiplying this number by the
 ratio of national market volume in each of the 13 product categories to the volume of surveyed
 products.

  National Reformulation Cost = Reformulation Cost of Survey Products * (Market Volume by
                             category/Survey Volume by category)
This is completed for each AIM coating market segment and summed across all market segments
to get the industry total.

       For this method to be an accurate measurement of national reformulation costs, the ratio
of market volume to survey volume must equal the ratio of the number of a market products to
the number of survey products. This is because reformulation costs are incurred on a per
product, rather than a per unit, basis. Therefore, the average product size of nonsurveyed
products must be the same as the average product size of surveyed products. With little
information on the nonsurveyed products, the analysis implicitly assumed that the surveyed
products are representative of all products that are produced in the industry and thereby used this
information to calculate the national estimate of social cost.

       While the survey comprises about three-quarters of the total AIM coatings market
volume, the surveyed companies (116 in number) account for less than one-quarter of the more
than 500 companies reported to produce AIM coatings. As a result, the companies not
responding to the survey are likely to be small AIM coating producers, which would increase the
total number of nonsurveyed products subject to reformulation.

-------
        Based on survey responses from small companies, the data indicate that average product
 size for small company products is much smaller (about one-sixth the size) of the average size of
 all products in the survey.  If there exists a combination of the factors of:  (a) the nonsurveyed
 companies being small companies, and (b) small company products being lower in volume, then
 the ratio of market volume to survey volume understates the ratio of the number of market
 products to survey products. In other words, the survey potentially omits a relatively large
 number of small products.  In addition,  it is expected that some surveyed companies combined
 products to some extent for reporting purposes. The extent to which this occurred is not known,
 but would tend to understate the total number of products. As a result, the methodology used in
 the El A best represents a lower bound of national reformulation costs by not adequately
 accounting for the number of nonsurveyed products. To determine an upper bound of costs, the
 methodology can be modified to assume that all of the nonsurveyed products are small volume.
 Since the EPA has recently received information that indicates that not  all of the companies
 omitted from the survey can be considered small since some large companies were also omitted,
 this alternative measure of national cost would produce an upper bound of the estimate of social
 cost.

 Alternative Method

       For each of the 13 defined market segments in the AIM coatings industry there is data on
 total market volume derived from the Census of Manufacturers data for the baseline year (1991)
 and the total volume of surveyed products for that category.  From the  data, the total volume that
 is omitted from  the survey is computed as follows:

    Omitted volume = Market volume - Survey volume (by category)

 If the average size of omitted products is known, the number of omitted products can be
 estimated as follows.

    Omitted products = Omitted volume/Average volume of an omitted  product.

 If the proportion of omitted products needing reformulation is known, then the number of omitted
product reformulations can be computed as:

    Omitted product reformulations — Omitted products * Proportion of omitted products needing
                                  reformulation,

and the corresponding reformulation costs are then:

   Cost of Omitted Product Reformulations = Omitted product reformulations * Reformulation
                                          cost per product.

-------
 The national reformulation costs can then be computed as'

    National Reformulation Costs = Survey Product Reformulation Cost + Omitted Product
                                 Reformulation Cost.

 Since there is not specific data on nonsurveyed products, the average product volume is not
 available to directly determine the number of omitted products and thus the number of omitted
 product reformulations needed. However, the information from the survey can be used to impute
 the values for the nonsurveyed products. If the analysis assumes that: (1) the average size of the
 omitted products in each market segment equals the average size of small company products
 reported for that market segment in the survey, and (2) the omitted product reformulation rate is
 each  market segment equal the reformulation rate for small company products reported for that
 market segment in the survey data1, then the estimation of national costs will increase due to an
 increase in the number of product reformulations assumed to occur in the nonsurveyed
 population

        Table  1 presents the total number of reformulations resulting from this methodology and
 the total "static" reformulations costs for the entire AIM coatings industry, which refers to the
 cost resulting from the imposition of the reformulation costs on all products that exceed the limits
 set by the regulation without consideration of the potential for products to be withdrawn from the
 market. In Table 1, the ratio of estimated total AIM reformulations to total survey reformulations
 is approximately 2:1, which is in contrast to  the estimate derived from the original  methodology,
 based on total market volume to survey volume of 1.28.1.

       The social welfare cost estimate is also presented in Table 1, which captures the potential
 responses of the affected firms by allowing affected producers to either reformulate or withdraw a
 product from the market. In this way, producers can minimize the impacts of the regulation.  If it
 is less costly to the firm to withdraw the product than it is to reformulate, the presumption is that
 they will do so
Summary:
       In summary, modifying the cost aggregation methodology to bound the costs for
nonsurveyed AIM products raises the social cost estimate for the regulation to about $38.9
million, which is 55% higher than the lower bound estimate provided by the original analysis of
the Economic Impact Analysis. These estimates derived largely from modifying the methodology
to explicitly estimate the costs incurred by the nonsurveyed AIM coatings producer, under the
assumption that the omitted products are products by small companies and possess characteristics
similar the small company products reported in the survey.  Because reformulation costs are
incurred on a per product basis, rather than a per unit (volume) basis, the process of aggregating
       1       In addition, there is also a potential for a greater number of the nonsurveyed
products to exceed given VOC limits since small volume products are often used in specialty
(niche) markets  This bias would increase the estimate of national cost, however, since not all of
the nonsurveyed products are small volume products, or in specialty markets, it is likely that this
bias is captured within th euper and lower bound estimates provided in these analyses.

-------
the cost estimates from the survey population to the national population needs to be based on the
ratio of total products in the market to total products in the survey  The aggregation as
previously done used the ratio of market quantity for each category to survey quantity as the
aggregation factor.  Because the ratio of products exceeds the ratio of product quantities, the
previous cost estimate serves as a lower bound and the new estimate provides an upper bound of
national costs
                      Table 1. National Reformulation Costs under the
                                 Alternative Methodology
No. Reform-
ulations for
Surveyed
Products
1,111
No. Reform-
ulations for
Nonsurveyed
Products
1,234
Total
Reformulations


2,345
Total Static
Reformulation
Cost

$41.6 million
Total Social
Cost


$38.9 million

-------