745R98019
EPA 0090
NYlSl
Information is our business.
ECONOMIC ANALYSIS OF TOXIC SUBSTANCES
CONTROL ACT. SECTION 403: HAZARD STANDARDS
ABT ASSOCIATES, INC.
CAMBRIDGE, MA
MAY 1998
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
Tailored to Your Needs!
Selected
Research In
Microfiche
SRIM€ is a tailored information service that
delivers complete microfiche copies of
government publications based on your
needs, automatically, within a few weeks of
announcement by NTIS.
SRIM® Saves You Time, Money, and Space!
Automatically, every two weeks, your SRIM' profile is run against all new publications
received by NTIS and the publications microfiched for your order. Instead of paying
approximately $15-30 for each publication, you pay only $2.50 for the microfiche version.
Corporate and special libraries love the space-saving convenience of microfiche.
NTIS offers two options for SRIM8 selection criteria:
Standard SRIMs-Choose from among 350 pre-chosen subject topics.
Custom SRIM*-For a one-time additional fee, an NTIS analyst can help you develop a
keyword strategy to design your Custom SRIM" requirements. Custom SRIM5 allows your
SRIM* selection to be based upon specific subject keywords, not just broad subject topics.
Call an NTIS subject specialist at (703) 605-6655 to help you create a profile that will retrieve
only those technical reports of interest to you.
SRIM* requires an NTIS Deposit Account. The NTIS employee you speak to will help you set
up this account if you don't already have one.
For additional information, call the NTIS Subscriptions Department at 1-800-363-2068 or
(703) 605-6060. Or visit the NTIS Web site at http://www.ntis.gov and select SRIM* from
the pull-down menu.
U.S. DEPARTMENT OF COMMERCE
Technology Administration
National Technical Information Service
Springfield, VA 22161 (703)605-6000
http://www.ntis.gov
-------
ECONOMIC ANALYSIS
OF
Toxic SUBSTANCES CONTROL ACT
SECTION 403: HAZARD STANDARDS
Economic and Policy Analysis Branch
Economics, Exposure and Technology Division
Office of Pollution Prevention and Toxics
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
May 1998
-------
CONTRIBUTORS
The economist responsible for completion of this report was Nishkam Agarwal. Analytical and draft
preparation support was provided by Abt Associates Inc. of Cambridge, MA, and Bethesda, MD, under
EPA Contract Nos. 68-D2-0175, 68-W6-0021, and 68-W98-005.
-------
Table of Contents
Executive Summary. . . . . .. ........................ ............................. ES-1
Regulatory Background. . . . . . . . . . . . . . . . . . . . . . .. ........................ .... ES-l
Analytic Approach. . . .. .. .. .. ...................................... ES-2
Baseline. . . . . . . . . .. ........................... . . . . . . . . . . . . . . . . . . . . . .. ES-4
Estimating Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ..... .................. ES-5
Estimating Benefits. . . . . . .. .. ....................... .. . . . . . . . . . . . . . .. ES-5
Identifying Hazard Standards that Maximize Net Benefits. . . . . . . . . . . . . . . . . . . . . . . . .. ES-6
Sensitivity Analysis.. [[[ ES-8
Other Impacts of Section 403 ................................. ............... ES-9
1.
Introduction. . . . . . . . . .. ........ ............................ .. ...... .... 1-1
1.1 Background. . . . . . . . . . . .. ....................................... ... 1-1
1.2 Purpose of The Proposed Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. 1-1
1.3 Goals of The Economic Analysis. . .. ................................... 1-2
1.4 Organization of this Report. . . . . . . . . . .. ................................ 1-4
2. Regulatory Background. .. .. .. ..... ...... ................................ .. 2-1
2.1 Lead as a Public Health Problem.. ........ ......................... .. 2-1
2.2 Regulation of Lead Products, Environmental and Workplace Releases of Lead,
and Lead in Drinking Water. . . . . . .. .................................. 2-2
2.2.1 Lead in Paint.. .. .. ........ .................................. 2-2
2.2.2 Lead in Gasoline. . . .. ........................................... 2-2
2.2.3 Other Products Containing Lead. . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. 2-2
2.2.4 Environmental and Workplace Releases of Lead. . . . . . . . . . . . . . . . . . . . . . . .. 2-3
2.2.5 Lead in Drinking Water. . . . . . . . . . . .. ............... ............. 2-3
2.2.6 Resultant Reduction in Blood Lead Levels. . . . . . . . . . . . .. .............. 2-4
2.3 Regulatory Efforts to Reduce Lead-Based Paint, Dust and Soil in
Residential Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2-4
2.3.1 Current Estimates of Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2-4
2.3.2 Federal Regulatory Activities to Decrease Exposure to Lead-Based Paint in Existing
Housing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2-5
2.3.3 Federal Guidelines and Other Activities Related to Lead in Soil and Dust. . . .. 2-8
2.3.4 State and Local Programs to Reduce Exposure to Lead-Based Paint,
Dust and Soil. . . . . . .. ............................................ 2-9
2.4 Non-Regulatory Initiatives to Reduce Lead-Based Paint, Dust and Soil Exposure. 2-10
2.4.1. Joint Initiatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2-11
2.4.2 Blood Screening Programs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2-11
2.5 Benefits of Increasing Lead Awareness and Lead Poisoning Prevention Programs 2-11
-------
3-1
3. Problem Definition, Regulatory Options, and Overview of Analytic Approach. . . . . . ., ....., 3-1
3.1 LeadContaminationProblem................. ..... ........,. .....,
. . . . . . .' 3-1
3.1.1 Exposure Sources .................... .......... ..., 3-2
3.1.2 National Blood Lead Levels and Health Effects. . . . . . . . . . . . ., .....,. .' 3-3
3.2 Market Failure. . . . . . . . . . . . . . . . . .. ........ ..., .................... 3-6
....., .
3.3 Need for Federal Regulation. . . . . . . . . . . . . . . . .. .............,.. 3-8
.,..' .
3.4 Regulatory Options. . . . . . . . . . . . . . . . . . . . . . .. ................... 3-8
3.4.1 Information Provision. . . . . . . . .. ...... ... ............... .....,..
. .' 3-9
3.4.2 Other Regulatory Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., ...,.. 3-9
3.5 Overview of Analytic Approach .................. ....,.. ., .. """"3-11
3.5.1 Summary of Risk Assessment ...................... ...., 3-13
3.5.2 Linkages Between Risk and Economic Analysis. . ., .................. 3-15
3.5.3 Summary ofIntegrated Analysis. . . . . . . . . . . . . . . . . . . . . . . . .. .,. 3-23
Appendix 3. A. Brief Summary of Current OMB Guidance. . . . . . . . . . . ., ., 3-24
Chapter 3 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ............. .,
. . ., ...... 4-1
4. Cost Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., ........
4-1
4.1 Introduction. . . . . . . . . . . . . . . . . . . . .. ....,... . .....,. ., ...........
4.2 Estimating Aggregate Costs. . . . . . . . . .. .. . .. ............... ...,. . . 4-1
4.3 EPA and HUD Approaches to Estimating Costs. . . . . . . . . . . . . . . ., . ..,.... 4-4
4.3.1 EPAApproach .................. ......,. .......... ... .... .. 4-4
4.3.2 HUD Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... . ...... 4-6
4.3.3 Differences Between the Approaches. . . . . . . . . .. .....,. .,. ..,. ..... 4-6
4.4 Data Sources. . . . . . . . . . . . . . . . . . . . . .. ..... ................ ..,.. ... 4-7
4.4.1 Uncertainty. . . . . . . . . . . . . . . . . . . . . . . .. .... .........,. .. .. .... 4-8
4.4.2 Multifamily Housing Considerations. . . . . . . . . . . . .. ...... .. . ....... 4-8
4.5 Hazard Evaluation Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .... 4-9
4.5.1 Lead Hazard Screen.. ... ......................................... 4-10
4.5.2 Risk Assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....,... . 4-10
4.6 Intervention Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ... 4-11
4.6.1 Dust Intervention Costs. . . . . . . . . . . . . . . . .. ..............., ........ 4-11
4.6.2 Paint Intervention Costs. . . . . . . . . . . . . . . . . . .. ....... .. ........ .. 4-13
4.6.3 Soil Removal Costs. . . . . . .. ............. ......... ...... ....... 4-17
4.6.4 Overall Intervention Strategies. . . . . . . . . .. .. . . . .. ...... ...... 4-23
4.6.5. Enforcement Costs. . . . . . . . . . . . . . . . . . . . . . . . . .. ......... ......... 4-23
4.6.6 Implementation Costs. . . . . .. ............ ...... ....... . . . . . . . . 4-23
4.7 Number of Interventions and the Number of Housing Units that Exceed
the Candidate Hazard Standards .. .................. ... . . . . . . 4-23
4.7.1 Number of Homes Performing Interventions for Alternative Floor
Dust Standards. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ...... ............. 4-24
4.7.2 Number of Homes Performing Interventions for Alternative Windowsill Dust
Standards. . . . . . . . . .. ............ .... ............ . . . . . . . . . . . 4-25
4.7.3 Number of Homes Performing Interventions for Alternative Soil Standards... 4-28
4.8 Likely Rates of Intervention and Their Impact on the Cost Estimates. . . . . . .. 4-28
Appendix 4.A: Estimating Soil Removal Costs.. ... .............. ... . . . . . . . . 4-31
Chapter 4 References. . . . . . .. ............ .... ......................... 4-36
ii
~403 EA
-------
5. Benefits....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
5.1 Introduction ........... ........ ........ ....... ..... ...... ...5-1
5.2 Benefits as Reduced Exposure and Adverse Health Consequences. . . . . . . . .. 5-2
5.3 Valuation of Benefits.............. ......... ... . .. ... ........ .. 5-4
5.3.1 Valuing Changes in IQ Points. . . . . . . . . . . . . . . . .. .... ......... . . .. 5-5
5.3.2 Valuing Increased Educational Resources. . . . . . . . . . . . . . . . .. ......... 5-11
5.3.3 Valuing Increased Blood Lead Screening and Medical Treatment. . . . . .. 5 - 1 2
5.4 Aggregation of Benefits . . .. . . . . . . . . . . . . . . .. ............. . . . . . .. 5-14
Appendix 5.A Screening and Medical Costs for Risk Groups. . .. .. ...... .. .... 5-15
Appendix 5.B Aggregating Benefits from Environmental Lead Reduction. ............ 5-20
Chapter 5 References. . . . . . . . .. . .................... .,........... 5-23
6. Net Benefits. . . . . . . . . . .. ......... .. ... . . . . . . . . . . . . . . . . . . .. .. ........... . 6~ 1
6.1 Costs, Benefits and Net Benefits for Various Candidate Floor Dust Hazard Standards6-3
6.2 Costs, Benefits and Net Benefits for Various Candidate Window Sill Dust Hazard
Standards. .. """""'" .... ................,.. ................ 6-7
6.3 Costs, Benefits and Net benefits for Various Candidate Soil Hazard Standards. .. 6-11
6.4 Hazard Standards that Maximize Net Benefits. . . . . . . . . . . . . . . . . . . . . . . 6-15
Appendix 6A. Net Benefits with Paint Intervention and Testing Excluded. . . . . . . . . . . .. 6-18
Chapter 6 Reference. . . . . . . .. .......... ............. ..................... 6-20
7. Sensitivity Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-1
7.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-1
7.2 Analyses Involving Parameter Changes.. .............. ....... ... 7-2
7.2.1 Discount rate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7 - 2
7.2.2 Value of an IQ Point. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-4
7.2.3 Hazardous Waste Disposal of Soil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-5
7.3 Analyses Involving Changes in Modeling Procedure. . . . . . . . . . . . . . . . . . . . . . . .. 7-7
7.3.1 Benefits from SmallIQ Changes. .. .. . .. .. . . . . . . "" .. '"'''''''''' 7-7
7.3.2 Transaction Triggerfor Interventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-10
7.3.3 Single Medium Analysis. . . . . . . . . . . . . . . . . . . . . . . . . .. ..... ..... ... 7-12
7.4 Additional Elements of Uncertainty. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-14
7.4.1 Unit Costs ofInterventions . . . . . . . . . . . . . . . .. ....... . . . . . . . . . . . . .. 7-15
7 .4.2 Valuation of Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7-16
7.4.3 Other Modeling Issues. . . . . . . . . . . . . . . . . . . . . . . .. ... .............. 7-16
Chapter 7 References. . . . . . . . . . . . . . . . . . . .. ......... ............. ..... ... 7 -19
8. Supplementary Analyses.. .................... .................................. 8-1
8.1 The Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement
Fairness Act (SBREFA) ............................... .. ............ 8-1
8.1.1 Impact of ~403 on the Lead Testing and Abatement Industry. . . . . . . . . . . . . .. 8-1
8.1.1.1 Major Findings. . . . . . .. .......................,. ............. 8-2
8.1.2 Impact of ~403 on the Rental Real Estate Sector. . . . . . . . . . . . . . . . . . , . . . .. 8-2
8.1.2.1 Definition of Small Entity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 8-4
8.1.2.2 Expected Impacts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8-4
~403 EA
iii
-------
8.2 Unfunded Mandates Refonn Act (UMRA) ... ........., ................ . 8-6
8.3 Paperwork Reduction Act (PRA) . . . . . . . . . . . . .. ......................... 8-6
8.4 Executive Order 12898 Federal Actions to Address Environmental Justice. . . . . . .. 8-7
8.4.1 The Distribution of Costs and Benefits by Race and Income. . . . . . . . . . . . . .. 8-8
8.4.2 Anticipated changes in results if assumptions are relaxed. . . . . . . . . . . . . . . .. 8-10
8.5 Executive Order 13045--Protection of Children from Environmental Health Risk
and Safety Risks. . . . .. ................ . . . . . . . . . . . . . . . . . . . .., . .. 8-13
Chapter 8 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8-15
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ........................ R-1
iv
~403 EA
-------
List of Exhibits
Exhibit ES.l: Summary of Resu1ts for Proposed ~403 Standards. .. .............. . . . . . .. ES-2
Exhibit ES.2: Linkages Between the Risk Assessment and Economic Methodologies. . . . . . . . . . . ES-4
Exhibit ES.3: Comparison of Standards Under A1temative Risk Assessment Models. . .. ....... ES-7
Exhibit ES.4: Summary of Resu1ts of Sensitivity Analysis. . .. .. ...... . . . .. ES-9
Exhibit ES.5: Beneficial Health Impacts on Children Resulting from ~403 .... . .. .......... ES-ll
Exhibit 3.1: The Demand for Abatements under Alternative Information Scenarios . . . . . . . .. 3-4
Exhibit 3.2: Other Regulatory Alternatives. . . . . . . . . . . . . . . . . . . . . . . . . .. ...... .. .. 3-10
Exhibit 3.3: Linkages Between the Risk Assessment and Economic Methodologies. . . . . .. .. .. 3-14
Exhibit 3.4: Duration for Each Intervention. . . . . . . . . . .. ................... ............ 3-17
Exhibit 3.5: Post-Intervention Ambient Conditions. . . . . . . . . . . . . . . . . . . . . .. .. ......... 3-18
Exhibit 4.1.A: Determining Total Costs - Outline of Calculation. . . . . . . . . . . .. .. .. ....... 4-3
Exhibit 4.1.B: Determining Total Costs - Estimating Present Value of Aggregate Costs. . . . . . . .. 4-5
Exhibit 4.2: Summary of Lead Hazard Evaluation Costs. . . . . . . . . . . . . . . . . . . .. .. ......... 4-10
Exhibit 4.3: Summary of Intervention Costs for Lead in Dust, Paint, and Soil. . . .. .. .. 4-12
Exhibit 4.4: Unit Cost Estimates for Soil Removal - Single-Family. . . . . . . . . . . .. .. ......... 4-18
Exhibit 4.5: Soil Abatement Costs - Single-Family Home. . . .. ................. .. ..... 4-19
Exhibit 4.6: Unit Cost Estimates for Soil Removal- Multifamily. . . . . . . . . . . . . . . . . . . . . . . .. 4-21
Exhibit 4.7: Soil Abatement Costs - Multifamily Building of 30 Units. . . . . . . . . . .. ......... 4-22
Exhibit 4.8 : Number of Homes Performing Interventions (Over Model Period) . . . . . . . .. ..... 4-26
Exhibit 4.9: Number of Homes Performing Interventions (Over Model Period) ............... 4-27
Exhibit 4.10: Number of Homes Performing Interventions (Over Model Period) ..... ... .... 4-29
Exhibit 4.11: Costs Under Alternative Assumptions About Intervention Rates: Proposed Standards 4-30
Table A.l: Unit Cost Estimates for Soil Abatements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . " 4-31
Table A.2: Median Lot Sizes (sq. ft.) from the American Housing Survey. . . . . . . . . . . . . . . . . . . .. 4-33
Table A.3: Median Footprint Sizes (sq. ft.) Derived from the American Housing Survey. . . . . .. 4-34
Table A.4: Perimeter Area Sizes (sq. ft.) Derived from the American Housing Survey. . . . . . . .. 4-34
Table A.5: Remote Area Sizes (sq. ft.) Derived from the American Housing Survey. . . . . . . . . . .. 4-35
Table A.6: Weighted Average Perimeter Areas and Remote Areas By Geographic Location. . . . .. 4-35
Exhibit 5.1: Summary of Post-Intervention Conditions for Various Intervention Alternatives.. ... 5-3
Exhibit 5.2: Summary of Benefits Analysis Estimate. . . . .. ........................ ...... 5-6
Exhibit 5.3: Risk Groups and Associated Screening and Medical Costs Per Child . . . . . . .. 5-13
Exhibit 6.1: Costs, Benefits and Net Benefits for Alternative Floor Dust Standards,
With Other Media Set at Proposed Standards. . .. """"""""""""""'" 6-4
Exhibit 6.2: IEUBK-based Model Costs, Benefits, and Net Benefits. . . . . . .. .. ............. 6-5
Exhibit 6.3: Empirical-based Model Costs, Benefits, and Net Benefits. . . . . . . . . .. ..... 6-6
Exhibit 6.4: Costs, Benefits and Net Benefits for Alternative Window Sill Dust Standards,
With Other Media Set at Proposed Standards. . . . . . . . . . . . . . . . . . . . . . . . . . .. ........ 6-8
Exhibit 6.5: IEUBK-based Model Costs, Benefits, and Net Benefits for Alternative
Window Sill Dust Standards. . . . . . . . . . . . . . . . . .. ..... ...... ............ ..... 6-9
Exhibit 6.6: Empirical-based Model Costs, Benefits, and Net Benefits for Alternative
Window Sill Dust Standards. . . . . . . . . . . . . . . . . .. ..........
Exhibit 6.7: Costs, Benefits and Net Benefits for Alternative Soil Standards,
With Other Media Set at Proposed Standards. . . .. ............... """""'" 6-12
. . . . .. ... 6-10
fi403 EA
v
-------
Exhibit 6.8: IEUBK-based Model Costs, Benefits, and Net Benefits for Alternative
. 6-13
Soil Standards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Exhibit 6.9: Empirical-based Model Costs, Benefits, and Net Benefits for Alternative
Soil Standards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., ..,. 6-14
. . . . . . . . . . . . . . . . . . . . . 6-17
Exhibit 6.10: Comparison of Standards Under Alternative Risk Assessment Models. . . . . . . . . . . .
Exhibit 6.A.1: Costs, Benefits and Net Benefits for Alternative Floor Dust Standards,
With Other Media Set at Proposed Standards. . . .. ...............
Exhibit 6.A.2: Costs, Benefits and Net Benefits for Alternative Window Sill Dust Standards,
With Other Media Set at Proposed Standards. . . . . . . . . . . . . . . . . . . . . . .. ., 6-19
Exhibit 6.A.3: Costs, Benefits and Net Benefits for Alternative Soil Standards,
. . ., .. ..,. 6-19
With Other Media Set at Proposed Standards. . . .. ............... ..
Exhibit 7 -1 a: Effects on Costs and Benefits of Proposed Standards due to Changing Discount Rate
Assumption. . . . . . . . . . .. ........... . .... ...... ................... .,.. 7-3
Exhibit 7-1b: Effects on Net Benefit-Maximizing Standards due to Changing Discount Rate
Assumption. . .. .........,......... ...... ....................... .. .,... 7-3
Exhibit 7 -2a: Effects on Costs and Benefits of Proposed Standards due to Changing IQ Valuation
Assumption. . . . . . . .. ............... . . .. ............... ..... ......... 7-4
Exhibit 7-2b: Effects on Net Benefit-Maximizing Standards due to Changing IQ Valuation
Assumption. . .. ...... .......................................... .. .. .. 7-5
Exhibit 7-3a: Effects on Costs and Benefits of Proposed Standards due to Changing Assumptions
Regarding Whether Removed Soil Must Ever Be Treated as Hazardous Waste.. ..... .. 7-5
Exhibit 7-3b: Effects on Net Benefit-Maximizing Standards due to Changing Assumptions
Regarding Whether Removed Soil Must Ever Be Treated as Hazardous Waste. . . . . . . . . . 7-6
Exhibit 7-4: Effects on Costs and Benefits of Not Counting Benefits from Individual IQ
Point Changes of Less than One. . . . . . . . . . .. ...................... ..... ..... 7-9
Exhibit 7-5: BaselinelPost-Intervention Blood Lead Difference by Percentile Group for
HUD Home 411207 . . . . . . . . . . . . . . . . .. .......................... .. ........ 7 -10
Exhibit 7 -6a: Effects on Costs and Benefits due to Changing Assumption about
Intervention Trigger. . . . . . . . . . . . . . . . . . . . .. .. ............................ 7-11
Exhibit 7-6b: Effects on Net Benefit-Maximizing Standards due to Changing Assumption about
Intervention Trigger. .. .. ................ ... . . . . . . . . . . . . . . . . . . . . . . . . . .. 7 -12
Exhibit 7-7: Net Benefit-Maximizing Standards in a Single Medium Analysis. . . . . . . . . . . . . . . .. 7-12
Exhibit 8.1: Characteristics of Establishments. . . . . . . . . . .. .. ............................ 8-3
Exhibit 8.2: Frequency of Lead Interventions in Multi-Family Housing. . . .. ............... 8-5
Exhibit 8.3: Ratio of Annual Compliance Costs to Annual Rent Payments, by size of business ..... 8-6
Exhibit 8.4: Distribution ofIntervention Costs and Benefits by Race and Income. . . . . . . . . . . . . .. 8-10
Exhibit 8.5: Birth Rates per 1,000 Households by Race. . . . . . . . . . . . . . . . . . . . . . .. """"" 8-11
Exhibit 8.6: Characteristics of Households Living in Pre-1978 Housing and Pre-l 978 Housing
with Interventions. . . .. ............ ..... .. ............................... 8-12
Exhibit 8.7: Beneficial Health Impacts on Children Resulting from ~403 .................. .. 8-14
. . . . ., 6-19
vi
i403 EA
-------
Executive Summary
This report presents an economic analysis of proposed regulations setting standards for lead-based paint
hazards. The regulations are being proposed under authority of 9403 of the Toxic Substances Control
Act (TSCA). This section was established by the Residential Lead-Based Paint Hazard Reduction Act of
1992, also known as "Title X. "
TSCA 9403 requires that EPA promulgate regulations to "identify... lead-based paint hazards, lead-
contaminated dust and lead-contaminated soil" for purposes of other parts of Title X. The lead-based
paint hazards addressed in this economic analysis include residential hazards from deteriorated paint and
contaminated dust and soil! These standards apply directly to "target housing" (most housing
constructed prior to 1978). In addition, various parties may also apply the standards to newer
residences.
The analysis compares alternative candidate standards in terms of their net benefits. Net benefits are
based on the benefits of risk reduction minus the costs of control activities needed to achieve the
reduction in risk. The benefit categories all measure health effects resulting from childhood lead
exposure. The analysis calculates net benefits for a wide range of alternative standards, including the
proposed 9403 hazard levels.
The analysis does not attempt to predict precisely how much remediation of residential lead-based paint
hazards will occur as a result of promulgating these standards. It is designed to provide comparisons of
different standards rather than absolute measures of costs and benefits for the different standards.
The proposed 9403 standards define the level of lead in soil and dust, and the condition of lead-based
paint, at which intervention activities should be undertaken. The proposed standards are listed in Exhibit
ES-l along with their estimated costs, benefits, net benefits, the number of homes affected, and number
of children affected. There are two separate estimates of benefits, and thus net benefits, because two
models were used to estimate differences in blood-lead levels resulting from differences in environmental
lead levels due to the proposed standards. Since these two models predict different blood-lead levels,
they result in different estimates of benefits.
Regulatory Background
Lead's advantageous properties, including its malleability, resistance to corrosion, good insulating
properties, and low cost, have made it attractive for many applications; lead has been used in gasoline,
ceramics, paint, and many other products. These uses have resulted in lead's release to and distribution in
all environmental media, which has complicated efforts aimed at reducing lead in the environment.
The proposed rule also defines "levels of concern." Levels of concern are set at lower levels of
contamination than the hazard standards, and are intended for use in risk communication. They were
based on health risks only and are not part of this economic analysis.
*403 EA
ES-1
-------
Exhibit ES-1
Summary of Results for Proposed ~403 Standards
Interior Paint Standard
1995 HUD Guideline for
deterioration
Exterior Paint Standard
1995 HUD Guideline for
deterioration
Floor Dust Standard
50 ~g/sq ft
250 ~g/sq ft
2,000 ppm
Window Sill Dust Standard
Soil
...!. ~~~~. ~-~.~.t. .~~~~~. ~~. y.~~:. .~~~.~.'. .~.i_~~~.~-~:_~-~. .~-~-~"I~~_.u... ---.. - -.. u_..... -....... u -. u~-~-~:~_~~I.I_i~_~u - u -. - -. - u.- - u. u. - -. u u
Total Benefits based on the IEUBK Model (over 50 year span,
discounted at 3%)
Net Benefits, based on IEUBK Model (over 50 year span, discounted
at 3%)
--------------------------------------------------------------_.-----.._----------------------------____-.0------------ --------------------------------------------------
$160.1 billion
$107.2 billion
Total Benefits, based on Empirical Model (over 50 year span,
discounted at 3%)
Net Benefits, based on Empirical Model (over 50 year span,
discounted at 3%)
.---------------------------------------------------------------------------------------------------------------------- -.---------.--------.--.--------------------------
$42.4 billion
-$10.5 billion
Number of Homes that Exceed the Standards
25.4 million
- - - - - - - - - - - - - - - - - - - - - - - - - -.. - *. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -.. -. - ~... -.. - - - -. - - - - - - - - - - - - - -..... -.... - - - ~ - - - - - - -. - - - -. - -. -. -....
Number of Children who will experience reduced exposure to
household lead in soil, dust, and paint
43.8 million
As our understanding of the negative health effects of lead has increased, a variety of federal, state and
local regulations have been developed to reduce exposure to lead. The presence of lead in certain
consumer products, su"h as gasoline, has been prohibited or restricted by regulations. Environmental
releases of lead to air and water, and lead concentrations in waste, also have been controlled. OSHA has
set limits on allowable workplace concentrations of lead. In addition, regulatory efforts have been made
to remediate exposure to lead through drinking water systems.
One of the largest remaining lead exposure sources for children is existing reservoirs of lead in paint,
dust and soil present in many residential areas. In an effort to reduce exposure to residential lead
hazards, regulatory efforts to address these hazards have been increasing for several years.
Analytic Approach
As envisioned by Congress, the lead hazard standards to be established under 9403 are intended to tell
people when they should act for the safety of their children, i.e. when interventions should take place.
As such, the analysis identifies hazard levels at which the interventions maximize net benefits. More
specifically, in defining when intervention actions should occur, the relevant measure is maximizing net
benefits for children (the population that is both at greatest risk from exposure and the most in need of
protection since children are not in a position to protect themselves). Therefore, the birth of a child is
used as the event that triggers intervention activities in the analysis used to evaluate alternative standards.
ES-2
~403 EA
-------
The factors considered in estimating the costs and benefits include:
.
The adverse health effects to children from lead exposure -- what they are and how they differ
with differing lead-related characteristics of the housing unit, and across demographic,
geographic and/or socio-economic groups.
.
The costs to individuals and society of interventions which reduce lead exposure.
.
The effectiveness and duration of these interventions and the resulting benefits in tenns of
reduced exposure and reduced adverse health effects.
.
The value of these benefits to individuals and society.
By comparing the total present value of net benefits (benefits minus costs) for a reasonable array of
standards, it is possible to identify the standard that yields the largest net benefits and to compare
standards in terms of their net benefits. For example, as standards become more protective (i.e. more
stringent), they also become more costly. By looking at net benefits, the analysis shows how much
society is giving up to acquire more protection. Likewise, costs can be reduced by making the standards
less stringent. The analysis estimates what benefits society is giving up with this reduction in costs.
The analysis considers children, from birth through the sixth birthday, living in homes that were built in
1997 or earlier.2 Because the potential for lead exposure in currently contaminated homes may remain
for some time, the analysis considers children born during the next 50 years, from 1997 to 2046.
Exposure, health effects, and benefits are calculated separately for the cohorts born in each of the 50
model years.
The methodology used to evaluate the economic return from the 9403 hazard levels has several linkages
with the risk assessment methodology, which in turn relates paint condition and the amount of lead in
dust and soil to blood lead levels and health effects. Exhibit ES-2 illustrates the basic linkages between
the risk assessment and economic modeling. The primary links are those that connect candidate hazard
standards and the presence of lead in each housing unit with hazard control choices and costs, and those
that connect the presence of environmental lead to blood-lead levels and thus to health effects and
economic damages and benefits. The first set of linkages outlines the cost estimation process, while the
latter describes the benefit estimation process.
In Exhibit ES-2, the boxes along the top represent the analysis of baseline conditions, yielding an
estimate of the economic damages resulting from the baseline residential lead levels. The boxes along the
bottom of the exhibit represent the analysis of ex post conditions; each scenario or potential lead hazard
definition is analyzed separately. A comparison of the baseline economic damages and the ex post
economic damages yields an estimate of the benefits of actions perfonned under the scenario. This is
According to Title X, the regulations will apply to housing constructed before 1978. However, the
analysis assumes that once EPA promulgates these standards, they will be generally applied to all housing
regardless of year constructed. Of the over 25 million homes that exceed the proposed standards, only an
estimated 890,000 (or 3.6 percent) were built after 1978.
~403 EA
ES-3
-------
represented by the far-right box in the middle row of Exhibit ES-2. From these benefits, the analysis
subtracts the corresponding costs to get net benefits.
Exhibit ES-2
Linkages Between the Risk Assessment and Economic Methodologies
Baseline
Ambient
Conditions
Baseline
Blood Lead
Distribution
Baseline
Human Health
Effects
Baseline
Economic
Damages
Hazard
Control
Costs
NET
BENEFITS
Benefits
Post-Action
Ambient
Conditions
Post-Action
Blood Lead
Distribution
Post-Action
Human Health
Effects
Post-Action
Economic
Damages
The evaluation of candidate hazard standards consists of calculating net benefits for a wide range of
levels for the candidate standards and then comparing the candidate hazard standards in terms of net
benefits. The candidate hazard standard yielding the largest net benefits provides the greatest benefit to
society.
Baseline
The benefit-cost analysis compares alternative futures over a 50-year time span: a baseline or "no-action"
alternative for which it is assumed that no changes are made to current ambient lead exposure conditions,
and a "post-action" alternative for which it is assumed that the ambient lead exposure conditions are
reduced in specific ways in response to the 9403 standards. In other words, it is a marginal analysis with
a baseline of no intervention. The baseline residential lead conditions are defined by data on dust and soil
lead levels, and condition of lead-based paint, collected by the U.S. Department of Housing and Urban
Development for a representative sample of homes. The baseline population blood-lead levels are
defined by data from the NHANES III Part 2 survey.
ES-4
~403 EA
-------
Estimating Costs
Costs of hazard control are calculated using unit costs of individual intervention activities, in
combination with the timing and number of interventions. Unit costs are calculated for each of the
testing and intervention activities and represent average costs of intervention nationwide. Intervention
activities include maintaining, removing or encapsulating deteriorated lead-based paint, removing dust,
and removing and replacing soil. Separate cost estimates are developed for single and multi-family
housing units; multi-family unit costs are developed by adjusting the single family cost estimates to
reflect the smaller size of multi-family units and the smaller yards of multi-family units.
When a child is born into a housing unit whose ambient lead levels exceed the candidate standards levels,
the appropriate intervention is assumed to occur. This initial intervention is repeated as necessary until
the child turns six years of age, at which time additional interventions will cease unless an additional
child has been born during this time period. If needed, interventions in the home will recommence if
another child is born after this period, and will follow the same repeat routine.
Costs incurred after the first year are discounted to the first year using an annual discount rate of 3
percent.3 The total cost estimate is the sum of the costs of hazard controls in all homes for each year and
represents the present value of the assumed stream of intervention costs.
Estimating Benefits
Benefits of hazard control are calculated using estimates of "avoided" economic damages corresponding
to avoided adverse health effects. The model defmes "avoided" as the difference between the baseline
scenario, which assumes no intervention activity and various intervention or ex post scenarios. Each of
the scenarios assumes a different specification of lead hazard standards, and hence intervention activities.
In the analysis, benefits are calculated for children whose exposure to lead is reduced for the period from
birth to age six.
These avoided economic damages include reductions in IQ, plus increased educational and medical costs
connected with high levels of exposure. In each case, the economic value is a proxy for society's
willingness to pay to avoid the health effect. Changes in IQ levels make up the vast majority of benefits
in this analysis. The economic value of avoiding lost IQ points is approximated by using an estimate of
the foregone lifetime income due to IQ point loss. The estimated value per IQ point lost is $8,346 (1995
dollars ).
Since lead exposure has been linked to a variety of health hazards for children and adults in addition to
those analyzed here, the benefit estimates are likely to understate the societal return from the 9403 hazard
levels. Furthermore, secondary benefits such as improved energy efficiency due to new windows and
increased aesthetic appeal due to repainting are not included.
Benefits accrue over time depending on hazard control choices and assumptions regarding exposure of
children. All benefit estimates are discounted to the present using an annual rate of 3 percent. Total
The sensitivity analysis presents costs, benefits and net benefits calculated with a discount rate of 7 percent
as well.
~403 EA
ES-5
-------
benefits are the sum of benefits calculated for each year or cohort of children protected and represent the
present value of the stream of benefits from the hazard controls. Net benefits are simply the difference
between the total benefits estimate and the total cost estimate. As such, they are an indicator of the
societal gains from hazard controls.
Identifying Hazard Standards that Maximize Net Benefits
The analysis evaluates alternative standards for floor dust, window sill dust and soil, assuming the
following responses to the paint standards:
Amount of Deteriorated
Medium Lead-Based Paint Intervention Activity
10 sq.ft. or more Repair
..--------.-.....---......-.--------------.------- ------______0.-----------'.---------------
Deteriorated Interior Lead- Based Paint 50 sq.ft. or more Abate
20 sq.ft. or more Repair
-------------------------------------------------- -------------------.--.-------------------
Deteriorated Exterior Lead- Based Paint: 100 sq.ft. or more Abate
The analysis assumes that homes receiving any lead intervention will receive interventions for all media
(floor dust, window sill dust, paint and soil) that exceed the standards. This assumption, combined with
the fact that there are interactions among the interventions in terms of both costs and benefits, means that
standards can not be accurately evaluated one at a time. Instead, the standards for a single medium must
be evaluated in the context of specified standards for all other media. To allow for this, the analysis
calculated costs, benefits and net benefits for all combinations of standards. For each medium, the
alternative standards were defined in terms of incremental changes in the levels of lead. For example,
floor dust standards varied by increments of 10 jlg/ft2 (e.g., 40 jlg/ft2, 50 jlg/ft2, 60 jlg/fe, 70 jlg/ft2,
etc.). Likewise, soil standards were analyzed in increments of 50 ppm (150 ppm. 200 ppm. 250 ppm,
300 ppm, etc.). All combinations of standards were analyzed.
Two EPA blood-lead models (IEUBK and Empirical) were used in this analysis and they generate
different benefit estimates for any given combination of standards. In addition to differences in the
overall size of benefits, the benefit estimates vary at different rates under the two models and thus the set
of standards that maximize net benefits is different under the two models. As a result, there is no unique
answer to the question: which combination of standards maximize net benefits? There is one answer
when benefits are estimated using the IEUBK Model and a different answer when the benefits are based
on the results of the Empirical Model.
The two blood-lead models differ in several ways, including the incorporation of different variables and
very different functional forms relating environmental lead levels to children's blood lead levels.
Notably, the functional form of the IEUBK Model is such that it is much more sensitive to changes in
environmental lead than the Empirical Model. Also, the IEUBK Model uses lead dust concentrations and
the Empirical Model uses dust lead loadings as input variables. Since dust lead concentrations and
loadings are not well correlated in the actual housing unit data collected by HUD and used for this
ES-6
!403 EA
-------
analysis, these differences in input variables result in differences in estimated blood-lead changes and
thus benefits.
For each of the two blood-lead models, Exhibit ES-3 presents the set of standards that maximize net
benefits, along with the costs, benefits and net benefits for each standard. The net-benefit maximizing
standards, based on the IEUBK Model, are much more stringent than the net benefit maximizing
standards based on the Empirical Model. In addition, the exhibit presents the standard that EP A is
proposing, along with its cost, benefit and net benefits. The proposed standard is shown twice, once with
benefits calculated using the IEUBK Model and once using the Empirical Model. The top half of the
table presents results using the IEUBK Model.
Exhibit ES-3
C
omparison of Standards Under Alternative Risk Assessment Models
IEUBK Model Results
Standards that
Maximize Net Benefits i Proposed Standards
Floor Dust Standard 40 ~ g/ft2 50 ~g/ft2
Window Sill Dust Standard 1 00 ~ g/ft2 250 ~g/ft2
Soil Standard 250 ppm 2,000 ppm
Total Cost $100.4 billion $52.8 billion
Total Benefit $273.6 billion $160.1 billion
Net Benefit $173.2 billion $107.2 billion
Empirical Model Results
Standards that
Maximize Net Benefits i Proposed Standards
Floor Dust Standard 80 ~ g/ft2 50 ~g/ft2
Window Sill Dust Standard 31 0 ~g/ft2 250 ~g/ft2
Soil Standard 4,350 ppm 2,000 ppm
Total Cost $44.0 billion $52.8 billion
Total Benefit $35.1 billion $42.4 billion
Net Benefit -$8.9 billion -$1,0.5 billion
The IEUBK net benefit maximizing standards are more stringent than the proposed standards. Because
of the large number of homes in the lower range of environmental lead levels, the IEUBK standards
would cost nearly twice as much as the proposed standards and the benefits would be nearly 1.7 times
those of the proposed standards. In addition, the net benefits, at $173 billion, would be substantially
higher than the net benefits of the proposed standards, at $107 billion.
The Empirical Model net benefit maximizing standards, on the other hand, are less stringent than the
proposed standards. Since there are many fewer homes with environmental lead levels in the range of
these standards, the Empirical Model net benefits maximizing set of standards would cost less ($44
billion as compared to $53 billion) and would produce smaller benefits ($35 billion as compared to $42
billion) than the proposed standard. However, its net benefits are somewhat larger than those of the
proposed standard, while still negative.
t403 EA
ES-7
-------
Sensitivity Analysis
To gain a better understanding of the relationships embedded in the analysis, and the impact of certain
parameter values, six sensitivity analyses were performed. The particular elements of the model chosen
to include in the sensitivity analysis reflect those elements identified as likely to have significant effect on
the results or for which there was a particular interest in determining what the potential effects might be.
Three of the sensitivity analyses dealt with alternative values for specific parameters:
.
Discount rate - 7% in place of 3%;
Monetary value of an IQ point loss or gain - $6,847 in place of $8,346;
Elimination of the additional cost of disposing of high-lead soil as a hazardous waste.
.
.
The other three sensitivity analyses involved changes in the modeling procedures used in the benefit-cost
analysis:
.
Exclude from the benefits small changes in IQ;
Use real estate transactions, rather than pending birth, as the intervention trigger;
Analyze standards for each medium, assuming there were no standards for other media.
.
The sensitivity analyses consider the effect on two outcomes of the benefit-cost modeling. The first
outcome is the impact on the estimated costs and benefits of the standards. The second outcome is the
effect on the determination of the set of standards that produce maximum net benefits. In some cases,
both cost and benefit estimates are changed but there is no shift in the standard found to maximize net
benefits. In other cases, there are changes in estimated costs and/or benefits that affect which standard
maximizes net benefits. In all cases, the sensitivity analyses are conducted separately using the IEUBK
and the empirical blood-lead models.
The results of the six sensitivity analyses are summarized in Exhibit ES-4. In the first three analyses
costs and/or benefits are decreased; where there are changes in the net benefit maximizing standards, they
become less stringent. The largest changes in the standards result from increasing the discount rate; the
other two cases experience relatively small shifts in the standards that maximize net benefits.
The impacts of the other three sensitivity analyses are mixed. Excluding small changes in IQ reduces
benefits while leaving costs unchanged. The net-benefit maximizing standards become less stringent.
Assuming that real estate transactions, instead of births, trigger intervention activities, increases costs
and decreases benefits because more actions take place but fewer occur where there is a child to benefit.
Again, the impact is to reduce the stringency of the net-benefit maximizing standards. Due to the
potential for double counting, neither costs nor benefits can be accurately estimated when standards are
analyzed for one medium at a time. The impact on standards that maximize net benefits is to make the
floor and soil standards more stringent when the Empirical Model is used to measure benefits, and to
leave the rest of the standards unchanged.
ES-8
~403 EA
-------
Exhibit ES-4
S fR
ummary 0 esults of Sensitivity Analysis
Factor Examined Impact on Impact on Impacts on Standards That
Estimation of Estimation of Maximize Net Benefits
Costs Benefits
Discount Rate: 7% in Decrease in costs Decrease in Standards substantially less
place of 3% relatively less than benefits relatively stringent
decrease in benefits more than
decrease in costs
------...----------...----------------....- _."'--0.-.------------------------ .-------------.".-----..---"---.- ---..--------.---------------------------------.------."-
Value of IQ Point: No impact on costs Decrease in IEUBK: No change in standards
decrease from $8,346 benefits Empirical: Standards slightly less
to $6,847 stringent
---------------------.........--...-------- ------------------...........-..... ........--------------.---------- --------------................-...-----------------------
Hazardous Waste Decrease in costs No impact on IEUBK: Soil standards slightly
Disposal of Soil: not benefits less stringent
required Empirical: Soil standards more
stringent
-------------------------..---...---------- ______0__"'_---------------------- --------------------------------- ____0-_____,--,----.""---------------------------------
Exclude Small No impact on costs Decrease in Standards would become less
Changes in IQ (i.e., benefits stringent, with larger impact on
less than 1 IQ point) Empirical than on IEUBK results
from Benefits
------.-------------..-...........-........ ................................... ...............-..........-....-. ...............-._...-...-..._.-...----------------------
Event That Triggers Increase in costs Decrease in IEUBK: Floor dust and soil
Interventions: use real (annual transaction benefits (most standards unchanged; window sill
estate transaction rates rate greater than interventions occur dust standard becomes less
in place of birth rates annual birth rate) where no child is stringent
present to benefit) Empirical: all standards become
less stringent
- - - - - - - - - - - -.. - -.. - -. ~ - -. - -.. ~.... -.. -. - - - - ---........---.----..-.---...--...- .... - -.... - - - - - - -. - -.. ~ - - - - - - - - - - --.-...-....---..----------------------------------------
Single Medium NA NA IEUBK: No change in standards
Analysis: standard for Empirical: Floor dust and soil
each medium set standards become more
assuming no other stringent
standards in effect
Other Impacts of Section 403
In addition to the benefit-cost analysis, several other types of impacts are important to consider in
evaluating a regulation. These additional analyses include the impact on small entities, minority and low-
income groups, and children, as well as the burden placed on state, local and tribal governments and the
private sector. In addition, the costs of complying with paperwork requirements for g403 were
considered. The results of these analyses are presented below.
The Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement
Fairness Act (SBREFA)
As described in the Preamble, the g403 standards do not require or mandate any actions by homeowners,
landlords, or personnel perfomring lead-based paint identifications and interventions. Instead, g403
standards infoTI11 decision-makers about what conditions constitute a hazard and recommend potential
actions. As a result, EP A is not required to conduct an analysis of small entity impacts under the
Regulatory Flexibility Act (RFA), as amended by the Small Business Regulatory Enforcement Fairness
~403 EA
ES-9
-------
Act (SBREFA). The RFA requires analysis of a rule's economic impact on the small entities that will be
subject to the rule's requirements. It requires that the analysis identify the types, and estimate the
numbers, of small entities "to which the proposed [or final] rule will apply," and describe the rule
"requirements" to which small entities "will be subject" and any regulatory alternatives, including
exemptions and deferrals, which would lessen the rule's burden on small entities. (Sections 603 and 604
of the RFA.) Rules that do not establish requirements applicable to small entities are thus not susceptible
to RFA analysis and may be certified as not having a significant economic impact on a substantial
number of small entities, within the meaning of the RFA. This is particularly true when the national
standards do not themselves require any particular action, as is the case with ~403.
Nevertheless, EP A has conducted a more limited analysis of the potential impact on small entities of
these standards as they work within the market. Two groups of entities are considered: lead-based paint
inspection and abatement firms, and landlords. To the extent that the ~403 standards may increase the
number of hazard identification and intervention actions, this is likely to help small businesses, who make
up the majority of inspection and abatement firms. In terms of impacts on landlords, the analysis found
that there would not be a significant economic impact on a substantial number of small firms.
Unfunded Mandates Reform Act (UMRA)
Under Title II of the Unfunded Mandates Reform Act, the cost to state, local and tribal government or the
private sector of compliance with federal regulations must be calculated and considered during the
regulatory process. Because ~403 is a regulation which provides information to consumers about
household lead safety and does not require households or public entities to take any action with respect to
that information, no costs are imposed on state, local and tribal governments or the private sector. As
such, this action is not subject to the requirements of sections 202 and 205 of (UMRA) because this
action does not contain any "federal mandates." Similarly this regulation contains no regulatory
requirements that might significantly or uniquely affect small governments, so no action is needed under
Section 203 of UMRA.
Paperwork Reduction Act (PRA)
The Paperwork Reduction Act (PRA) requires EP A to prepare an Information Collection Request (ICR),
which estimates the reporting and recordkeeping burden imposed by their regulations. Under the PRA,
"burden" means the total time, effort, or financial resources expended by persons to generate, maintain,
retain, or disclose or provide information to or for a Federal agency. This includes the time needed to
review instructions; develop, acquire, install, and utilize technology and systems for the purposes of
collecting, validating, and verifying information, processing and maintaining information, and disclosing
and providing information; adjust the existing ways to comply with any previously applicable
instructions and requirements; train personnel to be able to respond to a collection of information; search
data sources; complete and review the collection of information; and transmit or otherwise disclose the
information.
Section 403 contains no reporting or recordkeeping requirements, and thus no ICR is necessary for this
rule. However, an ICR was previously prepared and filed for the promulgation of regulations for TSCA
~402(a) and 404, and these burden estimates were based on estimates of the number of lead-based paint
identification and intervention activities anticipated. EPA re-examined the ~402(a) and 404 ICR and
determined that these estimates would not change due to the standards being proposed for ~403.
ES-1 0
~403 EA
-------
Executive Order 12898--Federal Actions to Address Environmental Justice
EP A investigated environmental justice related issues with regard to the potential impacts ofthis action
on the environmental and health conditions in low-income and minority communities. Non-white
households are more likely to live in housing with lead-based paint hazards, and their children are
expected to receive greater reductions in blood-lead levels if these hazards are mitigated. As a result,
non-white households are expected to bear more of the costs of complying with g403 but also receive
more of the benefits. Lower-and upper-income households face roughly the same cost of compliance and
are expected to receive the same blood-lead reductions. However, this means that lower-income
households will have to forego a larger share of their income to comply with g403.
Executive Order 13045--Protection of Children from Environmental Health Risk and
Safety Risks
The focus of the g403 regulation is on the protection of children's health. The household lead standards
were chosen based on an analysis of the health risks to children only. The benefits from g403 outlined in
Chapter 5 are a reflection of benefits to children under 6 years old only.
Of the estimated 173 million children born between 1997 and 2046, approximately 131 million children
will be born into housing built prior to 1979. It is estimated that g403 will result in reductions in
exposure to household lead in soil, dust, and paint for 43.8 million children born over that 50 year span.
This reduction in exposure, in turn, will result in reductions in the incidence of elevated blood-lead levels
and increases in average IQ. Exhibit ES-5 presents blood-lead and IQ statistics for both the baseline and
post-compliance scenarios.
Baseline
Post-9403
IEUBK
Post-9403
Empirical
Exhibit ES-5
Beneficial Health Impacts on Children Resulting from ~403
~ Number with ~ Number with ~ Number with ~
. . . .
: elevated i blood-lead: blood-lead:
Mean blood- ~ blood-lead ~ greater than ~ greater than ~ ~ Number
lead level i due to pica i 10 IJg/dl : 20 IJg/dl i Average i avoiding 10
(lJg/dl) ~ (millions) ~ (millions) ~ (millions) ~ 10 point gain ~ less than 70
I I I I
4.1~ 2.4~ 10.6~ 1.0~ NA~ NA
3.3 ~ 1.1 i 3.7 ~ 0.2 i 3.1 ~ 26,000
3.9 ~
1 . 1 ~
8.2 ~
0.7 ~
0.8 ~
9000
As shown, the health impacts of g403 are positive and substantial. The reduction in the number of
children suffering from elevated blood-lead levels due to pica (direct ingestion of paint chips) is on the
order of 1.3 million. The reduction in the number of children with elevated blood-lead levels (greater
than 10 /-lg/dl) from all sources is estimated at 2.4 to 6.9 million. The increase in average IQ depends
greatly on which benefits model is used but is between 0.8 and 3.1 points. Similarly the number of
children who will avoid an IQ less than 70 points is between 9,000 and 26,000 depending on the benefits
model employed.
~403 EA
ES-11
-------
ES-12
~403 EA
-------
1.
Introduction
1.1
Background
This report presents an economic analysis of proposed regulations setting standards for lead-based paint
hazards. The regulations are being proposed under authority of 9403 of the Toxics Substances Control
Act (TSCA). This section was established by the Residential Lead-Based Paint Hazard Reduction Act of
1992, also known as "Title X."
Use of lead-based paint in residences has been an important source of lead exposures historically.
Production and sale of lead-based paint for residential use was banned in the United States in 1978. This
ban, in combination with a phase-out of lead in gasoline, more stringent standards for lead in drinking
water, and reduced use of lead as solder in food cans, has dramatically reduced environmental lead levels.
Recent studies have suggested, however, that exposure to lead poses hazards at levels once thought to be
safe. In addition, past use of lead-based paint and other sources of lead have resulted in contamination
that continues to pose human health hazards. Many older residences (especially those built before 1978)
have lead-based paint that has chipped or peeled and become available for ingestion, especially by
children. In addition, residential dust and soil are contaminated with lead from past lead-based paint use
and other sources.
A variety of both voluntary and mandatory programs have been established at the federal, state and local
levels to encourage remediation of these residual hazards. The proposed rule considered here will
support implementation of these programs, either directly or indirectly, by establishing a definition of
lead paint-based hazards. By issuing these hazard definitions, EPA hopes to encourage improved
targeting of programs that address residential lead problems, thereby enhancing the cost-effectiveness of
lead-hazard programs as a whole.
1.2
Purpose 01 The Proposed Rule
TSCA 9403 requires that EPA promulgate regulations to "identify... lead-based paint hazards, lead-
contaminated dust and lead-contaminated soil" for purposes of other parts of Title X. This economic
analysis addresses the proposed 9403 standards identifying lead-based paint hazards, which include
residential hazards from deteriorated paint and contaminated dust and soill These standards apply
directly to "target housing" (most housing constructed prior to 1978). In addition, various parties may
apply the standards to newer residences as well.
Property-owners are not required to take action to address lead-based paint hazards, as defined by the
proposed rule. However, the proposed rule is expected to encourage such interventions by providing a
specific definition of hazard, along with guidance on actions that can be taken to address the hazards.
The rule should encourage more effective targeting of activities under various federal, state and local
programs, by communicating EP A's best judgement about conditions which require intervention.
The proposed rule also defines "levels of concern." Levels of concern are set at lower levels of
contamination than the hazard standards, and are intended for use in risk communication. They were
based on health risks only and are not part of this economic analysis.
~403 EA
1-1
-------
The lead-based paint hazard standards therefore define conditions of lead-based paint and levels of lead
in dust and soil at which EPA believes hazards should be addressed2. Paint hazards should be addressed
by repairing deteriorated paint, removing or enclosing the paint, or replacing the painted component.
Dust-lead hazards should be addressed by intensive cleaning. Soil-lead hazards should be eliminated
through soil removal or permanent cover of the soil. EP A is planning to develop a guidance document to
describe the recommended responses in more detail.
The proposed standards support implementation of key provisions of Title X, including eligibility criteria
for the Department of Housing and Urban Development's (HUD's) abatement grant program and
requirements that owners of HUD-associated housing and federal agencies evaluate and control lead-
based paint hazards in residential properties being sold. In addition, sellers and lessors of housing built
before 1978 are required to disclose known lead-based paint and lead-based paint hazards prior to sale or
rental, under Title X ~ 1018. Certified workers must be used for evaluation and cleanup where lead-based
paint hazards are present.
In addition, the proposed standards will have broader uses. The standards communicate the Agency's
best judgement about the identification of lead-based paint hazards to property owners, state and local
officials, tenants and other decision-makers. EPA expects that public and private institutions may
incorporate these standards into state and local laws, housing codes, and lending and insurance
underwriting standards.
1.3
Goals of The Economic Analysis
The purpose of this report is to analyze the choice of standards relating to the U.S. Environmental
Protection Agency's (EPA's) decisions under ~403 of the Toxic Substances Control Act. The report also
meets the requirements for economic analysis of Executive Order 12866 -- Regulatory Planning and
Review; the Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement Fairness Act
(SBRFA); Executive Order 12898 -- Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations; Executive Order 13045 -- Protection of Children from
Environmental Health Risks and Safety Risks; the Unfunded Mandates Act and Executive Order 12875
-- Enhancing the Intergovernmental Partnership; and the Paperwork Reduction Act (PRA).
The analysis compares alternative candidate standards in terms of their net benefits. Net benefits are
based on the benefits of risk reduction and the costs of hazard control activities needed to achieve the
reduction in risk. The analysis calculates net benefits for a wide range of alternative standards, including
the proposed ~403 hazard levels.
The analysis does not attempt to predict precisely how much remediation of residential lead-based paint
hazards will occur as a result of promulgating these standards. Rather, it is designed to provide
comparisons of different standards rather than absolute measures of costs and benefits for the different
standards.
The definition of lead-based paint itself has already been established by statute, and is not addressed by
this rule. This standard focuses on the conditions under which lead in deteriorated paint surfaces
constitutes a hazard.
1-2
~403 EA
-------
The lead hazard standards to be established under 9403 are intended to tell people when they should act,
i.e. when interventions should take place. In the economic literature this is referred to as a normative
analysis (as opposed to a positive analysis, which describes what will happen). As such, the analysis
identifies hazard levels at which the interventions maximize net benefits. More specifically, in defining
when intervention actions should occur, the relevant measure is maximizing net benefits for children (the
population that is both at greatest risk from exposure and the most in need of protection since children
are not in a position to protect themselves). Therefore, the birth of a child is used as the event that
triggers intervention activities in the analysis used to evaluate alternative standards.
The factors considered in estimating the costs and benefits include:
.
The adverse health affects from lead exposure -- what they are and how they differ with differing
lead-related characteristics of the housing unit, and across demographic, geographic and/or
socio-economic groups.
.
The costs to individuals and society of interventions which reduce lead exposure.
.
The effectiveness and duration of these interventions and the resulting benefits in terms of
reduced exposure and reduced adverse health effects.
.
The value of these benefits to individuals and society.
By comparing the total present value of net benefits (benefits minus costs) for each potential standard, it
is possible to identify the standard that yields largest net benefits and to compare standards with various
levels of costs or benefits in terms of their net benefits. For example, as standards become more
protective (i.e. more stringent), they also become more costly. By looking at net benefits, the analysis
shows how much society would give up to acquire more protection than that offered by the standard that
maximizes net benefits. Likewise, costs can be reduced by making the standards less stringent than the
one that maximizes net benefits. The analysis estimates what society gives up with this reduction in
costs. In focussing the estimations on interventions that take place at the time of the birth of a child, the
analysis identifies the standards that maximize the net benefits to the population of greatest concern --
children.
It is important to emphasize that both the costs and the benefits calculated in this report are subject to
substantial uncertainty. The net benefits measure, therefore, provides insight into the relationship
between costs and benefits for different hazard levels, but does not represent true net social benefits.
Important components of the benefits of reduced lead exposure could not be estimated. Adult health
effects were excluded, and only the most important of the benefits to children under the age of six were
included. In addition, there is uncertainty about the relationship between environmental exposures and
blood-lead levels. These uncertainties are highlighted by the fact that the two models provided by EPA
yield very different estimates of risk.
Finally, both costs and benefits reflect simplifying assumptions about how and when property-owners
will react to the hazard levels. Actual costs and benefits will vary, depending on how property-owners
actually react to the hazard levels. For instance, if fewer interventions occur or they occur later in time
than assumed in the analysis, both costs and benefits will be lower.
~403 EA
1-3
-------
While there is substantial uncertainty about the absolute estimates of aggregate costs and benefits, the
estimated relationship between the two for different standards is more reliable. Whenever costs are
incurred, benefits also result, although these benefits may occur at some time in the future.
1.4
Organization of this Report
Chapter 2 provides a historical overview of regulations that control lead use and risk. Programs that
address residential lead-based paint hazards are described, including both regulations and non-regulatory
initiatives. Many of these programs or regulations may be directly or indirectly affected by the proposed
lead hazard standards.
Chapter 3 describes the problems presented by lead-based paint hazards in residences, and discusses the
rationale for federal action to address these problems. This chapter also describes possible approaches
that might be used to address residual lead-based paint hazards, and provides an overview of the entire
economic analysis and its linkages with the risk assessment.
Chapter 4 describes in detail the methods used to calculate costs for different hazard standard levels. It
describes the development of unit costs and the methods used to aggregate costs across home types and
over time, and reports the predicted number of homes affected and number of interventions for alternative
standards.
Chapter 5 describes in detail the methods used to estimate benefits, including the calculation of unit
benefits and the aggregation of benefits across home types and over time. Finally, the chapter reports the
number of children affected and the predicted differences in blood-lead and IQ distributions for the
proposed standards.
Chapter 6 presents the results of the cost and benefit analyses. Aggregate costs and benefits are shown
for different standards for dust and then for soil, holding constant the standards for the other media. The
benefits estimated using the two blood-lead models are compared and reasons for their differences
discussed. Then net benefits are calculated for both blood-lead models, and these results are used to
identify the standards with maximum net benefits.
Chapter 7 discusses the uncertainties of the analysis and estimates the sensitivity of the results to specific
assumptions and input values. These include the discount rate, dollar values for lost IQ points,
assumptions about hazardous wastes costs, assumptions about the timing of interventions, the valuation
of fractional changes in IQ points, and other topics.
Finally, Chapter 8 presents findings relevant to specific rule-making requirements, including small
business impacts, unfunded mandates, paperwork burdens, environmental justice, and protection of
children.
1-4
~403 EA
-------
2. Regulatory Background
2.1
Lead as a Public Health Problem
Exposure to lead is one of the more serious public health problems currently facing the United States
(USDHHS, 1988). Lead's advantageous properties, including its malleability, resistance to corrosion,
good insulation, and low cost, have made lead attractive for many applications; lead has been used in
gasoline, ceramics, paint, and several other products. These uses have resulted in lead's release to and
distribution in all environmental media, which has complicated efforts aimed at reducing lead in the
environment (USDHHS, 1988). Much oflead in the environment is accessible to humans through a
variety of exposure pathways, and because it does not degrade, continued use of lead results in
accumulation in the environment. Human exposure to lead is of concern because lead interferes with the
normal functioning of cells, causing a range of toxic effects in the nervous, red blood cell, and kidney
systems (USDHHS, 1988). Fetuses and young children exposed to lead are especially at risk from
damages to the developing brain and nervous system (CDC, 1991a).
Knowledge of some of lead's negative health effects dates back hundreds of years. In the United States,
reproductive and developmental effects of lead were recognized in the 18th and 19th centuries in females
working in the lead industry and wives of lead workers. These women demonstrated health problems,
including sterility, spontaneous abortion, stillbirth, and premature delivery, and their offspring exhibited
high mortality, low birth weight, convulsions and other effects. The recognition of some of the health
effects of lead resulted in better industrial hygiene, which in turn reduced reproductive problems
(USDHHS,1988).
The prevalence of direct lead poisoning in children was first examined in Australia in the 1890s and
traced to exterior lead-based paint (USDHHS, 1988). In the US., physicians eventually defined lead
poisoning in children as a clinical entity in the early 20th century after a study reported that lead caused
acute encephalopathy in a number of children. In the 1930s and 1940s, epidemiologic data on childhood
lead poisoning began to expand, and the accrual of such data accelerated through the 1960s.
Rudimentary screening of children in the 1950s and 1960s clearly showed that they were being exposed
to excessive amounts of lead. Prevalence of lead poisoning was especially high among inner city youth.
Increased screening in the 1970s resulted in the recognition of lead poisoning as a widespread public
health problem (USDHHS, 1988).
Lead exposure's prominence as a public health concern is due to the magnitude of population blood lead
levels. The average blood lead levels in the US. population are estimated to have dropped in the last two
decades (USEPA 1989, 1991). However, the current geometric mean blood lead level in children is 3.1
~g/dL (USEP A, 1997), which is still about six times higher than the pre-industrial average of 0.5 ~g/dL
(USDHHS,1998).
The recognition of lead's adverse health effects has resulted in a lowering of the blood lead level that
triggers medical intervention. In 1970, the US. Public Health Service published guidelines that set the
level at 60 ~g/dl (CDC, 1991b). Shortly thereafter, the CDC set the guidelines at 40 ~g/dl, then revised
the recommendations to 20 ~g/dl, and finally set them at the current level of 10 to 14 ~g/dl in 1991.
Levels higher than this range should trigger various remedial actions; a child with a blood lead level
~403 EA
2-1
-------
between 15 to 1911g/dl should have nutritional and education interventions; a blood lead level greater
than 20 Ilg/dl should prompt medical evaluations and environmental investigations (CDC, 1991b).
The following two sections focus on existing regulations designed to decrease exposure to lead. Section
2.2 discusses regulations pertaining to lead products, releases of lead in the environment and the
workplace, and the concentration of lead in drinking water. Section 2.3 focuses solely on efforts at the
federal, state and local levels to decrease exposure to lead remaining in residential areas (including lead in
paint, dust, and soil). Section 2.4 describes a variety of non-regulatory initiatives that have been
undertaken by governmental agencies.
2.2
Regulation of Lead Products, Environmental and Workplace Releases of Lead, and Lead
in Drinking Water
The presence of lead in some consumer products has been prohibited or restricted by regulations.
Environmental releases of lead to air and water, and lead concentrations in waste, also have been
controlled. OSHA has set limits on allowable workplace concentrations of lead. In addition, regulatory
efforts have been made to remediate exposure to lead through drinking water systems.
2.2.1 Lead in Paint
In the 1950s, the paint industry voluntarily restricted sales of paint with lead content greater than one
percent (Mushak and Crocetti, 1990). Subsequently, the Lead-Based Paint Poisoning Prevention Act,
enacted in 1971, prohibited the use of paint with greater than one percent lead (by weight of nonvolatile
solids) in certain federally-owned or federally-assisted housing (HUD, 1990). As a result of 1976
amendments to this Act, lead paint was redefined as paint containing more than 0.06 percent lead by
weight (RUD, 1990). In 1978, the U.S. Consumer Product Safety Commission banned both the sale of
lead paint to consumers and the use of lead paint in residences or on other consumer-accessible surfaces
(16 CFR 1303).
2.2.2 Lead in Gasoline
The concentration of lead in leaded gasoline decreased significantly in the 1970s and the first half of the
1980s. In addition, the use ofleaded gasoline decreased significantly. The Clean Air Act of 1970 (CAA)
first indirectly controlled the use of lead in gasoline. Catalytic converters were necessary for automobiles
to achieve air standards, but lead rendered catalytic converters on automobiles inoperative. Therefore,
beginning in 1974, "unleaded" gasoline was introduced as a fuel for automobiles equipped with catalytic
converters. This "unleaded" gasoline contained no more than 1 0 percent of the lead concentration found
in leaded gasoline. The concentration of lead in leaded gasoline for use in vehicles and equipment
without catalytic converters went from unlimited, to 2 g/gal in 1978, to 1.1 g/gal in 1983, to 0.5 g/gal in
1985, to 0.1 g/gal in 1986. In 1986, the U.S. acted to phase out the use oflead in gasoline entirely (51
FR 24606).
2.2.3 Other Products Containing Lead
The U.S. canning industry began voluntarily phasing out the use of lead solder in food cans in the ] 970s
because alternative, affordable processes for sealing the seams of tin containers became available
(OECD, 1993). U.S. industry and the U.S. Food and Drug Administration have also undertaken efforts
to control lead exposure from ceramic ware (USDHHS, 199]), foil on wine bottles, and crystal ware
(USDHHS,1992).
2-2
~403 EA
-------
2.2.4 Environmental and Workplace Releases of Lead
Under authority of the Clean Air Act, EP A has established standards of perfonnance designed to lirllit
erllissions of air pollutants from lead smelting and processing facilities. In addition, lead emissions from
these and other industries are controlled via facility-specific permits written by states. These perrllits are
designed to reduce erllissions to the extent needed to meet EP A's national ambient air quality standard for
lead of 1.5 flglm3 (quarterly average), established in 1978 (43 FR 46246).
Under the Clean Water Act, federal effluent guidelines and pretreatment lirllits for lead-containing
effluents have been established for over 20 industries. These limits help achieve state-promulgated
surface water quality standards (which may be based on water quality criteria published by EPA). The
effluent lirllits are implemented by states through facility-specific permits and, depending on state water
quality standards, may be more stringent than federal effluent requirements.
Releases of lead as solid waste are regulated under the Resource Conservation and Recovery Act
(RCRA). A waste is defined as hazardous if, when tested, the leachate from the waste contains more than
5 ppm lead (40 CFR 261.24). In addition, certain lead-containing wastes are separately listed as
hazardous wastes. All of these wastes must be properly managed and disposed of (40 CFR 260-270).
EP A also initiated a voluntary program to reduce lead emissions, based on the Toxics Release Inventory
reporting. This project, called the "33/50 Program", encouraged industry to curtail emissions of 17
high-use toxic cherllicals, including lead, which are reportable to the Toxics Release Inventory. The
program sought comrllitments from companies to voluntarily reduce releases and transfers of the 17
pollutants and publicly recognized companies and facilities achieving their goals. The program's
reduction goals were broken into two phases -- 33 percent reduction by 1992 and 50 percent by 1995 --
using 1988 as the baseline year. A total of 1300 companies agreed to participate in this program, which
ended in rllid-September, 1996. Several other similar voluntary toxic emissions reduction programs have
been initiated by EPA (Environmental Science &Technology, 1996).
Efforts to reduce exposure to lead in the workplace have included setting perrllissible workplace air
concentrations of lead. The current Permissible Exposure Limit (PEL) is 50 flg/m 3 for most industries
except the construction industry (OECD, 1993). Under the Residential Lead-based Paint Hazard
Reduction Act, passed in October 1992, OSHA is required to issue interim regulations lowering the limit
for the construction industry (AECLP, 1993). On May 4, 1994, OSHA issued an Interim Final Rule,
Lead Exposures in Construction, setting the PEL to an 8-hour time weighted average of 50 flg/m3. The
interim rule also includes a list of three categories of tasks that are commonly known to produce
exposures above the PEL. Perfonnance of these tasks automatically triggers basic protective provisions
(USEP A, 1994).
2.2.5 Lead in Drinking Water
Exposure to lead in drinking water has continued because of past use of lead in plumbing. EP A has acted
to reduce these exposures through comprehensive measures (Mushak and Crocetti, 1990). In rules
promulgated in 1991, EPA outlined new treatment requirements for drinking water systems (56 FR
26460). The regulation requires tap water sampling from high risk homes (e.g., homes with lead service
lines or lead soldering installed since 1982). If at least 10 percent of home tap samples exceed 15 flg!l
(the "action level"), corrosion control treatment and public education is required. Replacement oflead
~403 EA
2-3
-------
service lines is required if corrosion control fails to bring water lead levels below the" action level." EP A
has also issued a maximum contaminant level goal of zero for lead in drinking water.
2.2.6 Resultant Reduction in Blood Lead Levels
As a result of these regulations, and other efforts, lead exposure and blood lead levels have decreased
significantly. The most recent national study (NHANES III Part 2) measured the mean of children's
blood lead levels to be 3.14 flg/dl. Thus, average blood lead concentrations have dropped over the last
two decades from about 15-20 flg/dl (USEPA, 1991) to approximately 3 flg/dl. In particular, reductions
of lead in gasoline have contributed dramatically to reductions in blood lead levels. Several studies have
specifically examined the relationship between blood lead and the lead content of gasoline, and have
found a strong positive correlation (Schwartz and Pitcher, 1989; Rabinowitz and Needleman, 1983).
Annest et al. (1983) noted a 37 percent drop in blood lead levels from 1976 to 1980 correlated with a
reduction in gasoline lead. Schwartz and Pitcher (1989) estimated that as much as 50 percent of blood
lead in the U.S. in the late 1970s may have been attributable to lead in gasoline.
Reductions in dietary lead have also contributed to declining exposures. Dietary lead intake for a two-
year-old child dropped from about 53 flg/day in 1978 to an estimated 13.1 flg/day in 1985; comparable
declines have been seen in adults (USEPA, 1989). These trends are attributable to the reduction in
gasoline lead emissions (and resulting reductions in deposition of lead from air to soil) and the voluntary
phaseout oflead-soldered cans by U.S. manufacturers since the 1970s. It can be calculated that these
changes in lead exposure from food have led to reductions of 1 to 2.5 flg/dl in average blood lead levels
(USEPA,1989).
2.3
Regulatory Efforts to Reduce Lead-Based Paint, Dust and Soil in Residential Areas
One of the largest remaining lead exposure sources for children is existing reservoirs of lead-based paint,
dust and soil present in many residential areas (USDHHS, 1988). In an effort to reduce exposure to
residential lead hazards, regulatory efforts to address these hazards have been increasing for several
years.
2.3.1 Current Estimates of Exposure
Although paint containing lead was banned for use in residences in 1978, exposure to existing lead-based
paint has continued due to prior use in residential and other buildings. In addition, leaded house paint can
contribute to lead in interior dust and soil. There also remains a significant soil burden of lead from
leaded gasoline and lead smelter emissions.
Several studies have demonstrated positive correlations between blood lead levels and lead in paint, soil
and dust (Charney et aI., 1980, Charney et al., 1983, and Bellinger et aI., 1986 as cited in HUD, 1990;
Clark et al., 1991). Blood lead levels are especially high in children. In a 1988 Report to Congress on
the extent of lead poisoning in children, USDHHS stated that the existing leaded paint in U.S. housing
and public buildings is "an untouched and enormously serious problem" (USDHHS, 1988). The Centers
for Disease Control conveys the seriousness of home lead exposure as a contributor to elevated childhood
blood lead by stating that lead poisoning exists in our society primarily because of exposure in the home
(CDC, 1991a). Infants and toddlers are especially susceptible to lead in the home because they may
ingest lead paint chips, dust and soil and because of the way they metabolize lead. Older children, up to
at least eight years old, are also at increased risk (USDHHS, 1988).
2-4
~403 EA
-------
Exposure continues mainly from paint in older homes, since houses built after 1978 are presumed to be
free of lead paint. Based on a 1987 HUD survey of 284 homes built before 1980, an estimated 64
million (83%) of all pre-1980 pri v ate homes have lead-based paint (defined as a measured lead
concentration on any painted surface of 1.0 mg/cm2 or greater) somewhere in the building (USEP A,
1995). Of these units, an estimated 12 million units have families with children younger than seven years
old. Seventeen percent of pre-1980 housing have dust lead levels in excess of federal guidelines (listed
below). Twenty-one percent of all pre-l 980 homes have excessive soil lead levels (USEPA, 1995).
Although a large majority of pre-1980 homes have lead-based paint, most have relatively small areas of
it. Older homes have the most lead-based paint. Pre-l 940 homes have about three times as much lead-
based paint as units built between 1960 and 1979.
A significant number of public housing units also contain lead paint. Eighty-six percent of all pre-1980
public housing family units have lead-based paint somewhere in the building (USEP A, 1995). Families
of any socioeconomic class may live in older housing, and thus be exposed to lead paint (USDHHS,
1988). However, families with the lowest incomes are disproportionately found in older housing
(USDHHS, 1988).
2.3.2 Federal Regulatory Activities to Decrease Exposure to Lead-Based Paint in Existing
Housing
Federal regulatory efforts and guidelines to limit exposure to lead-based paint in the existing housing
stock have evolved over the past twenty years. The following two sections chronicle these activities in
detail.
The Lead-Based Paint Poisoning Prevention Act and Amendments
The Lead-Based Paint Poisoning Prevention Act of 1971 (LBPPPA) and subsequent amendments (1973,
1976, 1987, and 1988) have resulted in a number offederal regulatory activities to reduce exposures and
risks from lead paint in housing. In addition to setting limits on the use of lead paint as described above,
the Act established grants for lead-poisoning screening and treatment, and required a report to Congress
on methods of abatement (HUD, 1990).
Abatement 01 Lead-Based Paint Hazards in Federally-Associated, Public and Indian Housing.
The 1973 amendments required HUD to eliminate, as much as was practical, hazards of lead-based paint
poisoning in pre-1950 housing covered by housing subsidies and applications for mortgage insurance and
in all pre-1950 federally owned housing prior to sale. HUD acted by issuing regulations to wam tenants
and purchasers of HUD-associated housing of the "immediate hazard" of lead-based paint (defined as
conditions associated with deteriorating lead paint surfaces). A 1983 court action resulted in broadening
the definition of "immediate hazard" to include intact paint; this definition was subsequently signed into
law in 1987,., In regulations issued by HUD in 1986 and 1987, the construction cutoff date was changed
from 1950 to 1973 in most cases. HUD again changed the cutoff date in response to 1987 amendments
to the LBPPP A; the new date became 1978 for all programs (HUD, 1990). The 1988 Amendments to
the LBPPPA specified the level which defines a lead paint surface as 0.5% by weight or 1.0 mg/cm2
(AECLP, 1993). HUD has also promulgated rules to eliminate lead paint hazards in public and Indian
housing (Mushak and Crocetti, 1990). Where children younger than seven years old are present in these
types of units, inspections for defective paint surfaces are required. If a child has an elevated blood lead
level, then the house must be inspected for chewable and defective surfaces, and abatement is required in
dwellings, common areas, or public child care facilities within the public housing.
~403 EA
2-5
-------
Grants. The LBPPPA authorized funding for States and cities to conduct extensive screening programs
to identify lead-poisoned children, refer them for medical treatment, investigate their houses for lead, and
require abatement (HUD, 1990).
Research and Reports to Congress. The 1971 LBPPPA required a report to Congress on the "nature
and extent of the problem of lead-based paint poisoning" and methods of removal. The 1987
amendments required an extensive research and demonstration project on lead-paint testing and
abatement technologies in HUD-owned housing, as well as additional reports to Congress (HUD, 1990).
In response to another mandate of the 1987 amendments, HUD conducted a survey of the distribution of
lead-based paint in the nation's housing stock and submitted a report on the results for privately-owned
housing to Congress in a comprehensive plan for abating paint in private housing. Additional
amendments in 1988 required a demonstration of abatement techniques in public housing as well as a
comprehensive plan to address abatement in public housing (HUD, 1990).
The Residential Lead-Based Paint Hazard Reduction Act
The most recent statutory activity related to the reduction of lead paint hazards is the enactment of the
Residential Lead-Based Paint Hazard Reduction Act in October of 1992. Also known as Title X of the
Housing and Community Development Act of 1992, this Act amends sections of the LBPPP A and adds a
new section (Title IV) to the Toxic Substances Control Act (TSCA), in addition to other important new
provisions. Described as "the most comprehensive and significant lead poisoning prevention legislation
in more than two decades" (AECLP, 1993), the Act aims to provide attainable goals for reducing lead
hazards in residential settings by targeting specific housing in the greatest need of abatement (AECLP,
1993).
Federally-owned and assisted housing. Title X allows for more targeted lead hazard evaluation and
reduction activities in federally-owned and assisted housing (AECLP, 1993). Whereas provisions under
the 1987 amendments to LBPPPA indicated that any and all houses built before 1978 that contain lead-
based paint constitute hazards that may be acted upon, Title X provides a more strategic approach to
reducing the hazards from lead-based paint. This approach involves requirements for risk assessments,
inspections, and interim controls for pre-1978 housing (targeted housing) and also requires deadlines for
action. Title X also extends federal lead-based paint requirements to all housing that receives more than
$5,000 in project-based assistance under any federal housing or community development program (in
addition to the federally assisted and insured houses covered under previous Acts) (Section 1012);
inclusion of these houses significantly expands the universe of federally-insured and assisted housing
subject to lead-based paint related requirements (AECLP, 1993).
Additional provisions apply to federally-owned housing being sold (AECLP, 1993). Properties built
prior to 1978 must be inspected and their condition disclosed to the prospective buyer. Units built before
1960 that have lead-based paint (defined as priority housing) must be abated (Section 1013).
Private housing. Private housing has received greater focus under Title X than under LBPPPA.
Although states, local governments or common law still determine whether landlords provide safe
housing, Title X includes several features to encourage evaluation and reduction of lead-based paint
hazards in private housing. One feature included formalizing into law a grant program run by the
Department of Housing and Urban Development for reducing lead-based paint hazards in low-income
privately owned housing. Under Section 1014, State and local governments are required to develop a
2-6
~403 EA
-------
Comprehensive Housing Affordability Strategy (CHAS) before receiving Federal housing or community
development grants. The CHAS must include an accurate estimate of the number of housing units that
contain lead based paint that are occupied by low-income families (HUD, 1997).
Other Title X provisions also affect targeted private housing (AECLP, 1993). These mandates include
integration of lead hazard evaluation and reduction into local housing programs, and certain disclosure
and warning requirements to be met at the time of sale or rental of any pre-1978 housing unit (Sections
1014 and 1018). Under Section 1018, a rule has been passed that requires purchasers or lessees to
receive an EPA pamphlet on hazards of lead in the home, and purchasers are allowed a 10 day period for
inspection for lead-based paint hazards. The Act also requires establishment of a national task force on
lead-based paint hazard reduction and financing; this group is to be made up of an array of groups
involved in housing, real estate, insurance, lending, abatement, and other groups (Section 1015). This
Task Force has recently published a report called Putting the Pieces Together: Controlling Lead
Hazards in the Nation's Housing, which recommends methods for controlling lead hazards in housing.
Safety of residents and workers. This law requires promulgation of a number of regulations
addressing the safety of workers undertaking interventions and the safety of families who live or will live
in treated housing. Section 1021 amends TSCA by adding a new Title IV, which primarily addresses
EPA requirements for contractor training and certification. The economic analysis presented in this
document supports the development of the regulation that responds to TSCA 9403 (in 91021); this
regulation requires EPA to define a "lead-based paint hazard" and dangerous levels oflead in dust and
soil.
EPA also must set standards of minimum performance for lead-based paint activities and ensure that
individuals engaged in activities are trained, that training programs are accredited, and that contractors
are certified (TSCA 9402). On August 29, 1996, EPA promulgated final regulations under 9402 of
TSCA to apply to target housing and child-occupied facilities. The Agency is also developing additional
9402 regulations, which will apply to training, certification and work practice standards for public
buildings constructed before 1978, commercial buildings, and steel structures such as bridges and water
tanks. These rules are under development and the promulgated regulations will deal with activities that
are intended to abate, delead or remove lead-based paint. In addition, the Agency is developing 9402
regulations to apply to renovation and remodeling projects. While renovation and remodeling projects
are not specifically intended to remove lead-based paint hazards, they may create a risk of exposure to
dangerous levels of lead.
HUD and EP A are to assist in funding state certification and training programs and issue standards for
a model state program (TSCA 9404). As of February 1997, $36.2 million dollars in grant money had
been awarded to 46 States, the District of Columbia, and 27 Native American tribes for the purpose of
certification and training programs (HUD, 1997). In addition, at the same time that rules were passed
under Section 402, the Agency published final rules under TSCA 9404 "that will allow States and Indian
Tribes to seek authorization to administer and enforce the regulations developed under section 402."
OSHA published interim final regulations requiring that contractors protect their workers from excessive
exposure to blood lead in May 1993. In addition, the regulations require employers to determine lead
exposure levels so that appropriate protective measures can be taken (HUD, 1997).
~403 EA
2-7
-------
EP A must assure that a program is in place to certify environmental sampling laboratories and must
provide for development of products and devices for testing and abatement (TSCA 9405).
Organizations that have authority to accredit laboratories as well as a list of accredited laboratories
have been established (NCSL, 1996b). On the product side, EPA and HUD evaluated 13 different
protocols for testing lead based encapsulants. This research led to the establishment by the American
Society for Testing Materials (ASTM) of two standard specifications for testing lead encapsulant
products. (HUD, 1997).
Education regarding lead paint hazards. Title X also mandates a variety of public educational
efforts. A hotline designed to inform the public about lead hazards was set up soon after passage of
the 1992 Act. The National Clearinghouse on Childhood Lead Poisoning was then established in
April, 1993. The Consumer Product Safety Commission, in coordination with EPA, published
'Reducing Lead Hazards When Remodeling Your Home" which has been distributed widely including
placement in hardware stores such as Home Depot that sell paint removal products (HUD, 1997).
Under TSCA Section 406, regulations are currently being developed which require renovators and
remodelers to inform residents of the hazards of renovation (NCSL, 1996b). HUD published the
"Guidelines for the Evaluation and Control of Lead-Based Paint Hazards in Houisng" in August 1995
in fullfillment of Section 1017 of Title X. This document is a comphrehensive guide on how to
identify and reduce lead-based paint hazards (HUD, 1997).
Research and development. A variety of research is also required under Title X. EPA is required
to conduct a study, currently in development, on the hazard potential of renovation and remodeling
(Section 1021: TSCA 402).
Section 405 (a) mandates EP A and other appropriate agencies to develop a program to promote safe,
effective and affordable monitoring and abatement measures. A wide variety of analyses and field
studies have been done to support Section 405 (a) (HUD, 1997). Under Section 405 (b) EP A has also
developed studies to establish minimum performance standards for laboratory analysis of lead in paint
films, soil and dust.
Section 405(c) requires the Centers For Disease Control and Prevention (CDC) to identify sources of
lead exposure for children. The CDC has numerous ongoing studies to evaluate the sources of lead
exposure. One such study is the NHANES III project which collected data regarding health indicators
and environmental exposure on 30,000 persons over a six year period (HUD, 1997). Section 405(c)(2)
mandates the National Institute for Occupation Health and Safety (NIOSH) develop studies on methods
of reducing occupation lead exposure.
A comprehensiv,e listing of all EPA, CDC, and NIOSH studies completed as of February 1997 can be
found in "Moving Toward a Lead-Safe America" (HUD, 1997).
2.3.3 Federal Guidelines and Other Activities Related to Lead in Soil and Dust
Guidelines for Levels of Lead in Soil and Dust
As mentioned above, under 9403 EP A is required to determine dangerous levels of lead in dust and
soil under Title X. While developing these definitions, interim guidelines specifying levels down to
which lead-based paint should be abated have been recommended by EPA. These levels are 100 j.lg/ftZ
for uncarpeted floors, 500 I1g/ftZ for window sills and 800 j.lg/ft2 for window wells (Goldman, 1994).
2-8
~403 EA
-------
A level of 400 ppm has been set as the hazard level for soils; lead abated to this level is expected to
result in no more than a 5% probability of a child having a blood lead level exceeding 10 )lg/dL
(Goldman, 1994).
Other Activities
Under authority of Title III of the 1986 Superfund amendments, EP A has funded projects in Boston,
Baltimore, and Cincinnati to test the health effects of abating soil with high lead content in residential
areas. This research has been considered in developing the final guidance on soil clearance levels.
2.3.4 State and Local Programs to Reduce Exposure to Lead-Based Paint, Dust and Soil
Activity to address lead-based paint hazards has recently increased at the state and local levels,
although certain areas (e.g., Baltimore) have had programs in place for many years. The following
sections discuss several state and local systems.
State Activities
Thirty-eight states and the District of Columbia have laws pertaining to lead poisoning prevention
activities (NCSL, 1997). Some notable examples of state programs are described below.
In 1971, Massachusetts passed a comprehensive Lead Poisoning Prevention Act (NCSL, 1996a). The
law established a statewide program for the prevention, screening, diagnosis and treatment of lead
poisoning, which included elimination of sources of poisoning and provided for research, educational,
epidemiologic and clinical activities. The law requires universal screening of children less than 6
years old. Because of passage of this law, a higher percentage of children were screened in
Massachusetts (in 1986) than in any other state (Prenney, 1987). Beginning in July 1988, regulations
were established to govern the training and licensing of deleaders and lead paint inspectors (NCSL,
1996a) .
Maryland's law (1996) prohibits the use of lead-based paint on any interior surface, any exterior
surface commonly accessible to children, or anything intended for household use. The law also
requires physicians to report people with elevated blood levels, and creates an advisory council to
explore the lead poisoning problem. The law also establishes a lead poisoning prevention commission,
a lead poisoning prevention fund, and risk reduction standards for affected properties. In addition, the
law established a comprehensive program to address lead removal, and also established maintenance
standards, registration of affected properties, and liability provisions (NCSL, 1997).
Rhode Island's Lead Poisoning Prevention Act was passed in 1991. Under the law, the Department of
Health is authorized to expand blood lead screening to all children under six at designated intervals. In
addition, child care providers must provide evidence of blood-lead screening for each child enrolled in
the facility (NCSL, 1997). Houses where children with blood lead levels of 25 )lg/dl or greater have
been identified must be inspected, and nonintact lead hazards in these homes must be abated (RIDH,
1993) .
California initiated its activities to reduce lead-based paint hazards in 1986 (FIorini et al., 1990). The
California law, updated in 1993, established the Childhood Lead Poisoning Prevention Program which
requires identification of target areas and analysis of information to design and implement a program
of medical follow-up and environmental abatement to reduce exposure. Specifically, it requires blood
~403 EA
2-9
-------
lead screening for children who are covered by health insurance. The statute also implements a lead
poisoning prevention and protection program in schools.
Delaware's Lead Poisoning Prevention Act requires that all child care facilities, public and private
nursery schools, preschools and kindergartens screen every child born after March 1, 1995 for blood-
lead levels unless the parents object. In addition, all health insurance agencies must provide coverage
for blood-lead screening at one year of age and all primary care physicians must provide this screening
at that time. Laboratories that conduct the tests are required to participate in a universal reporting
system.
Local Programs
Similar to state programs, local lead-poisoning programs have had limited resources with which to
carry out their programs (HUD, 1990). Differences between typical state schemes and selected city
programs lie more in the extent than in the substance of the activities (HUD, 1990). In general, city
programs are more focused and seem to receive higher priority, which may be due to the urgency of
the lead-paint problem in larger cities.
In the Comprehensive and Workable Plan for the Abatement of Lead-Based Paint in Privately Owned
Housing (HUD, 1990), the Department of Housing and Urban Development outlined several
distinguishing features of local programs as determined by investigation of ten selected cities:
A city that is governed both by local ordinances and state regulations for lead-poisoning
prevention and detection activities usually has local laws that are more stringent than state
laws and may supersede the state requirements.
.
In addition to providing intervention after cases of lead poisoning have been detected,
local programs may require intervention as a result of targeted inspection or tenant
complaints. Several cities, including Baltimore, Chicago, Louisville, New York, and
Philadelphia, are authorized to take such preventive measures.
.
In general, the city programs show more cooperation and coordination between agencies.
.
City programs usually screen for high blood lead levels more systematically and target
high-risk areas for screening.
Of the local programs for childhood poisoning prevention, Baltimore has one of the most extensive
schemes. In addition, as early as 1951, the city banned the use of lead paint in the interior of
residences (Mushak and Crocetti, 1990).
2.4 Non-Regulatory Initiatives to Reduce Lead-Based Paint, Dust and Soil Exposure
In addition to programs, policies and rules developed in response to Congressional mandates, several
governmental agencies have undertaken programs to reduce lead exposure that go beyond the Title X and
TSCA requirements.
2-10
~403 EA
-------
2.4.1. Joint Initiatives
The EP A has developed two initiatives in conjuction with several other federal, state and local agencies
to reduce lead exposure in high-risk, low-income communities. The Environmental Justice Initiative
was constructed to: (a) demonstrate that an effectively designed program can reduce poor children's
blood lead levels, (b) demonstrate the benefits of public, private and community organization
cooperation in the fight against lead poisoning, (c) conduct lead screenings, hazard reduction and
education efforts, (d) document the projects' successes and shortcomings and (e) foster self-sufficiency
through job creation and empowerment. To these ends the EPA formed an inter-agency work group
which has committed $3.7 million for pilot programs to test community-based approaches to lead
poisoning reduction (HUD, 1997).
The EP A has also designed the Whole House Initiative to establish and evaluate programs which
reduce multimedia exposure to lead hazard in the home. Again the EP A helped establish an inter-
agency task force whose primary activities were developing a multimedia Geographic Information
System (GIS) database and creating a multimedia environmental training proceedures (HUD, 1997).
2.4.2 Blood Screening Programs
The CDC State and Community-Based Childhood Lead Poisoning Prevention Program provides grant
monies to state and local agencies for the formulation and execution of blood lead testing programs.
These programs not only screen young children for lead poisoning but also identify potential sources of
lead, monitor the medical and environmental management of children who have been diagnosed with
elevated blood lead, and provide information to the public on methods to reduce lead hazards (HUD,
1997) .
In addition, the CDC is working with the State of Massachussetts to develop a model blood-lead
database. The database provides information on the health, housing, medical and environmental
management of children diagnosed as having elevated blood lead levels (HUD, 1997).
The Health Care Finance Administration of the U.S. Health and Human Services Department also
supports the Early and Periodic Screening, Diagnostic and Treatment (EPSDT) Program. The
EPSDT provides comprehensive and preventative health care benefits to lower income children
through the State Medicare programs. All Medicaid-eligible children between the ages of six months
and 72 months are required to have their blood lead tested under this program.
2.5
Benefits of Increasing Lead Awareness and Lead Poisoning Prevention Programs
Although several states and localities have taken action on lead-based paint, many have no standards
for paint, dust, or soil abatement. In addition, among the states and local areas that do have standards,
the levels of paint, dust, and soil considered unacceptable differ. By providing definitions at the
federal level for lead paint hazards and dangerous levels of lead in dust and soil, those states that do
not have standards may be prompted to adopt standards more quickly. In addition, the federal
guidelines will provide consistency between the states.
In addition to federal regulation on lead hazards, there is clearly room for non-regulatory initiatives to
reduce exposure to lead in paint, soil and dust. The early evidence from non-regulatory measures to
~403 EA
2-11
-------
reduce lead hazards indicates that community-based, joint government-private programs can help fill
the gaps left by statutory regulation alone.
2.12
~403 EA
-------
References
Alliance to End Childhood Lead Poisoning (AECLP). 1993. Understanding Title X: A Practical
Guide to the Residential Lead-based Paint Hazard Reduction Act of 1992. AECLP,
Washington, DC. January.
Annest, ]., J. Pirkel, D. Makuc, J. Neese, D. Bayes, and M. Kovar. 1983. Chronological Trend in
Blood Lead Levels Between 1976 and 1980. New England Journal of Medicine, 308: 1373-
1377.
Centers for Disease Control (CDC). 1991a. U.S. Department of Health and Human Services, Public
Health Service. Strategic Plan for the Elimination of Childhood Lead Poisoning. February.
Centers for Disease Control (CDC). 1991b. Preventing Lead Poisoning in Young Children. U.S.
Department of Health and Human Services, Public Health Service. Atlanta, GA. October.
Clark, S., R. Bornschein, P Succop, S. Roda, and B. Peace. 1991. Urban Lead Exposures of
Children in Cincinnati, Ohio. Chemical Speciation and Bioavailability, 3(3):163-171.
Environmental Science and Technology. 1996. 33/50 Program Ends, EPA Declares Success.
30 (11) : 482A -483A.
FIorini, K.L., G.D. Krumbhaar, and E.K. Silbergeld. 1990. Legacy of Lead: America's Continuing
Epidemic of Childhood Lead Poisoning. Environmental Defense Fund, Washington, DC.
March.
Goldman, Lynn. 1994. USEPA, Office of Prevention, Pesticides and Toxic Substances.
Memorandum on Agency Guidance on Residential Lead-Based Paint, Lead-Contaminated
Dust, and Lead-Contaminated Soil. July 14.
Mushak, P., and A. Crocetti. 1990. Methods for reducing lead exposure in young children and other
risk groups: an integrated summary of a report to the U.S. Congress on childhood lead
poisoning. Environmental Health Perspectives, 89: 125-135.
National Conference of State Legislatures (NCSL). 1997. State Lead Poisoning Prevention Statutes.
July 23. Compiled by Doug Farquhar.
National Conference of State Legislatures (NCSL). 1996a. State Training/Certification/ Accreditation
Programs for Lead. November 8. Compiled by Doug Farquhar.
National Conference of State Legislatures (NCSL). 1996b. Update on EPA Lead-Based Paint
Activities. Compiled by Doug Farquhar.
Organization for Economic Cooperation and Development (OECD). 1993. Risk Reduction
Monograph No. 1: Lead. Background and National Experience with Reducing Risk. Paris:
Environment Directorate, OECD. Document No. OCDE/GD(93)67.
~403 EA
2-13
-------
Prenney, B. 1987. The Massachusetts Lead Program: Moving Toward Phase 2. Pr~vention ~p.date.
Developed by the Maternal and Child Health Consortium Project and the National CoalItIOn
on Prevention of Mental Retardation. April.
Rabinowitz, M. and H. Needleman. 1983. Petrol Lead Sales and Umbilical Cord Blood Lead Levels
in Boston, MA. Lancet, 8314/5 (1):63.
Rhode Island Department of Health (RID H) . 1993. R 23-24.6-PB: Rules and Regulations for Lead
Poisoning Prevention.
Schwartz, J. and H. Pitcher. 1989. The Relationship Between Gasoline Lead and Blood Lead in the
United States. Journal of Official Statistics, 5:421-431.
U.S. Department of Health and Human Services, Public Health Service, Agency for Toxic Substances
and Disease Registry (USDHHS). 1988. The Nature and Extent of Lead Poisoning in the
United States: A Report to Congress. July.
U.S. Department of Health and Human Services, Public Health Service, Food and Drug
Administration (FDA). 1991. FDA Talk Paper: FDA Issues New Guidance on Lead in
Ceramic Ware, November 11.
U.S. Department of Health and Human Services, Public Health Service, Food and Drug
Administration (FDA). 1992. Statement by Michael R. Taylor, Deputy Commissioner for
Policy, Food and Drug Administration, Public Health Service, Department of Health and
Human Services Before the Ad Hoc Subcommittee on Consumer and Environmental Affairs,
Committee on Government Affairs, U.S. Senate. March 27.
U.S. Department of Housing and Urban Development (HUD). 1990. Comprehensive and Workable
Plan for the Abatement of Lead-Based Paint in Privately Owned Housing: Report to
Congress. Washington, DC. December.
U.S. Department of Housing and Urban Development (HUD). 1997. Moving Toward a Lead-Safe
America: A Report to the Congress of the United States; Office of Lead Hazard Control.
Washington, DC. February.
U.S. Environmental Protection Agency (USEPA). 1989. Review of the National Ambient Air
Quality Standards for lead: Exposure Analysis Methodology and Validation. Office of Air
Quality Planning and Standards. Research Triangle Park, N.C. June.
U. S. Environmental Protection Agency (USEP A). 1991. U. S. Environmental Protection Agency
Strategy for Reducing Lead Exposures. February 21.
U.S. Environmental Protection Agency (USEPA). 1994. Lead; Requirements for Lead-based Paint
Activities, Proposed Rule, Federal Register, Friday, September 2.
2-14
~403 EA
-------
U.S. Environmental Protection Agency (USEPA). 1995. Report on the National Survey of Lead-
Based Paint in Housing. Base Report. USEPA, Offiee of Pollution Prevention and Toxies.
EP A 747- R95-003. April.
U. S. Environmental Protection Agency (USEP A). 1997 Risk Assessment for the Section 403
Rulemaking; USEP A, Chemical Management Division, Office of Pollution Prevention and
Toxics.
~403 EA
2-15
-------
2-16
~403 EA
-------
3. Problem Definition, Regulatory Options, and Overview of Analytic
Approach
This chapter begins by characterizing the lead contamination problem to be addressed under Section 403.
The various sources of exposure, along with related blood lead levels and health effects, are presented.
Section 3.2 discusses sources of market failure, both on a theoretical basis as well as specifically how
incomplete information has resulted in too few lead interventions taking place. This section also
introduces regulation as a reasonable solution for such a market failure, and discusses why regulation
should take place at the federal level. Alternative regulatory options are presented in Section 3.4.
Finally, Section 3.5 serves as a summary of the risk assessment and its linkages to the economic analysis.
This summary serves as an introduction to the detailed presentation of cost estimates (Chapter 4),
benefits estimates (Chapter 5), and net benefits (Chapter 6) associated with the various lead hazard
standards.
3.1
Lead Contamination Problem
Despite recent reductions in air, water, and food contamination, important sources of lead exposure for
children remain, due largely to the widespread presence of lead-based paint in home environments.
Exposure to lead results in elevated blood lead levels associated with a suite of health effects, including
notably loss of IQ and other cognitive effects. Recent data suggest that the blood lead level of more than
one in 200 children exceeds the threshold for lead poisoning as defmed by the CDC, 20 /lg/dL.
3.1.1 Exposure Sources
Although lead may cause adverse health effects in any individual, exposed at any stage of life (in utero
through adulthood), the focus of Section 403 and this analysis is on children exposed from birth through
the sixth birthday. Young children are particularly susceptible to lead hazards because they are at a stage
of rapid development of the central nervous system, and because their normal behavior is likely to result
in greater exposure than older groups experience.
Currently the most significant high-dose source of lead exposure in children under school age is lead-
based paint. Through the 1940's, paint manufacturers used lead as a primary ingredient in many oil-
based interior and exterior house paints. During the 1950's and 1960's, the usage gradually decreased as
new paints were developed, and in 1978 the Consumer Product Safety commission (CPSC) ruled that
paint used for residence, toys, furniture, and public areas must not contain more than 0.06% lead by
weight. Nevertheless, an estimated 64 million (or 83%) of privately-owned, occupied housing units and
86% of public housing units built prior to 1980 contain some components covered with lead-based paint
(USEP A 1995). Children's exposure to lead from lead-based paint is likely to be high when the paint is
in a deteriorated state or is found on accessible, chewable, impact, or friction surfaces, making the lead
paint available to children who ingest paint chips (USEP A 1986; CDC 1991). This "pica" behavior
appears to be rare, but likely causes most of the highest blood lead levels observed in children.
In addition to being a source of direct exposure, deteriorated lead-based paint or the improper removal of
lead-based paint from a housing unit may contaminate soil and dust. Children are exposed to lead from
soil or dust in their homes as a result of typical hand-to-mouth activities. Lead-contaminated dust and
i403 EA
3-1
-------
soil are thought to be the major pathway through which most young children are exposed to lead from
lead-based paint hazards (USEP A 1986).
Young children are also exposed to a variety of other lead sources, which appear less important on a
national scale. Airborne lead is present in emissions from lead smelters, battery manufacturing plants,
and solid waste incinerators. The phase-out of leaded gasoline has contributed to the reduction of
airborne lead (CDC 1991). Drinking water may become contaminated with lead after it leaves the
treatment plant (CDC 1991). Although lead levels in drinking water generally do not have a statistically
significant effect on blood-lead concentrations as a result of the 1986 Safe Drinking Water Act, water is
still considered an important localized exposure source where lead solder and/or brass plumbing fixtures
are present because of the high absorption rate of lead in water. Lead exposure through food ingestion
has declined greatly in importance due to the phase-out of lead-soldered food cans and public education
(CDC 1991). Despite these improvements in exposure from air, water, and food, however, many children
still experience blood lead levels exceeding CDC health guidelines.
3.1.2 National Blood Lead Levels and Health Effects
Most studies of the health effects of lead use body-lead burden as biomarkers of lead exposure.
Measures of body-lead burden include lead in bones, teeth, and hair. Each of these options, however, has
a variety of disadvantages, including poor sensitivity and external surface contamination problems. The
most common measure used is blood-lead concentration. Although blood lead level reflects a mixture of
recent and past exposure, it has the advantage of being easily and inexpensively measured.
The widest recent survey of children's blood lead levels is the Third National Health and Nutrition
Examination Survey, Phase 2 (NHANES III). It was conducted from 1991 to 1994 and included
information from 987 children aged one to two years, the most appropriate age group for estimating
health effects. The national geometric mean and standard deviation blood lead levels for this group can
be calculated as 3.14 j..lg/dL and 2.09 j..lg/dL, respectively. (Battelle 1997)
Elevated blood lead levels are associated with an assortment of deleterious health effects, including IQ
point loss, other cognitive effects, neurological disorders, anemia and impaired heme synthesis, and
impaired hearing.
IQ point loss can be used to measure neurological loss that a child experiences due to any level of lead
exposure. Children's average IQ loss from lead exposure can be calculated from NHANES III data as
described in the Risk Assessment. Based on the one to two year olds surveyed, the national average
decrement is 1.06 points. (Battelle 1997)
The number of children with IQ less than 70 and the number of children with blood lead levels above 20
j..lg/dL can be used to measure cognitive effects seen mostly in children with high levels of lead exposure.
An IQ score of 70 is two standard deviations below the population mean; it is used as an indicator of
mental retardation and as the cut-off for special education requirements. Blood lead levels above 20
j..lg/dL are the level at which CDC recommends a complete medical evaluation, an environmental
assessment, and necessary environmental remediation for the child and his/her environment. Battelle
(1997) has calculated the fraction of one to two-year olds falling within these categories, based on
NHANES III data and certain assumptions. The results are that 0.115 percent of children have IQ's
under 70, and 0.588 percent have blood lead levels above 20 j..lg/dL.
3-2
~403 EA
-------
3.2
Market Failure
From an economic perspective, one necessary condition for regulations is the existence of an inefficiency
in the allocation of resources. This inefficiency is commonly labeled a market failure since the market is
the mechanism assumed to make efficient resource allocations possible. A market failure can come from
one or more of several sources. These include poorly defined property rights (such as negative
externalities, common property resources, and public goods); imperfect markets for trading property
rights (because of a lack of perfect information or of contingent markets; monopoly power; distortionary
taxes and subsidies and other inappropriate government regulations); and the divergence of private and
social discount rates. I
The occurrence of any of these conditions justifies further inquiry into the need for government regulation
to reduce inefficiencies in the allocation of society's resources. This section considers whether any of
these conditions are linked to excess exposures from lead contamination in residential soil, dust, and
paint. If so, a better understanding of the nature of the inefficiencies involved may facilitate the design of
effective regulations. The specific regulation considered here is the promulgation of hazard standards as
mandated by 9403.
The strongest case for the existence of a market failure can be built on the apparent lack of perfect
information. Correct information is an important prerequisite to the demand for intervention and other
forms of lead-based paint hazard controls.2 The property owner making the intervention decision has to
know the levels of lead in soil, dust, or paint; know what risks are implied by these levels; know the
significances of these risks; and know what can be accomplished through various forms of intervention.
Clearly, without knowing there is a lead problem, the owner will have too Iowa demand for intervention.
Misinformation on the other attributes of the intervention decision can also distort the demand for
intervention. Research into public views of risk indicate that misperceptions about latent risks, like those
associated with lead contamination, are common. These misperceptions can be biased upward or
downward, resulting respectively in excess and insufficient demand for intervention. Finally, reliable
information on the relative and absolute effectiveness of different intervention alternatives could be a
significant obstacle.
These relationships are illustrated in Exhibit 3.1, where the line labeled D represents the demand for
intervention under the condition of complete and perfect information. With this information, consumers
are able to accurately compare the value of intervention activities with their costs. In this case, the
number of interventions performed would be Q. There are several circumstances, however, that would
increase the demand for interventions above this optimal amount. If consumers overestimate the amount
of risk they and their families are currently subject to due to their housing conditions, and/or overestimate
the effectiveness of interventions, and/or overestimate the costs of alternative solutions (such as moving),
This taxonomy was developed from (Axelrad, 1993), and (Boadway, 1979).
Tbroughout this section, the term intervention is used to refer to the entire range of lead-based paint hazard
control activities.
j403 EA
3-3
-------
Exhibit 3.1. The Demand for Abatements
under Alternative Inform ation Scenarios
Price
PI
Supply of
Abatements
DI
P
Pz
Dz
D
Q2
Q
QI
Number of
Abatements
Where:
D=
Demand for Interventions where consumers have complete information.
D]=
Demand for Interventions where risks from lead are overestimated and/or effectiveness of
interventions are overestimated and/or cost of substitutes is overestimated.
Dz=
Demand for Interventions where risks from lead are underestimated and/or effectiveness of
interventions are underestimated and/or cost of substitutes is underestimated.
then the demand for interventions would exceed that under complete/perfect information, as represented
by line Dj. This situation would result in too many interventions occurring, at too high a price. Likewise,
if consumers underestimate the amount of risk they and their families are currently subject to due to their
housing conditions, and/or underestimate the effectiveness of interventions, and/or underestimate the
costs of alternative solutions (such as moving), then the demand for interventions would be less than that
under complete/perfect information, as represented by line Dz. This would result in too few interventions
occumng.
The market itself has not provided a means for correcting this situation. Although businesses that offer
testing or intervention services should find it in their vested interest to provide the kinds of information
cited above, thi& possibility has not closed the information gaps for the public. One impediment may be
public uncertainty about the reliability of the information that such businesses would provide. Their
information may be unreliable because they are not fully competent to assess the lead contamination and
what needs to be done, because the businesses are subject to moral hazard (which occurs, for example,
when a firm tells a homeowner that there is a lead problem that warrants a certain intervention it can
perform when the intervention is not necessary or suitable), or both. Since many property owners may
lack easy access to independent sources of information to motivate their intervention decisions, doing
nothing may be the likely response.
3-4
~403 EA
-------
While lamentable, this lack of action is understandable given limits on the time and money that an owner
can actually spend on obtaining information needed for many different decisions. For example, even
though homeowners, as parents, may be deeply concerned about the welfare of their children - a key
target of exposure from lead contamination - there are a host of other issues besides lead that affect
their children's welfare and for which parents need information to make important decisions. These other
information needs compete with the information needs of the lead intervention decision for scarce
household resources. Given how little intervention has been initiated by homeowners relative to the
prevalence of the problem, the likelihood that there is insufficient demand for intervention and that
information gaps contribute to this circumstance appears to be high. In conclusion, it appears that at
least one condition associated with market failures holds and, consequently, that inefficiencies may
characterize the market for lead testing and intervention.
Before a final determination can be made about the inefficiencies associated with the lack of information,
the costs of spanning the information gaps must be considered. One of the more important unknown
variables in setting hazard standards under ~403 is what constitutes an effective means of disseminating
this information; what approaches to making information available will actually get owners to test, to
consider the intervention alternatives, and to undertake intervention where appropriate. Simply setting
standards, without effective dissemination of the information, is likely to have little effect.
In attempting to answer the question of whether government regulation will make the market for
reductions in lead-based paint hazards more efficient, it is helpful to consider the public good aspect of
promulgating hazard standards. To the extent that the public finds these hazard standards credible and
takes steps to measure and reduce lead contamination, the standards are an independent benchmark for
action, providing at least part of the information needed to make an appropriate intervention decision. As
such, hazard standards can qualify as a public intermediate good since they can be used simultaneously
by many households in making their intervention decisions.3 Whether the hazard standards are a public
good or not depends ultimately on whether the interventions induced by the standards result in benefits
exceeding the costs. If so, the hazard standards are a public good. If not, they are a public bad. The
analysis in this report attempts to address this issue by discriminating among various forms and
magnitudes of hazard standards based on the magnitudes of their net benefits.
Other potential causes of market failure are the result of the persistence of lead intervention in residences.
By undertaking intervention, the owner creates positive externalities for any occupants outside of his or .
her immediate family (such as renters if the owner is a landlord) and subsequent owners who are
occupants of the home. If these renters and subsequent owners are fully informed about the implications
of lead contamination, the market may adequately compensate the original owner for undertaking lead
intervention and no externalities impede the intervention decision. If they are not fully informed, then the
original owner will not be sufficiently compensated for services provided to the renters and subsequent
owners. Under these circumstances, too few interventions will be undertaken. It is difficult to measure
how large this problem is since it requires information on the stock of knowledge about lead problems
held by tenants and purchasers today and in the future.4 Nevertheless, the development and wide
The term "public intermediate good" and its definition are adapted from Boadway, 1979.
Gathering such information would require a survey of homeowners and landlords as to their knowledge
about lead contamination and its adverse health effects.
~403 EA
3-5
-------
dissemination of credible lead hazards will increase the likelihood that renters and subsequent owners
will compensate the owners who undertook the interventions.
Compounding the problem of under compensation is a divergence between social and private discount
rates. Discount rates are particularly important, since this analysis anticipates that occupants as much as
fifty years in the future could potentially benefit from the intervention of a given housing unit. Even if
each renter or subsequent owner is willing to pay the full market value of the externality provided by the
original owner's lead intervention, it is likely that the private estimate of the present value of these future
payments to the original owner will be smaller than the present value based on the social rate of discount.
Consequently, by relying on private decisions, fewer interventions may be undertaken. If credible lead
hazard standards are widely disseminated, then the owner is likely to adopt these as his/her decision
criteria, as opposed to spending the time and money to develop personal definitions of lead hazard
standards. By basing the ~403 lead-hazard standards on the social discount rate, EPA will encourage a
level of intervention consistent with maximizing net social benefits.
This review suggests that there is one or more market failure affecting decisions regarding the
intervention of residential lead contamination. The lack of perfect information is a primary culprit. The
ultimate determination of a market failure, however, depends on whether gains in efficiency can be
accomplished by some form of regulation. An allocation of resources is deemed inefficient if someone
can be made better off without making someone else worse off. That is a core question of this analysis.
The examination of risks from lead contamination does indicate a substantial potential for making
individuals better off by reducing residential exposures from soil, dust, and paint. The benefit-cost
analysis presented in this report examines the questions of whether these actions would result in an
increase in net benefits. If the benefits of reducing lead exposures exceed costs, it is theoretically
possible to increase efficiency without making anyone worse off. The costs that have to be considered
include the costs of getting the right people to decide to act, to choose the right intervention, and to
perform and maintain the intervention in the specified manner as well as the direct costs of testing and
intervention.
3.3
Need for Federal Regulation
In the Residential Lead-Based Hazard Reduction Act of 1992 ("the Act"), the United States Congress
stated that the elimination of lead-based paint hazards was a national goal. Congress found that the
Federal Government must take a leadership role in building the required infrastructure, including an
informed public, State and local delivery systems, certified inspectors, contractors, and laboratories, and
trained workers (~1OO2(8)). By identifying what constitutes a lead-based paint hazard (defined as paint,
dust or soil conditions that would result in adverse human health effects), 9403 creates a crucial link in
the integrated federal regulatory approach necessary to adequately inform the public of the dangers of
lead-based paint, and to implement other portions of the Act that require either mandatory action, or in
some cases recommend voluntary action, if a lead-based paint hazard exists.
Justifying the need for a federal regulation requires two findings. First, there must be a market failure
that can be corrected through regulation. Second, it must be shown that this regulation should be carried
out at the federal level. The prior section argued that imperfect information may result in an inefficient
number of interventions. One of the objectives of the benefit-cost analysis presented in this report is to
demonstrate that net benefits to society can be increased. In the case of lead-based paint hazards, the
3-6
~403 EA
-------
necessary information has two parts: identification of situations which present a hazard and selection of
the appropriate response to address the principle source of the problem. A regulation in the nature of
9403 promotes the efficient identification of lead-based paint hazards by providing a metric to use as an
indicator of risk. This increases the amount of information available to the consumer at a very low cost
to the consumer. In this analysis, maximizing net benefits is the principal criterion used to determine a
range of possible candidate standards. In this way, actions under the rule can be targeted to address both
the source(s) oflead (e.g., soil, paint, dust on floors, dust on window sills) that are generating the greatest
risk within the housing unit and the households that will receive the greatest benefits (e.g. households
with new born children).
As written by Congress, various sections of Title X address different parts of the imperfect information
problem. By setting hazard standards, 9403 helps consumers identify situations that subject them to risk.
Without this information, consumers may be more likely to either underestimate or overestimate the value
of an intervention. Hazard standards would provide necessary, although not sufficient, information for
making an informed choice. In addition, the consumer needs information on the cost and effectiveness of
the various lead hazard control options available (e.g. removal oflead-based paint or lead contaminated
soil, encapsulation of lead-based paint, dust clean-up). This information need is addressed by 9402,
which assures a supply of trained and certified personnel to identify and control lead-based paint hazards,
and 9406 and 91018 which provide information about lead-based hazards to the population in general
and in particular at the time of property transactions. In addition, 9402 reduces transaction costs, by
assuring consumers that the information provided to them about their particular situation will be accurate
and complete.
Lead hazards are found in residences in all parts of the nation. Federal regulations can promote cost
savings by encouraging coordination among jurisdictions with resulting economies of scale. For
example, training and certification costs are reduced where states share the same requirements and
provide for certification reciprocity. They also promote partnerships in developing the most cost-
effective ways to address lead-paint hazards. In 9404, the Act encourages the individual States to adopt
the federal 9403 regulations, as well as federal regulations from other sections of the Act, adapting them
to the specific conditions that exist in the States. By establishing a benchmark, 9403 sets a standard for
action which holds throughout the nation, independent of state and local circumstances. States have the
option of imposing requirements that are more stringent than the federal procedures. States and localities
are in a better position to determine how the hazard standards are used and how to adapt their
implementation to local circumstances.
In addition, the Act authorizes certain federal expenditures to partially achieve the national goal of
eliminating lead-based paint hazards. Authorized federal expenditures include federal grants for
evaluating and reducing lead-based paint hazards in non-publicly owned or assisted housing, risk
assessments and interim controls in federally assisted housing, and inspections and intervention of lead-
based paint hazards in all federally owned housing constructed prior to 1960. The 9403 identification of
lead-based paint hazards is necessary to implement those federal expenditures in a manner that develops
the most cost effective methods.
~403 EA
3-7
-------
3.4
Regulatory Options
Options for government regulation fall into four general classes of instruments: (1) information provision
and labeling, (2) performance or technical standards, (3) bans or restrictions on use, and (4) economic
incentives. The first ofthese is most closely linked to the primary condition contributing to a market
failure, as described in Section 3.2. Consequently, directly addressing the lack of adequate information
will be the focus of this section and the analysis presented in this report. Examples of how the other three
classes of instruments might be applied are presented to illustrate their potential. To some extent, these
other instruments are used in other parts of Title X to provide an integrated approach. For example,
9402 of TSCA Title IV provides performance and technical standards in the form of training and
certification requirements and standards for performing lead-based paint hazard identification and control
activities. Prior laws have banned the use of lead-based paint, and 9402 restricts the use of certain
hazard control techniques.
3.4.1 Information Provision
A draft regulator's guide on economic incentives under TSCA identifies three circumstances that are
particularly favorable to making the provision of information an appropriate instrument for regulation
(Eyraud, 1993). The first circumstance - that there is a significant lack of information that generates
exposure problems - has already been identified as a strong likelihood. To rectify this circumstance, a
corollary condition has to be met. The new information has to be able to induce exposure-reducing
behavior. While additional information will alter the behavior of some portions of the population, it
should not be assumed to completely bridge the gap between socially-desirable and observed exposure-
reducing activites. Information programs are appealing as a means of regulation in part because they do
not impose direct burdens on the economy. One of the dangers, though, is that the absence of a direct
burden will come at the expense of being ineffective. This does not have to be the case. Collectively,
environmental and other public health programs have amassed substantial experience in learning about
what does work and what does not work in risk communication. This expertise should be tapped to
render any information approach effective.
The second circumstance favorable to effective information provision is situations where the exposure is
not created by an externality beyond the exposed individuals' control. In other words, the affected
population has to be able to put the information to good use. While externalities between current
intervention and future beneficiaries were identified as a possible cause of market failure, these do not
prevent information from being effective. For example, if a house is not abated over the next 30 years,
the occupants at that future time can decide to undertake intervention if they have the right information to
motivate their decision. This circumstance appears to apply to the exposures from lead in residential soil,
dust, and paint. There is, however, at least one major exception. Financial constraints can prevent even
the best informed household from taking effective steps to reduce exposure. As homeowners, households
may not be able to afford intervention or other hazard controls. As renters and buyers, they may not be
able to afford housing free of lead-based paint hazards. It is important to note, however, that this
impediment is not unique to an information approach.
The third circumstance favorable to the use of an information approach is where other regulations would
lead to greater adverse economic impacts. Although its effects are indirect (working through changes in
behavior rather than by direct enforcement), an information approach does create economic impacts.
3-8
~403 EA
-------
Whether these economic impacts are greater or less than those of other regulations is unknown at this
time because other regulations have not been studied in as much depth.
The analysis presented in this report focuses on defining a particular type of information - hazard
standards - by comparing the costs and benefits under assumed levels of activity by property owners.
Promulgating such hazard standards is one means of implementing 9403 of TSCA, which calls for EP A
to identify lead-based paint standards, lead-contaminated dust, and lead-contaminated soil. One
objective in promulgating such hazard standards is to fill part of the information gap that has been linked
to sluggish rates of intervention of lead-contaminated homes. Specifically, these hazard standards are
intended to indicate thresholds at which EP A recommends that certain forms of hazard controls take
place. As such, they can lower the information costs for homeowners making a decision about whether to
act and thus increase the demand for intervention and other control measures. It is important to note that
by providing such hazard standards, EP A will not be eliminating the information costs altogether. For
example, the costs of testing for levels of lead in soil, dust, and paint are still substantial. These costs are
considered in this analysis. Also, any public information campaign to motivate households to be
concerned about and test for lead contamination (analogous possibly to the public campaign currently
being waged for radon) will impose costs. These have not been explicitly considered here.
3.4.2 Other Regulatory Options
Other regulatory instruments may also be effective for addressing the market failures that have led to the
exposure of children to lead-based paint hazards. While some of these alternative instruments are
utilized in other parts of Title X, they have not been examined in the context of 9403 to the same extent
as the primary instrument considered - information provision. Suggestions for alternatives that might
be investigated further are provided in Exhibit 3.2. The list is meant to be illustrative rather than
exhaustive, particularly where economic incentives are concerned. The feasibility and advisability of
these alternatives could vary widely.
3.5 Overview of Analytic Approach
This section summarizes key aspects of the risk assessment and its linkages with the economic analysis,
and provides an overview of the methodology used to estimate the costs and benefits of the proposed
9403 lead hazard levels. The impetus for assessing the costs and benefits of lead hazard control is to
provide EP A with an estimate of society's potential return from the proposed regulation. The cost
estimates represent costs incurred by individuals, both testing and control costs to reduce lead exposure.
The benefit figures are estimates of the amount society will gain if the adverse health effects caused by
exposure to lead hazards are avoided. The evaluation based on a comparison of benefits and costs is
straightforward. If benefits exceed costs then the lead hazard levels are expected to result in a net gain to
society. Conversely, if costs exceed benefits then the lead hazard levels are expected to result in a net
loss to society.
~403 EA
3-9
-------
Exhibit 3-2
Other Regulatory Alternatives
Type of Instrument
Possible Application
Section 1018 ofthe Residential Lead-Based Paint
Hazard Reduction Act of 1992 requires the disclosure
of lead-based paint or any known paint hazards in the
sale of target housing (any housing constructed prior to
1978). This provision could be extended to provide
information on soil and dust hazards, not just paint, at
the time of sale of any housing, not just target housing.
Labeling
Technical and Performance Standards
Hazard levels could be enforced through performance or
technical requirements. Owners of homes where
children are present would, for example, have to keep
paint in good condition, and reduce and keep soil and
dust contamination below the hazard levels. Technical
standards could specify exactly what control actions are
necessary if hazard levels are exceeded. While ~402
establishes training and certification requirements, and
work standards, it does not target housing with young
children.
Bans and Restrictions of Use
Restrictions could be placed on the access of young
children to homes where lead contamination is of
concern. These restrictions could include exclusion
from occupying such homes or from spending extensive
amounts of time in them, or prohibitions from accessing
particular areas, such as rooms with paint in
deteriorated condition or bare play areas outdoors where
soil contamination is high.
Economic Incentives
A quota could be established for the numbers of homes
allowed to have excess lead contamination. The quota
could be allocated by a system of marketable
allowances. Homes without allowances would have to
undertake intervention or accept restrictions on their
accessibility to young children.
The most efficient policy in economic terms is the one that maximizes net benefits. Estimated benefits
and costs are typically interpreted as a proxy for society's gain or loss from the proposed regulation, since
the actual costs and benefits are unobservable to researchers and regulators. As such, the benefit-cost
comparison is often used with other criteria to evaluate the societal return from proposed regulations.
3-10
~403 EA
-------
3.5.1 Summary of Risk Assessment
Scope of Analysis
This analysis considers children, from birth through the sixth birthday, living in homes that were built in
1997 or earlier. Because the potential for lead exposure in the set of currently contaminated homes may
remain for some time, the analysis considers children born during the next 50 years, from 1997 to 2046.
Exposure, health effects, and benefits are calculated separately for the cohorts born in each of the 50
model years.
Characterization of Exposure
The effects of children's exposure to dust and soil contaminated by lead-based paint are the focus of this
study. Pica, the direct ingestion of paint chips, is also analyzed as a source of childhood lead exposure.
Air, water and food sources of lead are not analyzed because of prior national reductions in their
contamination.
The data source used for estimating exposure in children is the HUD National Survey of Lead-Based
Paint in Housing (USEPA 1995), referred to as the "HUD data" in this report. Conducted in 1989-
1990, this survey measured lead levels in paint, dust, and soil within 284 representative, privately-owned
and occupied housing units built before 1980. Each surveyed home was assigned a national sampling
weight (USEPA,1995). Units built in 1980 or later were not included in the survey since they were
assumed to be free of lead-based paint. Because this analysis considers homes built through 1997, the 28
BUD survey units built between 1960 and 1979 and containing no lead-based paint were used as
templates for describing homes built 1980-1997. The total number of homes built during this latter
period was simply divided by 28 to give an equal weight to each home type.
The unmodified characteristics of these total 312 "BUD home types" are used to represent the baseline
lead levels found in US housing stock throughout the modeling period, 1997-2046. To represent the
"post-intervention" lead levels caused by the introduction of national lead hazard standards, the values
for different BUD homes are reduced based on the candidate standards in question, the hazard control
interventions that consequently take place, and the effectiveness of interventions as specified in the Risk
Assessment (Battelle 1997).
Determination of Blood Lead Distributions: NHANES III and the IEUBK and Empirical Models
To characterize the baseline national distribution of blood-lead concentrations in each cohort of children,
this analysis uses data for one to two year-olds from the Third National Health and Nutrition
Examination Survey, Phase 2 (NHANES Ill). These data yield a blood-lead geometric mean (GM) value
of 3.14 flg/dL, and a geometric standard deviation (GSD) of 2.09 flg/dL. It is assumed that these figures
describe a lognormal distribution, and that without interventions, this baseline distribution would remain
constant for all cohorts born during the 50-year model period. The latter assumption is a basic premise
of the risk assessment.
The NHANES-derived blood lead GM and GSD are assumed to stem from exposure to the baseline
ambient lead levels reported in the HUD data. So while NHANES data can be used to characterize the
baseline blood lead level distribution, they cannot be used for post-intervention distributions, when
environmental lead levels are reduced. Instead, the EP A has two independent models for predicting the
blood-lead GM and GSD from ambient lead conditions in a given population. One model attempts to
~403 EA
3-11
-------
model the physical and biological processes underlying the relationship while the other uses a
statistical/epidemiological approach.
Each model is used within separate analyses to calculate blood lead distributions for both baseline and
post-intervention ambient lead levels. Model inputs for the baseline come from the baseline HUD data
for all 312 home types, while post-intervention inputs are based on reduced lead contamination levels at
the same homes. Predicted blood lead GM's and assumed GSD's for each HUD home type are weighted
and aggregated using statistical formulas to derive a national predicted blood-lead distribution for both
baseline and post-intervention scenarios.
Modeled post-intervention blood lead distributions cannot be directly compared to the baseline NHANES
distribution in a meaningful way. Therefore, the ratio of the predicted post-intervention blood lead GM
to the predicted baseline GM is multiplied by the baseline NHANES GM to derive the final modeled
post-intervention GM for each cohort. The same procedure is repeated for GSD's in order to define a
distribution, which is then compared to the baseline distribution and used to measure health effects
resulting from the standards.
The first of the two models used to predict blood-lead distributions from ambient lead conditions is
EPA's Integrated Exposure, Uptake, and Biokinetic (IEUBK) Model. As its name implies, this model
utilizes exposure, uptake and biokinetic information to predict a distribution of blood-lead levels in
children corresponding to a specific combination of environmental levels. Various parameters are fixed,
including daily intake of dust and soil, and the intake of lead from other sources such as air, water, and
diet. The variable inputs for this analysis are floor dust lead concentration and soil lead concentration.
The effect of lead-based paint on children who exhibit pica is estimated after the IEUBK model is run,
following a procedure described in Battelle (1997).
The second model used to predict blood-lead distributions is an empirical model. This model was
developed using data from the Rochester Lead-in-Dust Study to estimate the relationship between blood-
lead levels in young children and the observed level of lead in environmental media (paint, dust and soil)
from their primary residence. The empirical model as originally developed is log-linear in form,
expressing the natural-log transformed blood-lead concentration as a linear combination of natural-log
transformed exposure variables. The model was adapted slightly by Battelle to accommodate the HUD
data. The final variable inputs are floor dust lead load, window sill dust lead load, soil lead
concentration, and an indicator for the presence of deteriorated lead-based paint and incidence of pica.
Health Effect Endpoints
This analysis considers the following health endpoints: reduction in IQ, cases of IQ less than 70, cases of
blood lead levels greater than 20 f-lg/dL, and cases of blood lead levels falling into seven categories
defined by the CDC. A voidance of adverse health effects is then translated into monetary benefits.
Reduction in IQ may stem from any level of exposure to lead hazards. The next two effects are used to
represent cognitive effects more likely seen in children with unusually high levels of exposure. The
seven blood lead categories correspond to different sets of general symptoms, each with a different
screening and medical treatment protocol recommended by the CDC (CDC 1991).5
The categories cover all possible blood lead levels; children with blood lead under 20 flg/dL are not
expected to show significant symptoms or require medical intervention. Screening is still recommended.
3-12
~403 EA
-------
The IQ point loss for each cohort is determined by calculating the arithmetic mean total population blood
lead level from the OM and OSO. The arithmetic mean is then multiplied by the size of the cohort and
converted to the total number of IQ points lost using the factor 0.257 from Schwartz (1994). The
difference between the IQ points lost for the baseline and post-intervention scenarios yields the IQ point
loss avoided, which is the basis for benefits calculation.
The fraction of a cohort with IQ less than 70 is determined by a methodology based on Wallsten and
Whitfield (1986), using the cohort blood-lead OM and OSO. The same two parameters and the same
statistical procedure are used to calculate the fraction of each cohort with blood lead levels above 20
I-Lg/dL, and the fraction falling within the seven categories defined by the COc. The OM and OSO are
log-transformed to the mean and standard deviation of a normal distribution; then the areas under the
normal curve corresponding to the blood lead level ranges of interest are computed as percentages of the
total area under the curve.
Cohort fractions are multiplied by cohort size to yield the number of children falling within a given health
effect category. From the baseline to the post-intervention scenario, the difference in the population
within each category n IQ under 70, blood lead level over 20 I-Lg/dL, or blood lead level within a CDC
category n is the basis for benefits calculation.
3.5.2 Linkages Between Risk and Economic Analysis
The methodology used to evaluate the economic return from the ~403 hazard levels has several linkages
with the risk assessment methodology, which in turn relates paint condition and the amount of lead in
dust and soil to blood lead levels and health effects. Exhibit 3.3 illustrates the basic linkages between the
risk assessment and economic modeling, and the scheme of the overall analysis. The links that are most
relevant to this overview are those that connect candidate hazard levels and the presence of lead in each
housing unit with hazard control choices and costs, and those that connect the presence of lead to blood-
lead levels and thus to health effects and economic damages and benefits. The first set of linkages
outlines the cost estimation process, while the latter describes the benefit estimation process.
Costs of hazard control are calculated using the unit costs of individual intervention activities, in
combination with the timing and number of interventions. The occurrence of interventions depends on a
variety of factors linked with the risk assessment. The primary intervention "trigger" is the birth of a
child in a home. When a child is born into a housing unit whose ambient lead levels exceed the candidate
standards levels, the appropriate intervention is assumed to occur. This initial intervention is repeated as
necessary until the child turns six years of age, at which time additional interventions will cease unless an
additional child has been born during this time period. Interventions in the home will recommence if
another child is born after this period, and will follow the same repeat routine.
~403 EA
3-13
-------
Exhibit 3.3
Linkages Between the Risk Assessment and Economic Methodologies
Baseline
Ambient
Conditions
Baseline
Blood Lead
Distribution
Baseline
Human Health
Effects
Baseline
Economic
Damages
Hazard
Control
Costs
NET
BENEFITS
Benefits
Post-Action
Ambient
Conditions
Post-Action
Blood Lead
Distribution
Post-Action
Human Health
Effects
Post-Action
Economic
Damages
Benefits of hazard control are calculated using estimates of "avoided" economic damages corresponding
to avoided adverse health effects. As described above, these avoided economic damages include foregone
income due to reductions in IQ, as well as increased educational and medical costs connected with
unusually high levels of exposure. The model defines "avoided" as the difference between the baseline
scenario which assumes no intervention activity and various intervention or ex post scenarios, each of
which assumes a different specification of lead hazard standards and hence intervention activities. In the
analysis, benefits are calculated for children whose exposure to lead is reduced for the period from birth
to age six.
In Exhibit 3.3, the boxes along the top represent the analysis of baseline conditions, yielding an estimate
of the economic damages resulting from the baseline conditions. The boxes along the bottom of the
exhibit represent the analysis of ex post conditions; each scenario or potential lead hazard definition is
analyzed separately. A comparison of the baseline economic damages and the ex post economic damages
yields an estimate of the benefits of actions performed under the scenario. This is represented by the far-
right box in the middle row of Exhibit 3.3. From these benefits, the analysis subtracts the corresponding
costs to get net benefits.
i403 EA
3-14
-------
The evaluation of candidate hazard standards consists of calculating net benefits for a wide range of
candidate levels and then comparing the candidate hazard standards in terms of net benefits. The
candidate hazard standard yielding the largest net benefits provides the greatest benefit to society.
3.5.3 Summary of Integrated Analysis
This section is divided into five parts; the first four correspond to the linkages delineated above. The
first part discusses the connection between hazard levels and hazard control choices. The second briefly
discusses the link between hazard control choices and ex post ambient conditions and new blood lead
levels. The third part focuses on the costs of hazard control, while the fourth discusses the benefits of
hazard control. The final part discusses the choice of discount rate.
Hazard Levels and Hazard Control Choices
In 9403 of TSCA Title IV, Congress instructs EP A to define hazard levels for lead in soil, interior and
exterior paint, and window sill and floor dust. The analysis presented in this report assumes that hazard
levels induce specific intervention activities in homes where the levels are exceeded6. In other words, if
there is a problem then there is a response action, and there is a one-to-one relationship between the
problem and the lead hazard control response.? For example, if lead levels in the soil exceed the hazard
levels, there will be an appropriate soil intervention, but there will not be an interior paint intervention in
response to elevated levels of lead in soil.
Six intervention or hazard control activities are considered. These include high and low intensity
interventions for interior paint and exterior paint, and a single intervention each for soil and dust. High
intensity hazard controls of interior and exterior paint hazards occur when deteriorated lead-based paint
(LBP) is extensive. Low intensity hazard controls of paint hazards occur when deteriorated LBP is
present but not extensive. Soil intervention activities occur when the soil-lead concentration exceeds the
soil standard. Dust hazard control occurs when the floor dust loading exceeds the floor dust standard, the
window sill loading exceeds the window sill dust standard, or when it is required to accompany another
intervention type, either high intensity interior paint control or soil removal.
To determine whether or not levels are exceeded and thus when interventions will occur, baseline ambient
conditions for all model homes represented in the HUD data are compared to each candidate hazard
standard. Hundreds of alternative sets of hazard standards are specified to permit the comparison of
results and to examine the relative efficiency of different specifications. The hazard standard is not
varied for paint, however, and the level used in this model was determined by the US EP A. Baseline
ambient conditions are estimated using the US Department of Housing and Urban Development (RUD)
National Survey Data. See Battelle (1997) for a full description of the methodology used to characterize
ambient dust, paint, and soil conditions within the US housing stock.
This is a strong assumption (assuming hazard control occurs automatically) and overlooks factors such as
informational problems, differences in income, and variation in individual preferences and behavior.
Alternative approaches to modeling the response to lead hazard levels could generate different results.
A small portion of the homes will have no intervention even though the ambient lead levels are above the
standards because no children will be born in them.
~403 EA
3-15
-------
Risk Assessment: Moving from Ambient Conditions to Blood Lead Distributions to Health
Effects
The baseline scenario assumes that no additional hazard controls will occur. For each candidate hazard
level evaluated, the model calculates ex post ambient conditions in each HUD home type. Using the
approach described in 3.5.1, these ex post ambient conditions are used to estimate a single ex post blood
lead distribution for the entire cohort born each year of the model run, 1997-2046. In other words, the
"post-intervention" blood lead distribution reflects ambient conditions after induced intervention
activities. In addition, population blood lead distributions over time reflect changes in the housing stock
that are expected to alter the distribution of lead exposure within homes. This is because homes built
before 1980, which may have lead-based paint, are older and are lost to demolition at a faster rate than
homes built from 1980-1997, which do not have lead-based paint. As lead-free homes comprise a
greater percentage of the modeled national housing stock over time, exposure to lead hazards will drop
regardless of regulations aimed at controlling health hazards.
The incidence of four health effects are estimated as a function of attributes of the blood lead distribution.
These calculations are made for the baseline and "post-hazard control" blood lead distributions on an
annual basis. "A voided" health effects represent the difference between the baseline and "post-hazard
control" incidence estimates for each model year cohort, and serve as the basis for monetized benefits
calculations.
Cost EstimatesB
Drawing on a variety of sources described in chapter four, unit cost estimates are derived for each of the
hazard control activities, in terms of cost per intervention activity per home. Representing average costs,
unit costs are calculated for each of the testing and intervention activities. Separate cost estimates are
developed for single and multi-family housing units; multi-family unit costs are developed by adjusting
the single family cost estimates to reflect the smaller size of multi-family units and the smaller yards of
multi-family units. Hazard control activities include maintaining, removing or encapsulating deteriorated
lead-based paint, removing dust, and removing soil.
For each hazard standard scenario, the choice of intervention activity in a home depends on its baseline
ambient conditions and the hazard standard specified in the scenario. The hazard control standards work
in conjunction with unit costs, the birth trigger, and repeat rules to determine the home's total cost of
intervention during the model run period. It is assumed that non-pennanent interventions are repeated
when their duration is exceeded if a child under six is present, with attendant costs. Exhibit 3.4 presents
the assumed duration for each intervention. In other words, costs incurred by a particular housing unit
accrue over time depending on the assumed hazard control choices. For example, high intensity paint
interventions are assumed to have a duration of 20 years. Therefore, if a child under six is present 20
years after the ir}itial high intensity interior paint intervention in a home, the intervention will be repeated
at that time. If no child is present, an intervention will not be perfonned until such future time as a new
child is born into the home.9
A more detailed presentation of the estimation of costs appears in Chapter 4.
In some cases, repeat interventions will be performed after the last year of the modeling period, 2046, to
protect children born in 2046 or earlier (during the modeling period). For instance, if a high intensity
interior paint intervention is performed in a home in 2028, it will be repeated in 2048 if a child under six is
present who was born before 2047.
3-16
~403 EA
-------
Costs incurred after the first year are discounted to the first year using an annual discount rate of 3
percent. 10 The total model cost estimate is the sum of the cost of hazard controls in all homes for each
year and represents the present value of the assumed stream of intervention costs.
Exhibit 3.4
Duration for Each Intervention
Intervention
High Intensity Interior Paint
Low Intensity Interior Paint
Duration in Years
Paint: 20
Paint: 4
High Intensity Exterior Paint
Low Intensity Exterior Paint
Paint: 20
Paint: 4
Dust
Dust: 4
Soil Removal
Soil: Permanent
Benefit Estimates"
Lead hazard control choices work in conjunction with the birth trigger and intervention repeats to
determine the stream of costs. At the same time, control choices are less directly linked with benefits.
The effect of intervention choices on benefits is realized through their effect on ex post ambient
conditions, health effects, and medical costs. Exhibit 3.5 presents the ex post ambient conditions
associated with each intervention. The benefits of 9403 are "avoided" health effects, where "avoided"
health effects are the difference between the incidence of health effects in the baseline and post-hazard
control scenarios. The analysis quantifies the benefits from medical screening and treatment costs
associated with a wide range of blood lead level categories, as well as from three specific health effects:
lost IQ points, IQ less than 70 points, and blood lead levels greater than 20 /lg/dL. In each case, the
economic value is a proxy for society's willingness to pay to avoid the health effect.
Since lead hazards have been linked with a variety of health hazards for children and adults in addition to
those analyzed here, the benefits estimates are likely to understate the societal return from the 9403
hazard levels. Furthermore, secondary benefits such as improved energy efficiency due to new windows
and increased aesthetic appeal due to repainting also are not included.
10
The sensitivity analysis presents costs, benefits and net benefits calculated with a discount rate of 7 percent
as well.
11
A more detailed presentation of the estimation of benefits appears in Chapter 5.
~403 EA
3-17
-------
Exhibit 3.5
Post-Intervention Ambient Conditions
Floor dust lead loading level reduced to 40 ug/W
Window sill dust lead loading reduced to 100 ug/ft2.
Effect on dust lead concentration depends on other
interventions implemented (see Battelle 1997)
. Soil: Soil lead concentration reduced to 150 ppm.
. Oust: Floor dust lead loading level reduced to 40 I-Ig/ft2.
Window sill dust lead loading level reduced to 100 I-Ig/ft2
Effect on dust lead concentration depends on other
interventions implemented (see Battelle 1997)
Note: Lead levels remain constant in any case where starting levels are lower than assumed post-intervention ones.
Intervention
High Intensity Interior Paint
Low Intensity Interior Paint
High Intensity Exterior Paint
Low Intensity Exterior Paint
Dust
High Intensity Soil
Assumed Post-Intervention Condition
. Paint Deteriorated interior LBP made inaccessible
. Oust Floor dust lead loading level reduced to 40 I-Ig/ft2
Window sill dust lead loading level reduced to 100 1-19/ft2.
Lead concentration level reduced by 80%
. Paint Deteriorated interior LBP made inaccessible
. Oust Lead concentration level reduced by 80%
. Paint Deteriorated exterior LBP made inaccessible
. Oust Lead concentration level reduced by 80%
. Paint Deteriorated exterior LBP made inaccessible
. Oust Lead concentration level reduced by 80%
. Oust
The economic value of avoiding lost IQ points is approximated by using an estimate of the foregone
lifetime income due to IQ point loss. IQ affects wage earnings through ability, education, and labor force
participation. The estimation procedure, therefore, has two major steps. First the present value of the
earnings stream of an average newborn is estimated. Second, available economic literature was used to
estimate the percentage increase in lifetime earnings one would expect from a one point increase in IQ.
Both indirect and direct effects on earnings were considered; the direct effect is the sum of the effects of
IQ on employment and on earnings for employed persons, holding years of schooling constant, and the
indirect effect is the impact of IQ on years of schooling (with subsequent effects on probability of
employment and earnings of employed persons). One final correction is made to this estimate. The cost
of the marginal increase in educaton due to the increase in IQ must be subtracted from the gross increase
in lifetime earnings in order to obtain the net benefit per IQ point. There are two components to this cost:
the direct cost of the education (e.g. tuition) and the opportunity cost of lost income during education.
Subtracting these marginal costs yields an estimate of $8,346 per IQ point lost (1995 dollars).
The economic value of avoiding cases of IQ less than 70 is approximated by using avoided special
education costs. As defined, these education costs are incurred from age 7 though age 18. Similarly, the
economic value of avoiding cases of blood lead levels above 20 flg/dL is proxied by using avoided
compensatory education costs. In this case, the education costs are assumed to be incurred from age 7
through age 9. With increases in blood-lead, there are increases in monitoring and medical intervention
3-18
~403 EA
-------
costs as recommended by CDC. The economic value of reducing the population-wide blood-lead levels is
proxied by the reductions in these monitoring and medical costs. (CDC 1991)
Benefits accrue over time depending on hazard control choices and assumptions regarding exposure by
children. All benefit estimates are discounted to the present using an annual rate of 3 percent. Total
benefits are the sum of benefits calculated for each year or cohort of children protected and represent the
present value of the stream of benefits from the hazard controls. Net benefits are simply the difference
between the total benefits estimate and the total cost estimate. As such, they are an indicator of the
societal gains from hazard controls.
Discount Rate.
As prescribed by the Office of Management and Budget (OMB) in their most recent guidance, the 9403
analysis uses discounting to provide present value estimates of costs and benefits to be realized at
different points in time in the future. (US OMB 1996a) A summary of OMB guidance is presented in
Appendix 3.A. Several characteristics of the actions affected by 9403 require the use of discounting.
First, actions under this rule will occur over an extended period of time (the analysis considers 50 years).
Second, for any particular action, the costs and benefits might occur in different years. Third, unit
benefits from reduced exposure are estimated based on future medical and education costs avoided and
higher future income streams experienced. For accurate measures, future income and avoided future
costs must be discounted to yield present values.
In estimating the total net benefits of alternative definitions of lead hazard levels, the model used in this
analysis "assigns" intervention costs to the year in which an activity occurs, and "assigns" benefits to the
year in which a cohort is born. Due to the repetition of intervention activities, and births into homes
where interventions have already occurred, the costs and benefits may be assigned to different years.
Thus discounting is used to:
.
Estimate the unit benefit value for each type of benefit.12 These are:
the present value of the future increase in income for one newborn based on a one-point
IQ increase;
the future foregone special education costs for one newborn who crosses the threshold
from IQ under 70 to IQ over 70;
the future foregone compensatory education costs for one newborn who crosses the
threshold from blood lead over 20 f-lg/dL to blood lead under this level; and
the future foregone medical costs for one newborn who moves from a higher blood lead
category to a lower one.!3
12
Estimating the unit cost of an intervention activity does not involve discounting, since it is assumed that the
costs will all be incuITed at the time of the activity. Therefore no separate entry is listed for estimating unit
cost values.
13
There are many unit benefits of this type depending on which two categories the newborn passes between;
categories and their associated costs are described in chapter 5.
9403 EA
3-19
-------
.
Estimate the present value of the stream of costs and benefits resulting from ~403
activities in the future. Since the model considers actions taken over the next 50 years, the
benefits and costs from these future actions must be discounted so that they can be summed with
actions taken in the first year. In some cases, children are born into housing units where
interventions have occurred in earlier years and are still effective. Thus there would be benefits
but no costs that year. In other cases, a repeat activity might occur to support benefits counted in
earlier years, resulting in costs but no new benefits assigned that year.
While several factors affect discount rate estimates in general, the selection of the appropriate discount
rate is considered in the context of each of these situations.
General Considerations in Selecting the Discount Rate. Selection of the appropriate discount rate to
use in benefit-cost analyses is a long running controversy and is the subject of a large amount of
economic literature and government policies. The choice of discount rates can make a profound
difference in the results of a benefit-cost analysis, especially for programs like this one where costs
and benefits may accrue at different points over an extended period of time. Most economists
currently believe that no single rate is fully appropriate to use in all situations, even though that might
be desirable for pragmatic reasons. The debate on selecting a discount rate involves two distinct
issues: what type of discount rate is appropriate in a particular situation, and what is the magnitude of
the appropriate discount rate that is selected.
Some agencies specify a single rate to be used in all situations in order to promote consistency among
different analyses. For example, the Congressional Budget Office (CBO) uses the real (Le., inflation-
free) rate on long-term government bonds to discount costs and benefits of proposed legislation and in
preparing the annual budget of the United States. The CBO identified two percent as the appropriate
(and non-changing) discount rate (Hartman, 1990). The Government Accounting Office recommends
using the real rate of return on government bonds over one year in length. Currently this rate ranges
between 2.7 and 3.0 percent (US OMB 1996b). The Office of Management and Budget (US OMB,
1996a) also recommends using a single constant discount rate that reflects the real marginal pre-tax
rate of return on average private investments. The recent OMB guidance refers the reader to its 1992
guidance, in which OMB identified seven percent as the appropriate discount rate to use for
Regulatory Impact Analyses (and indicated that this may change in the future) .
The recent OMB guidance also says that there are circumstances where different discount rates are
appropriate, and encourages RIAs to include alternative calculations using other discount rates when
justified. In the extended guidance on discounting procedures (US OMB, 1988), the OMB identified
situations where the costs of the regulation are almost entirely borne by consumers as a situation where
a different discount rate is more appropriate. Such is the case for ~403, where much of the costs and
quantified benefits are due to lead hazard reduction activities due to voluntary consumer action and
paid for by the consumers.
The debate between using a rate of return on investment capital and the consumption rate of return
focuses on whether investment is being displaced. Some discounting theory emphasizes that one dollar
diverted from productive investment reduces the stream of future production, while a dollar diverted
from consumption would only substitute one type of consumption for another. This diverted capital
argument is the basis of the "shadow price of capital" approach to discounting, which treats displaced
3-20
~403 EA
-------
investment as "costing" more than displaced consumption. The practical difficulty in implementing
this approach is to identify which costs are diverted investments, and which are diverted consumption.
Various pragmatic approaches to this have been proposed and used by the EPA and other government
agencies for regulatory analysis, including the "two-staged" discounting approach (Kolb and Scheraga,
1990), or a single "blended rate" somewhere between the rate of investment return and the rate of
consumption return.
Recent developments in the economic literature have raised serious questions about the extent to which
capital is actually" displaced" today. The displaced capital approach holds that because regulation
diverts funds from alternative investments, some investment opportunities are not undertaken. In other
words, the pool of available capital is assumed to be fixed, forcing some investment to be foregone
when capital is diverted away from investment. While the pool of available capital is relatively fixed
(at least in the short run) in a closed economy, in an open economy capital can flow in from other
countries. The increased demand for investment capital in the United States (created in part to finance
the federal deficit) has attracted large amounts of capital into the country, and most economists feel
this has significantly reduced the pressure that federal borrowing has had on real interest rates. While
the supply of capital is not perfectly elastic, neither is it perfectly inelastic. An elastic supply of capital
reduces the difference between investment rates of return and consumer rates of return.
In addition, estimates of real financial rates of return are lower than many people believe. The real
rate of return on United States government bonds has been near zero percent for most of this century,
while the annual return on a broad portfolio of stocks has averaged near four percent. In general.
stocks have done better since 1980 (averaging 4.26 percent) than in the other periods this century, but
the rate ofreturn may return to historic norms in the future (Freeman. 1993). Thus the range ofreal
rates of return on investment opportunities range from near zero to four percent.
The issues involving the appropriate discount rates and procedures are very complex, and are not
likely to be resolved soon. Much of the recent economic literature summarizing the discount rate
debate concludes that discount rates reflecting either the social rate of time preference or the rate of
return on investments are the appropriate discount rates to use, and there is not that great a difference
between the rates. For example Moore and Viscusi (1990) find no evidence that the rate of time
preference for environmental-related health effects differs from financial rates of return, and cite
evidence that two percent rate is appropriate. Lind (1990) recommends a range of one to three
percent, and Freeman (1993) recommends two to three percent.
Applying these Findings to the Selection of a Discount Rate for the ~403 Analysis. In this analysis,
best practice suggests that both benefits and costs should be measured as consumption foregone and thus
the social rate of time preference has been used for discounting, although what the rate is called is a moot
point if Moore and Viscusi's findings are correct. The reasoning for basing the discount rate on foregone
consumption is that the benefits ofthe rule (e.g., avoidance of an IQ decrement) will provide the
beneficiary with a higher income and therefore greater consumption potential. In the case of costs, the
reasoning for using foregone consumption as the discount rate is based on the manner in which the funds
spent for rule compliance would otherwise be used. This is particularly true for homeowners, who are
likely to view expenditures on improving their home as a consumption expenditure and would not
divert funds from investments for lead hazard reduction activities. On the other hand, the argument
could be made that landlords would reduce investments to finance lead hazard reduction activities.
~403 EA
3-21
-------
There are two responses to this argument. First, since the action is voluntary under 3403, it can be
assumed that the rate of return (say in terms of increased property value) is at least as great as the
landlord would realize on the alternative investment and thus there is no decline in the future stream of
production. Second, even if the landlord were to divert his own investment funds to these activities,
owner-occupied housing units constitute the majority of properties affected by this rule, and thus the
discount rate relevant to homeowners would dominate.
Based on the information presented above, a 3 percent discount rate has been adopted as the most
appropriate rate for use in this analysis. It is used in Chapters 4, 5, and 6 for the estimation of the
present value of costs, benefits and net benefits. Since a 7 percent rate is often used for government
regulations. results using 7 percent are presented as a sensitivity analysis in Chapter 7 to facilitate
comparison among rules.
3-22
~403 EA
-------
Appendix 3. A. Brief Summary of Current OMB Guidance
In its latest economic analysis guidance, OMB discusses the use of discounting in the context of
comparing costs and benefits that occur at different points in time.]4 (US OMB 1996a) As they say:
"discounting takes account of the fact that resources (goods or services) that are
available in a given year are worth more than the identical resources available in a later
year."
The guidance refers the reader to Circular A-94, which specifies that the discount rate should
approximate the opportunity cost of capital -- the before-tax rate of return to incremental private
investment. (US OMB 1992) Circular A-94 (last revised in 1992) estimates the rate to be 7 percent in
real terms. Circular A-94 goes on to say that agencies may present alternative discount rates in the form
of sensitivity analyses, along with a justification for the consideration of these alternative rates.
The guidance then goes on to discuss the shadow price approach -- which it describes as the preferred
model for discounting. One of the main drawbacks of this approach, however, is its complexity. It
requires that investment-related benefits and costs be converted to their consumption-equivalents, which
in turn, requires the calculation of the "shadow price of capital." Estimating the shadow cost of capital
requires assumptions about the extent to which government actions -- including regulations -- crowd out
private investment, the social (i.e. before-tax) returns to this investment, and the rate of reinvestment of
future yields from current investment. In addition, the rate of time preference must be estimated, since
this is the appropriate rate to use when discounting future consumption to obtain the present value of the
future stream of consumption equivalents.
The guidance also briefly addresses the issue of intergenerational comparisons. OMB suggest that a
special social rate of time preference might be used because the social marginal rate of substitution
between the well-being of members of successive generation may be less than the individual rate of time
preference. ]5
14
Discounting should not be used to adjust for uncertainty nor changes in future prices (neither changes in
absolute prices, as in inflation, nor in relative prices).
15
As with the use of the shadow price approach, OMB asks that agencies consult with them before using a
special social rate of time preference in their benefit-cost analyses.
~403 EA
3-23
-------
References
Axelrad, D. 1993. Guidance on the Preparation of Economic Analyses and Regulatory Impact
Analyses in OPPT. Regulatory Impacts Branch, Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency, Washington, D.C., January, pp. 26-27.
Battelle. 1996. Procedures and Results for Input to the Economic Analysis for Section 403 and
Addenda to Report on Inputs to Economic Analysis. Prepared by Battelle, Columbus, OH, for
U.S. EPA, Office of Pollution Prevention and Toxics, Chemical Management Division, June 17
and August 29.
Battelle. 1997. Risk Assessment to Support Standards for Lead in Paint, Dust, and Soil. Prepared by
Battelle, for National Program Chemicals Division, Office of Pollution Prevention and Toxics,
U.S. Environmental Protection Agency. EPA 747-R-97-006, December.
Boadway, R.W. 1979. Public Sector Economics. Little, Brown and Company (Inc.), Boston, MA, pp.
29-43.
Centers for Disease Control. 1991. Preventing Lead Poisoning in Young Children: A Statement by the
Centers for Disease Control. Public Health Service, Centers for Disease Control and
Prevention, U.S. Department of Health and Human Services, October 1991.
Eyraud, J. 1993. Economic Incentives Under TSCA: A Regulator's Guide. Volume 1. Review Draft.
For the U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics,
Regulatory Impacts Branch. Washington, D.C., February, pp. 3-12.
Freeman, A. Myrick (1993). The Measurement of Environmental and Resource Values: Theory and
Methods. Resources for the Future, Washington, DC.
Hartman, Robert W. (1990). One Thousand Points of Light Seeking a Number: A Case Study of CBO's
Search for a Discount Rate Policy. Journal of Environmental Economics and Management,
Vol18(2):S3-S7.
Kolb, Jeffrey A. and Scheraga, Joel D. (1990). Discounting the Benefits and Costs of Environmental
Regulations. Journal of Policy Analysis and Management. Vol 9(3):381-390.
Lind, Robert (1990). Reassessing the Government's Discount Rate Policy in Light of New Theory and
Data in a World Economy with a High Degree of Capital Mobility. Journal of Environmental
Economics and Management, VoI18(2):S8-S28.
Miceli, T.S.; Pancak, K.A.; and Sirmans, C.F. (1996). An Economic Analysis of Lead Paint Laws,
Journal of Real Estate Finance and Economics, 12:59-75.
Moore, Michael 1. and Viscusi, W. Kip. (1990). Discounting Environmental Health Risks: New
Evidence and Policy Implications. Journal of Environmental Economics and Management, Vol
18(2):S51-S63.
3-24
~403 EA
-------
Schwartz, J. 1994. Low-Level Lead Exposure and Children's IQ: A Meta-Analysis and Search for a
Threshold. Environmental Research, Vol 65: 42-55.
United States Environmental Protection Agency (1986). Air Quality Criteria for Lead, June, 1986.
United States Environmental Protection Agency (1995). Report on the National Survey of Lead-Based
Paint in Housing, April, 1995.
United States Office of Management and Budget (1996a). Economic Analysis for Federal Regulations
under Executive Order 12866, January 11,1996.
United States Office of Management and Budget (1996b). Guidelines and Discount Rates for Benefit-
Cost Analysis of Federal Programs, Circular No A-94, Appendix C, Revised. February 6,
1996.
United States Office of Management and Budget (1992). Guidelines and Discount Rates for Benefit-
Cost Analysis of Federal Programs, Circular No A-94, Revised. October 29,1992.
United States Office of Management and Budget (1988). Regulatory Impact Analysis Guidance
Regulatory Program of the United States Government.
Wallsten, T.S. and Whitfield, R. G. 1986. Assessing the Risks to Young Children of Three Effects
Associated with Elevated Blood-lead Levels. Report by Argonne National Laboratory. Report
No. ANL/AA-32. Sponsored by the U.S. EPA Office of Air Quality Planning and Standards.
3-25
~403 EA
-------
3.26
~403 EA
-------
4. Cost Analysis
4.1
Introduction
As described in Chapter 3, candidate hazard standards are ranked in terms of their net benefits (i.e.,
benefits minus costs) to determine which standards yield the maximum net benefits. This chapter
describes the methodology used to calculate the costs associated with each candidate hazard standard.
In this analysis, two conditions are necessary for an initial intervention to occur in a housing unit.
Interventions are assumed to occur in households where ambient lead levels exceed the lead hazard
standards and a newborn child is expected in the ensuing year. Interventions are repeated as necessary to
provide six years of protection for the newborn child, plus protection for any children born in subsequent
years. Separate intervention activities are defined for interior and exterior paint, dust, and soil. The
overall intervention strategy for a housing unit is composed of a combination of the medium-specific
interventions. The appropriate combination of intervention activities is determined by comparing the
lead levels in the housing unit's soil and dust, and the condition of its paint, to the lead hazard standard
under consideration. For each of the housing units covered in the HUD survey, the unit costs are applied
to the specific intervention strategy appropriate to that housing type. Summing across all 312 home
types, weighted to represent the nation's housing supply, provides an estimate of the aggregate cost of
the hazard standard. All cost estimates in the subsequent sections are presented in 1995 dollars.!
4.2
Estimating Aggregate Costs
As described in the prior chapter, the benefit-cost analysis compares alternative futures over a 50-year
time horizon: a baseline or "no-action" alternative, for which it is assumed that no changes are made to
current ambient lead exposure conditions, and a "post-action" alternative for which it is assumed that the
ambient lead exposure conditions are reduced in specific ways in response to the ~403 standards. In
other words, this is a marginal analysis with a baseline of no intervention, and the marginal costs are the
costs of inspecting and testing housing units for lead, and performing the various intervention activities.
The model used to estimate the aggregate costs of each candidate hazard standard analyzes each of the
312 home unit types in the HUD dataset separately (RUD 1993). These 312 home types include housing
constructed after 1978. According to Title X, the regulations will apply to housing constructed before
1978. However, the analysis assumes that once EP A promulgates these standards, they will be generally
applied to all housing regardless of year constructed.2 While the use of lead-based paint was banned in
1978, dust and soil continue to be contaminated with lead from a variety of sources. Since some homes
built after 1978 may have lead levels that exceed the candidate hazard standards, it is likely that the legal
system and lending institutions will adopt these standards for all housing. For example, the ~403
The Gross Domestic Product (GDP) implicit price deflator was used to convert estimates to 1995 dollars
(Office of the President, 1996). The GDP implicit price deflator is estimated by the Department of
Commerce's Bureau of Economic Analysis and incorporates real income and price inflation changes over
time.
Of the over 25 million homes that exceed the proposed standards, only about 890,000 (or 3.6 percent)
were built after 1978.
9403 EA
4-1
-------
standards may be adopted by courts in detennining when a property owner's decision to not intervene is
considered an act of negligence for which the owner can be held financially liable. Likewise, mortgage
lenders are likely to be more hesitant to finance property acquisitions if the properties exceed the 9403
standards. This was the reaction of lending institutions to the asbestos regulations.
By tracing intervention activities in a single hOllsing unit over a 50-year period, the model estimates the
total present value of costs for that unit. This calculation of total per-housing-unit costs is performed for
each of the 312 housing unit types under each of the candidate hazard standards. Using the weights
assigned to each of the 312 housing unit types and the timing of initial interventions during the 50-year
period, the results are extrapolated to a national estimate. In other words, the present values for each
housing unit type are appropriately weighted and summed to provide an estimate of the aggregate
national costs for the candidate standard.
The model is illustrated in Exhibit 4.1.A. For each of the 312 housing unit types, the first stage is to
compare the unit's baseline or current ambient lead condition to the candidate hazard standards. If the
baseline conditions meet all the standards (presence and condition of lead-based paint, the amount or
loading of lead in household dust, and the concentration of lead in soil), no interventions will occur in any
housing units of that type and they will incur no intervention costs. In other words, these housing units
drop out of the cost calculations for this candidate hazard standard. 3 If the unit exceeds any of the hazard
standards (e.g., has lead levels in its soil that are greater than the standard), then the housing unit type
moves to stage 2 of the analysis.
Since the model assumes that testing and interventions begin only when a child is born into the home, the
housing units are divided into two groups: 1) those where a birth is expected that year, and 2) all the rest.
The housing units in the second group are subjected to this same bifurcation process in the next year,
after their number has been reduced due to demolition. This cycle continues through the 50-year period,
with some housing units receiving their initial intervention each year and units without an initial
intervention that year moving on to be considered in the following year. The current national birth rate,
adjusted in future years to reflect expected changes in the national birth rate, is applied to each of the 312
housing types used in this analysis. (Battelle, 1997)
Each year, the housing units in group one (those experiencing their first birth) move to stage 3. Initial
intervention activities that will bring the unit into compliance with the hazard standard are identified. In
addition, all activities that will need to be repeated for that housing unit are identified and their timing
detennined based on the assumptions that the newborn will live in the unit for at least six years, and
assumptions about future birth and demolition rates (Battelle, 1997). Using the unit costs specific to
each intervention activity, developed later in this chapter, and a 3 percent discount rate, the discounted
present value of costs to be incurred due to this stream of activities is calculated (see Exhibit 4.1.B).
All units will need to be inspected at the time the birth of the initial child occurs in the unit, and inspection
costs are incuITed at that time.
4-2
t403 EA
-------
101
~
o
(,,)
m
:t>
Exhibit 4.1.A
Determ ining Total Costs - Outline of Calculation
Stage 1
Each of the 312 housing unit types in the HUD dataset is analyzed year-by-year over a 50-year period, as follows:
Do the baseline conditions of
housing unit exceed the hazard
standards for paint, dust and/or soil
being anal 'zed?
No
No interventions needed for that
housing type under the hazard
standard being analyzed
~
,
(,,)
Yes
----..
Stage 2
Predict number of housing units of
this type in which the birth of a
child will occur that year for the
first time during the model period
(based on birth rate and number of
houses of this type)
Un;t' when no b;rth I
occurs that year ~
No interventions and no costs that
year for these units
,It
Reduce number of units of this type
by demolition rate
,It
N urn ber of housing units of this type
available for possible initial
interventions next year
Units where birth
occurs that year
Stag e 3
.
Detcrm ine expected stream of
intervention activities needed to
protect newborn for G years and to
protect all children subsequently
born into these housing units.
allow ing for housing unit attrition
(demolition)
U sing estimated unit costs for each
intencntion activity, discounted
from year activity occurs back to
year I of modeL calculate present
value of stream of interventions for
all housing units of this type (see
Exhibit .t.l.B for details)
-------
Since the present value of all costs to be incurred by a single housing unit over the life of the model is
calculated at the time the initial intervention action is taken, this group of housing units drops out of any
further cost calculations under this hazard standard. Each year, another group of housing units of this
type experiences their initial interventions and the present value of all the costs to be increased by those
units is calculated.4 The total costs for all units experiencing initial interventions in a given year are
calculated. The present value of the aggregate costs for all units over all years, is calculated by
discounting the present value of intervention costs from the year the initial intervention takes place back
to the fIrst year of the model and summing discounted costs across all years.
The next sections of this chapter describe the development of the intervention-specific costs, including
the matching of intervention activities to baseline conditions of housing units, and estimates of the
number of homes exceeding the candidate hazard standards. Section 4.3 summarizes the differences
between EPA's approach to estimating costs (as presented in this document) and the approach employed
by the U.S. Department of Housing and Urban Development (HUD) in its Regulatory Impact Analysis of
the Proposed Rule on Lead-Based Paint (HUD, 1996). Section 4.4 describes the data used in the cost
analysis. Section 4.5 presents cost estimates for each of the lead hazard evaluation activities, while
Section 4.6 offers cost, duration, and effectiveness estimates for each of the intervention activities.
Section 4.7 presents the impact of alternative candidate hazard standards in terms of number of homes
that exceed the candidate standards. The fInal section compares the rate of intervention activities
assumed in this normative analysis with likely rates of activity, and discusses the likely costs of
interventions.
4.3
EPA and HUD Approaches to Estimating Costs
While this analysis draws on the cost estimates developed by HUD in its Regulatory Impact Analysis of
the Proposed Rule on Lead-Based Paint (HUD, 1996), for reasons described below the analysis uses
somewhat different estimates of unit- or activity-specifIc costs.
4.3.1 EPA Approach
The g403 regulations will designate separate lead hazard standards for soil, dust, and paint that protect
the public from adverse health effects. Given that hazard standards are set independently for each
medium, costs were estimated on a medium-specifIc basis, and housing owners are assumed to respond to
the hazard levels on a medium-by-medium basis.
It is diffIcult to predict the exact mix of actions the public will take in response to the lead hazard
standards because the g403 regulations are voluntary and there are a variety of possible intervention
approaches. Therefore, for purposes of this analysis, a set of reasonable intervention choices was defIned
for each mediu~.. The types of intervention actions to be undertaken were designated as a function of the
level and characteristics of the lead hazard (e.g., concentration oflead in the soil, condition of lead-based
paint) in a housing unit. This analysis estimated costs for two levels of paint intervention (high and low-
intensity), and one each for dust and soil.
Since all the units of a given housing unit type have the same baseline characteristics, the discounted
present value of the costs to be incurred by one of these units is the same for all units of this type
experiencing their initial intervention in the same year. The discounted present value of costs for this
housing unit type will differ among different years of initial intervention, hazard standards, and across
different housing unit types.
4-4
~403 EA
-------
em -
8 I Cost~f Each of the Interventio;; - I
m I Activities I
» I I
I I
I Low-Intensity Interior Paint I
I I
I I
I High-Intensity Interior Paint I
I I
I I
I I
I I
I Low-Intensity Exterior Paint I
I I
I I
I I
I High-Intensity Exterior Paint I
I I
I I
I I
I I
I Dust Cleanup I
I I
I I
I I
I I
I I
I Soil Removal I
I I
I I
I I
I I
I I
g; I I
L.-- - - - - - - - - - - __I
Exhibit 4.1.B
Determining Total Costs - Estimating Present Val De of Aggregate Costs
I Calculate Present Value of Interventio~osts Incurred l
I by Single Housing Unit of Given Type. Sum the I
I discounted annual costs. I
I Year l: Calculate costs of any initial I
I intervention activities needed to meet I
I ~~~s. I
I I
I + I
I Year 2: Calculate costs of any repeat I
I intervention activities necessary and discount I
I back to Year 1. I
I I
I I
I I
I + I
I I
I I
I Year 3: Calculate costs of any repeat I
I intervention activities necessary and discount I
I back to Year 1. I
I I
I + I
I I
I . I
I . I
I . I
I I
I I
I + I
I I
I Year 50: Calculate costs of any repeat I
I intervention activities necessary and discount I
I back to Year 1. I
I I
'"'- -- - - - - - - - - - - - - - - ""'"
I Calculate Present Value of Interventio; Cost;;!ncurred I
I by All Housing Units of this Type. I
I Sum the discounted annual costs. I
I I
I Year l: Multiply number of housing units I
I experiencing initial intervention in Year I by I
I the per-housing-unit present value of initial I
I and repeat interventions in that house. I
I I
I + I
I I
I Year 2: Multiply number of housing units I
I experiencing initial intervention in Year 2 by I
I the per-housing-unit present value of initial I
I and repeat interventions. Discount back to I
I Year 1. I
I I
I + I
I I
I . I
I . I
I . I
I I
I + I
I I
I Year 50: Multiply number of housing units I
I experiencing initial intervention in Year 50 by I
I the per-housing-unit present value of initial I
I and repeat interventions. Discount back to I
I Year 1. I
I I
I I
I I
I.....- -- - - - - - - - - - - - - - -.....
-------
This approach represents a compromise between estimating a single cost for each medium (dust, soil, and
paint) and establishing cost estimates for a range of detailed lead hazard reduction activities. Data
limitations greatly complicated the cost estimation process. Few data were available on the public
responsiveness to lead hazards. Secondly, variation in housing units confounded the derivation of unit
costs. Furthennore, the limited environmental data collected in the HUD National Survey of Lead-Based
Paint in Housing imposed constraints on the level of detail at which costs could be calculated.
4.3.2 HUD Approach
In its Regulatory Impact Analysis of the Proposed Rule on Lead-Based Paint, HUD presented a cost-
benefit analysis of requirements for evaluation and hazard reduction activities in federally-owned and
federally-assisted housing (HUD, 1996). To do this, HUD estimated costs for various intervention
activities based on infonnation from abatement experts and the Task Force Report (Lead-Based Paint
Hazard Reduction and Financing Task Force, 1995). Costs were assigned to activities covered by the
proposed HUD requirements. This process was simplified by the fact that specific intervention activities
(e.g., visual evaluation, interior paint repair, exterior paint repair, and area cleanup) were designated for
each type of federally-owned and federally-assisted housing in the proposed requirements.
4.3.3 Differences Between the Approaches
There are two significant differences between the EP A and HUD approaches to estimating costs. First,
HUD estimated evaluation and intervention costs at a more disaggregated level than that used in this
analysis. For example, cleanup activities, clearance testing, and window work were separate
interventions in the HOD analysis, while they were combined and included within the medium-specific
interventions of this analysis. The distinction arises because the HUD requirements specifically dictate
which interventions occur and what activities comprise interventions. Because the ~403 regulations are
voluntary, they do not specify any particular actions that must be taken. These voluntary responses may
not include all the activities specified under the HOD regulations. Alternatively, some people may
choose to do more than HUD would require. This analysis, therefore, provides estimates of the costs
associated with average or reasonable levels of hazard reduction work done voluntarily by housing
owners.
A second notable difference is that HUD distinguished between the incremental costs associated with
lead hazard reduction activities and the costs incurred in painting or other rehabilitation activities that
would have been perfonned in the absence of the requirements. For example, the costs of scraping and
repainting a room containing lead-based paint were compared with those of repainting a room without
taking precautions for lead to derive the incremental cost of such an intervention. HOD assumed that
paint and rehabilitation costs not associated with lead abatement were offset by the market value of
benefits associated with improved housing conditions. Subsequently, their analysis only included the
incremental costs in designating the net benefits of the proposed rule. The analysis presented here does
not separate costs in this fashion largely because the housing units covered under the ~403 regulations
are privately owned. In contrast to private housing, public and federally assisted housing benefit from
federally assisted painting and rehabilitation programs. This analysis assumes that private units are
already maintained to the level at which additional expenditures do not return equal or greater offsetting
4-6
~403 EA
-------
financial benefits. Intervention costs include the full costs of lead hazard reduction. Benefits in the fonn
of increased market value were not considered.5
Where appropriate, this analysis combined the cost estimates developed by HUD for their RIA with other
data to calculate intervention cost estimates. Sources in addition to the HUD data were used because in
some cases it was not clear what activities were included in the HUD estimates and because of the wide
scope of the ~403 regulations. The HUD RIA (HUD, 1996) provided limited infonnation on the sources
of the cost data employed. Alternative sources of data were located when the HUD RIA sources could
not be identified or when the HUD RIA did not provide cost estimates for an activity (e.g., lead hazard
screen, soil removal) included in this analysis.
4.4
Data Sources
Costs were estimated for specific evaluation and intervention activities. Cost estimates were derived for
three types of evaluation activities, two levels of intervention for interior and exterior paint, and one level
of intervention for dust hazard reduction and one soil hazard intervention. Primary sources of data used
by this analysis included:
.
the Comprehensive and Workable Plan for the Abatement of Lead-Based Paint in
Privately Owned Housing: Report to Congress (HUD, 1990);
the National Center for Lead-Safe Housing's SpecMaster Database of intervention costs
(NCLSH, 1995) and a NCLSH report on lead hazard control for non-profit housing
organizations (NCLSH, undated);
the Lead-Based Paint Hazard Reduction and Financing Task Force report (Task Force,
1995)
the 1996 HUD RIA on lead paint hazards in federally-owned and federally-assisted housing
(HUD, 1996);
Hometech's Remodeling and Renovation Cost Estimator (1996);
R.S. Means' Repair and Remodeling Cost Data (1996);
American Housing Survey (HUD, 1995);
interviews with HUD research grantees (state and local governments) working on lead
interventions; and
selected interviews with lead testing and abatement finns, as well as landscapers,
commercial cleaning services, and hazardous waste disposal finns.
.
.
.
.
.
.
.
Cost estimates are presented for the evaluation and intervention activities in Sections 4.5 and 4.6,
respectively. Where appropriate, estimates from various sources were combined to assure reasonable
unit cost estimations. Descriptions of lead hazard characteristics by housing units were based on the
HUD National Survey data, and the extent of these lead hazard characteristics dictated the intervention
activities that occurred in each unit.
Housing conditions were compared with candidate hazard standards to determine the types of
intervention activities conducted. For interior and exterior paint hazards, alternative standards are not
Likewise, the analysis did not include any potential decline in market value for housing units that do not
receive hazard reduction activities even though they exceed the ~403 standards.
~403 EA
4-7
-------
analyzed. EPA has specified that high-intensity interventions will occur when damaged lead paint
exceeds 20 square feet and low-intensity interventions for interior and exterior paint hazards will occur
when damage is less than 20 square feet but over five square feet. Alternative hazard standards are
analyzed for soil and dust interventions. While dust interventions are always conducted as part of high-
intensity interior paint and soil interventions, dust interventions were also conducted independently in
homes with dust loading levels above the specified hazard level.
4.4.1 Uncertainty
The uncertainty associated with the cost estimates was largely due to data limitations. Point estimates of
intervention costs were the most common data, and frequently the exact services and/or methods included
in a single intervention estimate were not described in the data. This lack of complete information
occurred often for paint interventions, for which the information on the area and types of surfaces
stabilized or abated and the post-intervention testing were not clearly specified.
Intervention methods will change over time, as information on effectiveness becomes available and
technology changes. This analysis applied unit cost estimates from the available data to the period from
1997 to 2046. Future prices are unknown and it remains unclear whether the use of current prices will
overstate or understate the real costs of testing and abating. For example, with more competition in the
intervention market, prices may decrease in the future. Conversely, future prices may rise as standards
are introduced, because of training costs and performance requirements. In addition to the limitations
imposed by the data and lack of knowledge about future costs of intervention methods, variation in the
housing stock and regional prices introduced uncertainty into the costs of testing and conducting
interventions to address lead hazards.
Additional uncertainty arose from the modeling of hazard reductions. The impact of interventions was
modeled using effectiveness and duration estimates provided by Battelle (1997). In some cases, limited
information was available on effectiveness or duration of intervention activities for lead hazards; some
data on effectiveness with respect to changes in blood lead levels and dust lead loadings were available.
The assumptions about effectiveness and duration may have significant, but unknown, effects on the
model results. Effectiveness and duration estimates are presented for each intervention activity in
Section 4.6.
4.4.2 Multifamily Housing Considerations
Single-family housing unit cost estimates served as the basis for multifamily housing unit cost estimates.
Buildings with greater than four units were considered multifamily buildings by the HUD National
Survey (We stat, 1995). The same definition was adopted by this analysis. Based on this defmition,
multifamily units composed 17.5 percent of the U.S. housing stock in 1993 (DOC and HUD, 1993).
Typical multifamily housing intervention costs differ from single-family housing intervention costs for
several reasons, including smaller average unit sizes, shared costs for soil interventions, and the
possibilities for economies of scale in testing multiple units in a single building. In conformance with
HUD regulations and ~402 training guidance, this analysis assumed that lead inspections in multifamily
buildings use a random sampling approach, eliminating the need to test all units. By doing so, the
4-8
i403 EA
-------
average per-unit costs for testing were reduced in multifamily housing.6 It was assumed that multifamily
housing is predominantly rental housing and that testing will be done by landlords on entire buildings.
This appeared to be a reasonable assumption because the American Housing Survey (AHS) indicated that
89 percent of multifamily units are rental units (DOC and HUD, 1993).
Adjustments for unit size, economies of scale in evaluation, and shared soil areas were made to derive
multifamily cost estimates. Cost estimates for hazard evaluation, paint inspection, interior lead paint
intervention, and dust intervention were reduced to reflect the smaller average unit sizes. On average,
multifamily units were estimated to be 33 percent smaller than single-family units, so a factor of 0.67
was used to adjust costs.7 Additional reductions were made in hazard evaluation and paint inspection
cost estimates to account for the fewer tests required. A factor of 0.77 was used to scale costs for lead
hazard screens, risk assessments, and paint inspections in multifamily housing. This factor was based on
an EP A (1995a) suggestion that 23 units out of 30 (the average number of units in multifamily buildings
from DOC and HUD, 1993) be tested for a statistically valid sample. If testing of multifamily units is
done by individual renters or by landlords as units turn over, testing costs may have been underestimated.
Economies of scale for intervening in multiple units in a single building (e.g., simultaneous paint
abatements in several units of a building) were not addressed as data were not available for this
adjustment. Cost estimates for exterior paint intervention were based on costs of window replacement as
multifamily units do not contain much exterior LBP. Likewise, cost estimates for soil interventions were
based on estimated dimensions of areas in which soil work would be done.
4.5
Hazard Evaluation Costs
Exhibit 4-2 presents the costs estimated for two types of evaluations: lead hazard screen, and risk
assessment. Activities included and data sources are also displayed. A lead hazard screen is a limited
assessment used to determine the absence of lead hazards in a dwelling unit. A risk assessment is a full
inspection for lead hazards in a home and includes dust testing, soil testing, visual assessment of paint
condition, and limited XRF or laboratory testing of paint in bad condition. Testing scenarios were based
upon the HUD guidelines (HUD, 1995), EPA's risk assessment training materials (U.S. EPA, 1995b),
and EPA's regulations under Section 402 of Title IV of TSCA (Abt Associates, 1995a).
In estimating costs of each candidate hazard standard, the model assumes that either a lead hazard screen
(for single family units without deteriorated lead-based paint) or a risk assessment (all other units) is
performed. Testing is done at the time a childbirth is expected and testing is not repeated for a unit.
Costs for testing multifamily buildings may be reduced even further by using a targeted sampling approach
in which units most likely to contain lead hazards are sampled, but the information needed by risk
assessors to do this targeting is difficult to obtain.
American Housing Survey (AHS) data indicated that the median size of multifamily units was 1255 square
feet, the median size of single family units was 1775 square feet (DOC and HUD, 1993 as cited by HUD,
1996), and the mean size of multifamily units was 940 square feet as compared to 1,800 square feet for
single family units (DOC and HUD, 1993, calculated by Abt Associates directly). Because older units are
likely to be smaller than new units and single family average sizes may be skewed by very large units, such
as those built in recent decades, this analysis followed HUD (1996) in estimating multifamily units
affected by lead hazard regulations to be 1,000 square feet and single family units to be 1,500 square feet.
~403 EA
4-9
-------
Exhibit 4.2
Summary of Lead Hazard Evaluation Costs
Type of Testing
Activities Included
Cost per
Single-
Family
Housing Unit
in 1995
Dollars
Cost per
Multi-
Family
Housing Unit
in 1995
Dollars
Source
of
Information
Lead Hazard Screen
inspection of paint condition, collection and
analysis of two composite soil samples,
collection and analysis of two composite
dust samples
212
Mean of
NCLSH,
undated and
Task Force,
1995
Risk Assessment
collection and analysis of ten individual dust
samples, collection and analysis of two
composite soil samples, visual inspection of
paint condition, XRF and/or laboratory
testing of deteriorated paint
456
235
Mean of
NCLSH,1995
and Task Force,
1995
4.5.1 Lead Hazard Screen
Single-Family Unit Cost. A lead hazard screen requires an inspection of paint condition, collection
and analysis of two composite soil samples, and collection and analysis of a minimum of two composite
dust samples. Two sets of estimates were identified for the cost of conducting a lead hazard screen. The
first was from the NCLSH handbook (NCLSH, undated) which reported a cost of $199 for examination
of paint condition, housekeeping standards, six lead dust wipe samples, and soil lead testing. The second
was from the Task Force Report (Task Force, 1995), which estimates a cost for a lead hazard screen
ranging from $150 to $300. The mid-point of this range was used ($225). The cost of a lead hazard
screen was calculated as the average of the NCLSH (undated) and Task Force (1995) estimates: $212.
The analysis assumes that all single-family housing units that do not contain deteriorated paint will
perform a lead hazard screen rather than a risk assessment or paint inspection.
Multifamily Housing Cost. This analysis assumes that multifamily housing units only conduct risk
assessments, rather than lead hazard screens, due to the prevalence of children in multifamily buildings as
well as the risk of liability faced by landlords.
4.5.2 Risk Assessment
Single-Family Unit Cost. A risk assessment is a full inspection of lead hazards in a home, including:
dust testing, soil testing, visual assessment of paint condition, and limited XRF or laboratory testing of
paint in bad condition.8 NCLSH (1995) provided a figure of $537 per dwelling unit but did not indicate
the activities included in the assessment. The Task Force (1995) indicated a range of risk assessment
costs of $200 to $500, with an average of $375 for single-family units and $260 for multifamily units
A typical testing plan requires a visual inspection of paint condition and determination of the lead content
of painted surfaces by either in situ analysis using a portable x-ray fluorescence (XRF) analyzer or by off-
site laboratory analysis of paint chip samples. X-ray fluorescence analysis has the advantage of being
faster, cheaper, and non-destructive when compared with laboratory paint chip analysis. However, XRF
readings are not as accurate as laboratory analysis (HUD, 1990). Chemical spot testing is cheaper than
XRF testing but is not accurate enough for quantitative analysis (HUD, 1995).
4-10
~403 EA
-------
(assuming that composite dust sampling is used). The risk assessment cost estimate for a single-family
unit was calculated as the average of the NCLSH (1995) and Task Force (1995) single-family estimates:
$456. This analysis assumes that all single-family housing units containing deteriorated lead-based paint
will perform a risk assessment rather than a lead hazard screen or paint inspection.
Multifamily Housing Unit Cost. The cost of a risk assessment in a multifamily unit was estimated by
multiplying the single-family cost of $456 by a factor of 0.67 to reflect the smaller size of multifamily
units. See Section 4.4.2 for a complete discussion of this factor.
$306 = 0.67 x $456
The analysis calculates the total cost of evaluating a multifamily building based on the assumption that
23 units out of 30 (the average number of units in a multifamily building from DOC and HUD, 1993)
require a risk assessment. This ratio was based on an EPA (1995a) suggestion that 23 units out of 30 be
tested for a statistically valid sample. If testing of multifamily units is done by individual renters or by
landlords as units turn over, testing costs may be higher. Economies of scale in testing and intervening in
multiple units in a single building were not addressed as data were not available for this adjustment. In
addition, given assumed birth rates, the analysis assumes that at least one child is born into every
multifamily building in year one; thereby triggering a risk assessment in all multifamily buildings and soil
removal where required in year one. In contrast, interior actions taken to abate the lead hazards identified
by the risk assessment occur on a unit-by-unit basis as children are born into the multifamily unit.
4.6
Intervention Costs
Exhibit 4-3 summarizes the intervention cost estimates for lead in dust, paint, and soil. In addition to the
cost estimates, Exhibit 4-3 lists the activities included and the data sources. The approach used to
estimate costs for the different intervention elements is discussed below. High and low-intensity
intervention methods are discussed where appropriate. Duration and effectiveness assumptions are also
addressed. These assumptions were based on Battelle (1997).
4.6.1 Dust Intervention Costs
In practice, the cost of a dust intervention depends on the size of the dwelling unit and the thoroughness
of the cleaning. For purposes of this analysis, dust intervention for controlling lead hazards from dust
that include cleaning of the unit with a HEP A vacuum and wet mopping.
Single-family Dust Cleaning Cost. In addition to thoroughness and size of area cleaned, the cost of
dust intervention depends on whether carpeting and upholstered furniture are replaced. For this analysis,
the dust intervention was defmed as the vacuuming of all rooms in the unit, including floors, woodwork,
window wells, and furniture, with a high-efficiency particle accumulator (HEPA) vacuum, followed by a
wet wipe-down of the unit with a lead-specific detergent. No replacement of furniture or carpeting was
included, although this may be necessary for full cleaning of a highly-contaminated unit. NCLSH
(undated) estimated a cost of $484 for cleaning a three bedroom, one-and-a-half bathroom house of
moderate size after a "medium-intensity abatement." A second NCLSH (1995) source provided an
estimate of $48 per room for HEPA vacuuming followed by a wash with TSP (a lead-specific detergent)
and a second HEP A vacuuming. At an average of 6.2 rooms per single- famil y house (DOC and HUD,
~403 EA
4-11
-------
Exhibit 4.3
Summary 01 Intervention Costs for Lead in Dust, Paint, and Soil
Intervention Activities Included Cost per Cost per Multi-
Single-family family Housing Source
Housing Unit in Unit in 1995 of
1995 Dollars Dollars Information
Dust HEPA vacuuming of all floors, woodwork, 391 262 Mean of NCLSH, 1995 and
window wells, and furniture. Wet wipe-down of NCLSH, undated
unit with lead-specific detergent. No
replacement of contaminated furniture or carpets
included.
High-intensity Complete encapsulation, enclosure, or removal 6,587 4,687 Mean of Lim, 1996; HUD,
interior paint of LBP, including replacement of windows. (4,744 in repeat (3,450 in 1996; and lower bound of
Includes post-intervention dust cleanup with years) repeat years) range in NCLSH, undated
HEPA vacuum and clearance testing.'
Multifamily cost includes disposal of hazardous
waste.
Low-intensity Paint stabilization in one room: repair of 437 437 Mean of HUD, 1996 and
interior paint damaged LBP, repainting, covering and sealing NCLSH, undated
window wells and sills, post-intervention dust
cleanup in room where work was done, and
clearance testing.*
High-intensity Single-family: Complete encapsulation, 5,706 2,275 Mean of HUD, 1996 and
exterior paint enclosure, or removal of LBP. No dust NCLSH, undated for single-
intervention afterwards in interior of unit. family; cost of window
Multifamily: Replacement of seven windows. replacement from Santucci
(pers. comm.) for multifamily
Low-intensity Single-family: Repair of damaged LBP. 807 182 Mean of HUD, 1996 and
exterior paint Multifamily: Stabilization of seven windows. NCLSH, undated for single-
family; cost of window
stabilization from NCLSH
(1995) for multifamily
Soil removal Soil removed up to a depth of six inches and Unit costs from Hometech
disposal in a landfill. Removal may be from area Remodeling and Renovation
away from home and/or three feet around the Cost Estimator (1996) and
perimeter. Includes interior dust clean-up'. R.S. Means' Repair and
Remodeling Cost Data
Removal and replacement of perimeter soil. 2,046 399 (1996) plus area estimates
from HUD's America
Removal and replacement of remote area soil. 7,878 777 Housing Survey (1995) and
Santucci (personal
Removal and replacement of both perimeter and 9,008 901 communication) .
remote area soil.
, The dust cleanings performed in conjunction with the high-intensity interior paint and the soil interventions are essentially the same as
the full-house "stand-alone" dust intervention listed first on this table. The dust cleaning performed in conjunction with the low-intensity
interior paint intervention involves only the room where the paint intervention occurred.
4-12
i403 EA
-------
1993), this yielded an estimate of $298 per house. The final cost estimate for dust intervention in a
single-family home is $391, the average of the two values.
$391 = $484 + ($48 X 6.2)
2
Multifamily Housing. The single-family cost for dust intervention was multiplied by a factor of 0.67 to
reflect the smaller size of multifamily units. The resulting cost estimate is $262, as shown below. See
Section 4.4.2 for a complete discussion of this factor.
$262 = 0.67 X $391
Effectiveness. The effectiveness of dust cleaning depends on the circumstances in which it is used and
whether post-intervention lead levels are measured in terms of loadings or concentrations. Battelle
(1997) estimated that a single cleaning would bring floor dust lead loading levels to 40 Ilg/ft2 or the pre-
intervention levels, whichever is less. Window sill dust lead loadings would be brought to either 100
/lg/ft2 or the pre-intervention levels. Based on Battelle (1997), post-intervention dust-lead concentrations
will vary depending on:
.
whether or not a soil intervention also occurs;
whether or not a paint intervention has occurred, and, if not, whether any deteriorated lead-
based paint exists in the housing unit;
whether or not a dust cleaning has occurred in a situation where no soil or paint intervention
occurred and no deteriorated lead-based paint exists in the housing unit.
.
.
See Battelle (1997) for details on calculating post-intervention concentration levels.
Duration. The duration of dust abatement also depends on the circumstances in which it is used. The
dust intervention was assumed to be effective for four years based on a report by Battelle (1997). In
situations where a dust cleaning is performed immediately preceding the birth of a child, and no soil nor
paint intervention occur, then the dust cleaning is repeated in four years to continue the protection of that
child.
4.6.2 Paint Intervention Costs
Estimated costs of interior lead paint intervention vary widely, in part due to differences in the extent of
intervention:' For example, replacing windows is less expensive than removing all the molding, doors and
wall paint. Additional variation is due to the size of the homes abated and regional price variations.
Several general cost estimates for lead paint intervention existed, but few could be used here because the
work included in the interventions was not specified. Instead, estimated costs from sources that provided
information on the specific activities involved and the size of the unit abated were employed. Two
degrees of paint intervention are considered: high-intensity, where paint damage is extensive, and low-
intensity, where paint damage is limited to a relatively small area.
~403 EA
4-13
-------
Interior Paint Intervention
Single-Family High-Intensity Cost. A high-intensity paint intervention involves encapsulation,
enclosure or removal of LBP in the housing unit, including removal of windows with LBP, and post-
intervention cleaning using a HEP A vacuum. Several estimates from the literature were used to construct
the high-intensity paint cost estimate. Lim (1996) provided a cost estimate of $6,809 for window
replacement, floor smoothing, walVdoor/trim treatment including encapsulation and enclosure, and unit
cleanup for a small unit of up to 1,200 square feet. A second estimate was developed by combining
HUD's (1996) cost estimates for interior abatement, unit cleanup, and clearance testing, plus an
independent estimate of $242 for replacement of each wooden window (NCLSH, 1995) and an
assumption of 12 windows, yielding a total cost of $6,504.
$6,504 = $3,000 + $450 + $150 + (12 x $242)
NCLSH (undated) estimated a range of $6,449 to $16,122 for work on a three bedroom house of
moderate size. Their estimated cost included window replacement, enclosure of walls with gypsum,
occupant relocation, and unit cleanup as well as numerous other activities. The lower bound of this range
($6,449) was used in this analysis because some of the intervention activities included in deriving the
higher cost estimates (e.g., $8,383 for the abatement of a 2-story townhouse) were not included as part of
the defined high-intensity paint intervention. Averaging the Lim estimate ($6,809), the HUD estimate
($6,504), and the lower bound of the NCLSH estimate ($6,449) provided a cost estimate of $6,587 for a
high-intensity paint intervention in a single-family home. While this estimate may be low for removal of
LBP from all surfaces of all rooms in a large unit, it was reasonable for units with a mixture of LBP and
non-LBP and for encapsulation methods.
The analysis estimates costs over a 50-year time frame; therefore, high intensity interior abatements
(assumed to have a duration of 20 years) may occur up to three times in one home. However, the
estimated cost of a high intensity interior abatement includes some permanent measures that are not
recurring, such as the replacement of windows. The cost estimates for repeat high-intensity paint
interventions, therefore, exclude window replacement costs. Deducting window replacement costs used
in each estimate from Lim (1996) and HUD (1996) results in an estimated cost of $5,646 and $3,842,
respectively. The final cost estimate for a repeat high intensity paint intervention is $4,744, the average
of the two values.
Single-Family Low-Intensity Cost. Low-intensity paint intervention includes stabilization of the
deteriorated interior LBP in a room, including repair of the LBP, covering and sealing the window sills
and wells to ensure cleanability, and cleanup in the room. Using cost estimates from NCLSH (undated)
for paint repair, work on two windows, and dust cleanup in the room, a cost of $311 was estimated for a
single-family unit.
$311 = $215 + (2 x $27) + $43
Combining cost estimates from HUD (1996) for interior paint repair ($500), window work on two
windows ($50), and area cleanup ($13 based on $0.05 per square foot for non-HEPA vacuuming from
Task Force (1995)), a second cost estimate of $563 was developed. The analysis assumes an average
room size of 250 fe for a single-family home.
4-14
~403 EA
-------
$563 =
$500 + (2 x $25) + ( $0.05 x 250 ft 2)
ft2
The average ofthese two values was $437, and this was employed as the estimate of the cost oflow-
intensity paint intervention in a single-family home. As indicated, the cost estimate for low-intensity
paint intervention included dust cleanup only in the area where work was completed.
Multifamily High-Intensity Cost. The estimate of intervention costs for high-intensity paint
intervention for multifamily housing was based on the single-family cost adjusted by a factor of 0.67 to
reflect the smaller size of multifa~ly units. This scaling resulted in an estimate of $4,413 for high-
intensity intervention in multifamily housing units.
$4,413 = 0.67 x $6,587
Again, the analysis estimates costs over a 50 year time frame; therefore, an estimate was required that
excluded any permanent measures undertaken the first time the house was abated. This was
accomplished by scaling the single-family repeat intervention costs (i.e., $4,744) by a factor of 0.67,
generating a value of $3,178. In cases where exterior interventions are triggered at the same time interior
paint interventions are occurring, the scaled cost of $3,178 is used in order to avoid the double counting
of window replacement costs incurred as a result of the exterior abatement.
Portions of the waste generated from abatements on multifamily housing units may be subject to
Resource Conservation and Recovery Act (RCRA) hazardous waste requirements. Under current
regulations, only those portions of the waste that fail the Toxicity Characteristic Leaching Procedure
(TCLP) for lead are considered hazardous waste. Disposal costs depend on the quantity being discarded.
In the HUD abatement demonstration project, which involved extensive abatement, 217 pounds of
hazardous waste, not including architectural debris, were generated per housing unit, and cost $274 to
discard (U.S. EPA, 1992). All high-intensity, interior multifamily abatements are assumed to incur this
incremental waste disposal cost. The cost estimates assume that architectural debris is not handled as
hazardous waste (Abt Associates, 1995b).
Multifamily Low-Intensity Cost. Since low-intensity paint intervention was limited to work in one
room, costs were not assumed to vary between single-family and multifamily housing. Therefore, the
single-family estimate of $437 was used for multifamily units.
This analysis assumes that all interior hazard control work (high and low-intensity) in multifamily homes
will be conducted on a unit-by-unit basis as children are born into the home.
Effectiveness. High-intensity interior paint interventions were assumed to have an effectiveness
equivalent to that of the dust intervention described earlier in terms of dust lead load. All paint
interventions reduce dust lead concentration as specified by Battelle (1997), and eliminate the source lead
that would otherwise be available to children exhibiting pica.
~403 EA
4-15
-------
Duration. The high-intensity paint intervention was assumed to have a duration of 20 years. Low-
intensity paint interventions and the accompanying dust cleanup of the room where the work occurred
were assumed to last for four years. In other words, to provide continued protection from exposure, low-
intensity paint interventions will need to become repeated once during the fIrst six years of a child's life.
Exterior Paint Intervention
Single-Family High-Intensity Cost. High-intensity exterior paint intervention involves full
encapsulation or removal of all exterior LBP from a housing unit. Cost estimates reported in the
literature varied according to the activities undertaken and the size of the area abated. HUD (1996)
estimated the cost of encapsulation or removal of exterior LBP and interior cleanup from a single-family
home of about 1,500 square feet and interior cleanup as $5,500. The NCLSH handbook (undated)
provided an estimate of $5,911 to $16,122 as the cost of a complete exterior paint job designed to fully
enclose LBP on the exterior of a three bedroom house of moderate size. The associated duration of
interventions was reported to range from 20 to 60 years (NCLSH, undated). The lower bound of the
NCLSH estimates was used since this analysis is interested in a 20 year duration. The average of the
HUD (1996) estimate ($5,500) and the lower bound of the NCLSH estimate ($5,911) was $5,706, which
was used for the single-family unit cost of high-intensity exterior paint intervention.
Single-Family Low-Intensity Cost. The low-intensity paint intervention cost was derived by averaging
two cost estimates. Low-intensity exterior paint intervention involved repair of all damaged exterior
LBP The fIrst estimate of $613 was reported in the NCLSH handbook for exterior paint repair plus
complete paint work up to a height of five feet (NCLSH, undated). The second cost of $1000 was
estimated in HUD (1996) and included exterior paint repair for one side of a single-family house of 1,500
square feet. The average of these estimates ($807) was the cost estimate used by this analysis for low-
intensity paint intervention in single-family homes. While this estimate may be low for homes needing
extensive paint repair, it was likely to be reasonable for homes needing a moderate level of work.
Multifamily Housing. Multifamily buildings are not likely to have extensive amounts of exterior LBP,
based on data from the BUD survey and the American Housing Survey (AHS). HUD (1990) indicated
that multifamily units account for 5.5 percent of the total amount of exterior LBP, even though these
units (greater than four units per structure) represented 17.3 percent of the total housing units in the 1993
AHS. The cost consultant for the National Center for Lead-Safe Housing indicated that most exterior
LBP on multifamily buildings would be present on windows and fIre escapes, since most multifamily
buildings, and particularly the larger ones, are masonry structures (Santucci, pers. comm.). For this
analysis, exterior paint intervention was considered to be repair or replacement of the windows in the
multifamily unit; information about the prevalence of fIre escapes in multifamily buildings was not
available. Using an estimate of seven windows per unit,9 and using the cost of stabilizing a window of
This estimate of seven windows was based on personal communication with Gopaul Ahluwalia at the
National Association of Home Builders, 1993. Mr. Ahluwalia estimated the average number of windows
in a new single-family home (17) and the average number per unit in a new multifamily apartment building
(9) based on a recent construction-material-usage data base. There were two trends in home building that
needed to be considered before using these estimates as the number of windows present in homes built
prior to 1980 (our population of interest for lead abatement). The first is that new homes are larger now
than in the past and second, homes are currently built with more windows to increase light in the house.
No quantitative information was available about the latter trend, but the former trend, was compensated for
by multiplying the 1993 average number of windows by the ratio of the average square feet per single-
4-16
~403 EA
-------
$26 (NCLSH, 1995), the cost for low-intensity exterior paint intervention for multifamily units was
$182. For high-intensity exterior paint abatement, the cost was estimated as the cost ofreplacing seven
windows. Santucci (pers. comm.) estimated a cost of $250 to $400 for replacement of a window
opening in a multifamily building. Using the midpoint of this range, the cost for the whole unit was
$2,275. For those units that need encapsulation or enclosure of exterior wans or work on fire escapes,
this estimate may be low. For other units, this cost may be high as not all windows may need
replacement. For example, NCLSH (1995) estimated the cost of encapsulation of both interior and
exterior window components at a cost of $43 per window, which was much cheaper than replacement.
All multi-family buildings are assumed to conduct exterior paint interventions in year one if required by
lead-hazard levels.
Effectiveness. All exterior paint interventions reduce dust lead concentrations as specified by Battelle
(1997) and eliminate the source oflead that would otherwise be available to children exhibiting pica.
Duration. Low-intensity exterior paint intervention measures are assumed to have a duration of four
years based on a report by Battelle (1997). A high-intensity exterior paint intervention is assumed to
have a duration of 20 years, again, based on Battelle (1997).
4.6.3 Soil Removal Costs
The costs of soil intervention vary with the size of the area treated, the method used, and whether or not
the waste is considered hazardous under RCRA. Residential soil intervention is a relatively new industry
and no standards have been established on what constitutes an effective intervention other than removal.
Single-Family Soil Removal Cost. This analysis defines a soil intervention as the removal of topsoil to
a depth of six inches, replacement with uncontaminated soil, raking and seeding, and disposal of lead
contaminated soil. Two areas of the yard are potentially subject to a soil removal: around the perimeter
of the unit, which is the area likely to be affected by the chipping of exterior LBP, and in remote areas
away from the foundation. Unit costs for soil removal and replacement are based on the Hometech
Remodeling and Renovation Cost Estimator (1996). Removal and replacement activities include: soil
removal using a small backhoe, backfilling using a small backhoe, replacement soil costs, and raking and
seeding of the yard. The Hometech Cost Estimator recommends an opportunity cost of $525 for use of a
backhoe. In addition, costs are included for the transport and disposal of lead contaminated soil. Landfill
disposal costs are based on Perket (1994) and are specific to non-hazardous waste. U.S. EPA provides
an estimate of $0.22 per ton-mile for the transport of bulk solids. Unit costs are presented in Exhibit 4.4.
family home in 1980 to the footage estimated for 1993 (1600 fe/2100 fe). This resulted in an estimate of
13 windows per average single-family home and 7 windows per unit for a multifamily dwelling built in
1980.
~403 EA
4-17
-------
Exhibit 4.4
Unit Cost Estimates for Soil Removal - Single-Family
Abatement Activity Unit Cost
Soil removal using a small backhoe* $6.08/yd3 (+$525)
Backfill using a small backhoe* $O.20/ft2
Replacement soil cost $40.00/yd3
Raking and seeding (by hand) $O.30/ft2
Landfill (non-hazardous) $35/ton
Soil Transportation $O.22/ton-mile
A small backhoe is defined to have a 1/2 cubic yard bucket.
In order to calculate the total cost of conducting a soil abatement, the unit cost estimates were combined
with estimates of the area likely to treated. As described above, two areas of the yard are potentially
subject to a soil removal -- perimeter areas and yard areas away from the foundation (i.e., remote areas).
The HUD data do not provide any information on the size of the yard or perimeter area; therefore, data
from the American Housing Survey (AHS) were used to estimate these values for an average single-
family home based on lot size, square feet within the house, and the number of floors in the house. To
summarize the methodology: I) the house's footprint (i.e., the amount of land covered by the building)
was estimated by dividing the square footage of the house by the number of floors in the home; 2) from
the footprint ofthe house, a perimeter was extrapolated (assuming a uniform shape for all homes); 3)
from the perimeter of the home, an estimate of the perimeter area extending three feet from the
foundation of the home was calculated; and 4) the remote area was based on yard size (calculated by
subtracting the foundation and footprint areas from the total size of the lot). The average area addressed
in the removal of soil from the perimeter of a single-family home is 417 ft2, and the average remote area is
2,571 ft2 For a detailed description of the methodology, refer to Appendix 4.A.
Exhibit 4.5 provides a summary of the total soil removal costs for abating the perimeter area, remote
area, and perimeter and remote areas together. Added to each of these costs is the cost of a single-family
interior dust cleanup ($391) described in Section 4.6.1. Soil transportation costs are calculated assuming
a distance of 100 miles to the landfill. Economies of scale are achieved if soil is removed from both the
perimeter and remote area due to the opportunity cost of using a backhoe as well as the cost of an interior
dust cleanup.
Single-Family Hazardous Waste Disposal of Soil. In cases of highly-contaminated soil, the cost of
disposal of soil as hazardous waste is added to the soil intervention cost. Soil is considered hazardous if
it fails the Toxicity Characteristic Leaching Procedure (TCLP) for lead. Many factors affect the leaching
characteristics of lead in soil, including the soil type and pH. While there is great variability in response
to the TCLP test, soil is unlikely to fail the test if concentrations of lead are less than 2000 ppm (Spittler,
pers. comm,). The analysis conservatively assumes that houses with a soil lead level greater than 2000
ppm will fail the test, and these houses will incur the additional soil disposal costs. The incremental costs
triggered by the handling of soil as hazardous waste are calculated based upon: 1) the average quantity of
soil likely to be removed during a soil intervention; and 2) the per ton price for bulk treatment and
disposal of waste at a RCRA Subtitle C facility. As discussed above, the analysis assumes an area of
417 ft2 and 2,571 ft2 for soil removal from the perimeter and remote area, respectively. Assuming that
4-18
~403 EA
-------
CQ)
~
o
c,,)
m
»
Exhibit 4.5
Soil Abatement Costs - Single-Family Home
Perimeter Total Perimeter
(area or soil Remote Total Perimeter Total Remote and Remote
Abatement Activity Unit Cost ($) quantity) (area or soil quantity) Cost ($) Area Cost ($) Area Cost ($)-
Soil removal using small 6.08/yd3 8 yd3 48 yd3 572 814 861
backhoe (+$525)
Backfill using small backhoe 0.20/fe 417fe 2,571 te 81 501 583
Replacement soil cost 40.00/yd3 8 yd3 48 yd3 309 1,904 2,213
Raking and seeding (by 0.30/fe 417 ft2 2,571 te 125 771 896
hand)
Landfill (non-hazardous) 35/ton 10 tons 62 tons 352 2,166 2,518
Soil transportation costs 22/ton 10 tons 62 tons 216 1,330 1,546
(100 miles to
landfill)
Interior dust cleanup 391/home NA NA 391 391 391
Total (non-hazardous) 2,046 7,878 9,008
- Represents certain economies of scale if soil is removed from both perimeter and remote area.
~
I
...
CD
-------
six inches of topsoil are removed, a volume of 8 yd3 is removed from the foundation and/or 48 yd3 is
removed from areas away from the foundation.
Perket (1994) estimated a cost of $174 per ton for bulk waste disposal, including treatment costs, in a
RCRA Subtitle C facility based on a survey of hazardous waste landfill prices. A portion of this cost is
deducted to calculate the incremental cost of handling and disposing of the soil as a hazardous waste.
The data used to estimate the cost of a soil abatement already include the cost of disposing of the soil as a
non-hazardous waste. To avoid double counting, the disposal cost of $174 per ton for hazardous waste
was reduced by $35 (i.e., the per ton cost for landfilling as non-hazardous waste), resulting in a cost of
$139 per ton.
Assuming 1.3 tons per cubic yard, results in the disposal of 10 tons of contaminated soil from the
perimeter of the home and/or 62 tons of contaminated soil from the remote area of the yard. Multiplying
these quantities by the incremental cost of disposal ($139), resulted in a total incremental cost of
hazardous soil disposal of $1,397 (perimeter only), $8,608 (remote area only), and $10,005 (both
perimeter and remote area). If soil is removed from both perimeter and remote areas, but only one of the
two areas exceeds the lead concentration of 2,000 ppm, hazardous waste disposal costs mayor may not
be incUITed. They will not be incurred if the average lead concentration of all soil removed is under 2,000
ppm, following soil mixing. If mixing could not reduce soil lead concentration beneath this threshold,
then it will not be performed, and hazardous waste disposal costs will be incurred only for the soil
fraction exceeding 2,000 ppm.
Multifamily Soil Removal Cost. Estimating soil removal costs for multifamily units required
knowledge of the average number of units in a multifamily building, the size of the area being abated, and
the unit costs for each of the soil removal activities (e.g., raking and seeding). Again, two areas of the
yard are potentially subject to soil removal: perimeter areas and yard areas away from the foundation. A
different approach was used to estimate these areas for multifamily homes because the AHS does not
report lot size for homes with two or more units. Santucci (pers. comm.) estimated a perimeter of 400 to
450 feet (or 1,275 fe assuming an area extending three feet from the foundation of the unit) for a
multifamily building of 40 units, or about 11 feet of perimeter per unit. Flaherty (pers. comm.) estimated
that yards for urban multifamily buildings in the Minneapolis area were likely to be up to twice as large
as single-family yards, suggesting a maximum remote area for the building of 5,142 if, or 171 ft2 per
unit, based on an average of 30 units per multifamily dwelling, as estimated from the 1993 AHS.
Total soil removal costs were calculated by combining these area estimates with unit costs from R.S.
Means' Repair and Remodeling Cost Data (1996) book. The R.S. Means Cost Data book is
recommended for residential, commercial, and industrial repair and remodeling projects costing between
$10,000 and $1 million, and was therefore used in estimating multifamily costs. As with single family
homes, removal and replacement activities include: soil removal using a small backhoe, backfilling using
a small backhoe, replacement soil costs, and raking and seeding of the yard. The R.S. Means Cost Data
book recommends an opportunity cost of $390 for use of a backhoe. 10 In addition, costs are included for
the transport and disposal of lead contaminated soil. Landfill disposal costs are based on Perket (1994)
10
The fixed cost portion of a soil removal is greater for single family jobs than for multifamily jobs.
Presumably, this reflects the fact that larger jobs provide a greater return on equipment use and thus the
opportunity cost of the backhoe is less.
4-20
i403 EA
-------
and are specific to non-hazardous waste. u.s. EPA provides an estimate of $0.22 per ton-mile for the
transport for bulk solids. Unit costs are presented in Exhibit 4.6.
Exhibit 4.6
Unit Cost Estimates for Soil Removal - Multifamily
Abatement Activity
Soil removal using a small backhoew
Unit Cost
Backfill using a small backhoe*
Replacement soil cost
$7.00/yd3 (+$390)
$0.46/ft2
Raking and seeding (by hand)
Landfill (non-hazardous)
$40.00/yd3
$0. 24/ft2
Soil Transportation
$35/ton
$0.22/ton-mile
A small backhoe is defined to have a 1/2 cubic yard bucket.
Exhibit 4.7 provides a summary of the total soil removal costs for abating the perimeter area, remote
area, and perimeter and remote areas together. Added to each of these costs is the cost of a mutlifamily
interior dust cleanup ($262 per unit) described in Section 4.6.1. Soil transportation costs are calculated
assuming a distance of 100 miles to the landfill. Economies of scale are achieved if soil is removed from
both the perimeter and remote area due to the opportunity cost of using a backhoe as well as the cost of
an interior dust cleanup. All multifamily buildings are assumed to conduct soil abatements in year one if
required by lead-hazard levels.
Multifamily Hazardous Waste Disposal of Soil. As discussed above, in certain cases, the cost of
disposal of soil as hazardous waste is added to the soil intervention cost. Soil is considered hazardous if
it fails the Toxicity Characteristic Leaching Procedure (TCLP) for lead. It is assumed that multifamily
buildings with a soil lead level greater than 2,000 ppm will fail the test, and will, therefore, incur the
additional soil disposal costs. Soil disposal costs are calculated based upon the average quantity of soil
removed and the per ton disposal cost estimated in the "Single-Family Hazardous Waste Disposal of
Soil" section above. Assuming that six inches of topsoil are removed, a volume of 24 yd3 is removed
from the foundation and/or 95 yd3 is removed from areas away from the foundation. The incremental
cost of disposing of soil as hazardous waste is estimated to be $4,269 (perimeter only), $17,217 (remote
area only), and $21,486 (both perimeter and remote area) for an average 30-unit building. The same
mixing principles apply as apply to single-family hazardous waste disposal of soil.
Effectiveness. Soil removal was assumed to reduce the soil lead level to 150 ppm in areas where soil
was removed, based on the average of the lead levels in the replacement soil in the Urban Soil Lead
Abatement Demonstration Project (Elias, 1993). Soil removal was also assumed to affect dust. The
reduction of dust lead concentration was variable, depending on other interventions performed (see
Battelle 1997). The dust cleaning that accompanies soil removal was assumed to reduce dust loads to 40
flg/ft 2.
~403 EA
4-21
-------
~
I
N
N
CD»
~
C
t.)
m
)0
Exhibit 4.7
Soil Abatement Costs - Multifamily Building of 30 Units
Perimeter Remote Total Remote Area Total Perimeter and
(area or soil (area or soil Total Perimeter Cost Remote
Abatement Activity Unit Cost ($) quantity) quantity) Cost ($) ($) Area Cost ($)*
Soil removal using small 7. 00/yd3 24 yd3 95 yd3 555 1.057 1.222
backhoe (+$390)
Backfill using small backhoe 0.46/ff 1.275 ft2 5.142 ft2 587 2.365 2.952
Replacement soil cost 40. 00/yd3 24 yd3 95 yd3 944 3,809 4,753
Raking and seeding (by 0.24/ff 1.275 ff 5,142 ff 305 1,228 1,533
hand)
Landfill (non-hazardous) 35/ton 31 tons 124 tons 1.074 4,333 5,407
Soil transportation costs 22/ton 31 tons 124 tons 660 2,661 3,321
(100 miles to
landfill)
Interior dust cleanup 7,852/Building NA NA 7,852 7,852 7.852
(30 units per
building)
Total (non-hazardous) .. . 11,977 23,304 27,039
. Represents certain economies of scale if soil is removed from both perimeter and remote area.
-------
Duration. Soil removal was assumed to be permanent since the topsoil containing lead was removed
based on Battelle (1997). The dust effects were also assumed to be permanent as the presumed source of
lead had been removed.
4.6.4 Overall Intervention Strategies
Assumptions regarding evaluation activities and triggered intervention work determined which unit costs
were associated with different housing populations. As stressed in the introduction, this analysis
followed the language of the ~403 regulations and assumed that individual housing owners respond to
hazards on a medium-specific basis. In cases where multiple hazard levels were exceeded, adjustments
were made to avoid multiple counting of costs. For example, the costs for high-intensity interior and
exterior paint intervention in multifamily homes both include costs for window replacement, so
adjustments were necessary in units performing both interventions.
4.6.5. Enforcement Costs
There are no enforcement costs associated with ~403 since it requires that the Agency set hazard
standards for lead in paint, soil and dust that will be used in other sections of Title X to trigger
interventions. The enforcement costs of these actions, however, are not attributable to ~403 but to the
section of the rule requiring the intervention. All intervention activity under ~403 is voluntary and thus
incurs no enforcement cost.
4.6.6 Implementation Costs
The implementation costs associated with ~403 are of two types. The first is the cost of setting and
promulgating the ~403 hazard levels themselves; a negligible cost compared to the funding appropriated
in Title X for intervention ($250 million in 1994). The second implementation cost is incurred by states
or localities that voluntarily use the hazard standards set by the Agency as action levels in their own lead
management programs. The size of these costs depends on the current level of activity at the state and
local level, whether the hazard standards that the Agency sets are above or below those of the programs
in place, and the number of programs that implement the hazard standards. If the Agency standards are
more stringent than current practice, implementation costs could be significant. However, if the Agency
standards are higher than those in practice, implementation costs will be negligible. No quantitative
evaluation of the implementation costs was attempted because reliable information on current and
expected future programs at the state and local level was not available. If implementation costs are
proportional to the number of homes affected, which could be the case if state or local authorities decided
to track homes to assure intervention, then the inclusion of implementation costs in the benefit-cost
analysis would favor higher hazard standards over lower ones, all other things being equal, since the
number of homes to be tracked would be lower under higher hazard levels.
4.7 Number of Interventions and the Number of Housing Units that Exceed the Candidate
Hazard Standards
The cost of any particular hazard standard is a function of the number of intervention actions of each type
that occur and the unit cost for each of these actions. The number of intervention actions, in turn, is a
function of the number of housing units that exceed the hazard standard under consideration. This
section estimates the number of interventions; chapter 6 presents the resulting estimated costs and
compares them to the monetary value of the benefits for various hazard standards.
~403 EA
4-23
-------
As described in chapter 3, the model estimates net benefits of every combination of hazard standards and
compares these estimates to determine which combination of standards maximizes net benefits. Due to
all media which exceed the standards at the time of the arrival of a child, net benefits are estimated in the
context of responding to each of the individual standards (paint, floor dust, windowsill dust and soil).
dust clean-up following the removal and replacement of soil. Where a soil intervention is performed,
therefore, a separate dust cleaning is not required regardless of the pre-intervention level of lead in the
would be overestimated, since it would include homes that are actually receiving their dust cleaning as
part of their soil interventions. Likewise, homes that exceed both the floor and the windowsill dust
properly assign costs and benefits to specific combinations of standards, the analysis estimates costs,
benefits and net benefits for combinations of hazard standards.
combinations of hazard standards. Since we are interested in illustrating the relationship between
changes in a specific standard and the number of homes that exceed that standard as well as a reasonable
other media are set at the option which EP A has chosen to propose. The proposed option, and the
number of interventions for that combination of hazard standards, are:
Interventions during the
Proposed Standard Model Duration
Paint Repair/Maintenance Interior: 10 sq ft or more, but less than 50
sq ft, of damaged paint
Exterior: 20 sq ft or more, but less than 100
sq ft, of damaged paint 7.0 million homes with
Paint Abatement Interior: 50 sq ft or more of damaged paint paint interventions
Exterior: 100 sq ft or more of damaged
paint
Dust Cleaning floor dust loading = 50 IJg1ft2, or windowsill 15.2 million homes with
dust loading = 250 IJg1ft2, or both dust interventions
Soil Removal and lead in soil = 2000 ppm 2.0 million homes with soil
Replacement interventions
4.7.1 Number of Homes Performing Interventions for Alternative Floor Dust Standards
As shown in Exhibit 4.8, the number of homes performing dust interventions (not including dust cleaning
associated with either paint or soil abatements), and the total number of homes performing any
intervention varies only slightly with variations in the floor dust standard. The standards (shown on the
horizontal axis) become less stringent as we move from left to right. The horizontal axis gives the
loading of lead in the floor dust at which interventions are called for, the higher this number, the more
homes that "pass" the test (i.e. the fewer homes that exceed the standard). The number of homes
performing soil abatements (about 2 million homes at the soil standard given above -- 2000 ppm) and the
~403 EA
-------
number performing any form of paint intervention (about 7 million homes at the paint standard given
above) do not change with changes in the floor dust standard. The total number of homes performing
some form of intervention is less than the sum of homes performing soil, dust or paint interventions
because some homes will perform multiple types of intervention activities.
In this exhibit, the floor dust standard varies between 40 Ilg/ft2 (the assumed post-intervention dust lead
loading) and 380 Ilg/ft2 (the highest pre-intervention loading in the data set). The number of homes
performing dust interventions is relatively insensitive to the stringency of the floor dust standards. The
number of homes that perform a dust cleaning, over the model period, declines slowly for standards
between 40 and 90 Ilg/ft2, dipping at a standard of 100 Ilg/ft2 , and declining slowly after that,
approaching but staying above the number of homes that perform some form of paint interventions.
The line which represents the total number of homes performing an intervention over the model period
traces a very similar path, but overall declines slightly less, indicating that some of the homes that no
longer do dust interventions, as the floor dust standards become less stringent, continue to do soil and/or
paint interventions. Because both paint and soil interventions are much more expensive than dust
cleaning, the costs will not decline as rapidly as the total number of homes performing an intervention.
At highly stringent floor dust standards (the left-hand end of the horizontal axis), homes with dust
cleaning comprise a much larger percentage of total homes with any intervention than is true at the other
end of the spectrum.
4.7.2 Number of Homes Performing Interventions for Alternative Windowsill Dust Standards
The next exhibit (Exhibit 4.9) presents the same type of information for variations in the windowsill dust
standard. The windowsill dust standards vary over a much broader range than the floor dust standards.
The minimum value (100 Ilglft2) equals the assumed post-intervention lead loading. The maximum pre-
intervention loadings in the data set are substantially higher than the 1000 Ilg/ft2 shown on the exhibit,
but the number of homes performing dust cleaning is nearly constant above the 1000 Ilg/ft2 standard.
Again, the number of homes performing soil interventions and the number performing paint interventions
do not change with changes in the windowsill dust standards.
The number of homes performing dust cleaning is very sensitive to the windowsill dust standards. The
number of homes performing dust interventions declines rapidly at the most stringent standards, levels
out slightly, and then declines even more at the 200-210 Ilglft2level. After that steep drop, the number of
homes declines more gradually with decreasing stringency of the windowsill dust standards.
As with the floor dust standards, the line representing the total number of homes performing an
intervention over the model period traces a very similar path, but overall declines slightly less than the
number of h9.mes performing dust cleanings. This indicates that some of the homes that no longer do
dust interventions, as the windowsill dust standards become less stringent, continue to do soil and/or
paint interventions. Again, costs should not decline as rapidly as number of homes, because the mix of
interventions shifts towards the more expensive ones as the stringency of the windowsill standards
decreases.
~403 EA
4-25
-------
Exhibit 4.8
Number of Homes Performing Interventions (Over Model Period)
o
OJ
E
~ 20
'5
o
c: 15
~
:E
''jC
...1...1
Other ctandards:
Soil: 2000 ppm
Sill duct: 250 IJglft2
Pairrt damage
irnerior:
r',.1aintenance: 10ft'
A,batemerrt: 50 fF
exterior:
Mainerrtance: 20 f!2
A,batemerrt: 100 ft'
-Total Affec1:ed
- Du::;1 Cleaning
- Pairn Irrtervn.
- Soil Remo',/al
400
Note: The "Duct
Cleaning" value::;: do
not include homes
with cleanings done
accompanying soil
remclval or irnerior
pairn abatement.
30
I I t I I
I I I I I I I
I I I I I I I
- - - - - .'. - - - - - -' - - - - - _, - - - - - - 1- - - - - - ,I. - - - - - _I - - - - - _I - - - - - -
I I I I I 1 I
I I I I I I I
I I I I I I I
I I I I I I r
-----~-----~------t------~-----~-----~-----~------
I I I I I I I
I I 1 I I I I
I I J I I I I
--..... I J I t I I
- - - - - .:' ~ - - - - -~- - - - - - ~- - - - - - T - - - - - -I - - - - - -I - - - - - -
I I J I I r I
I I I I 1 I I
- - - - ~-~----~~....~ - - - - - -: - - - - - - ~ - - - - - ~ - - - - - ~ - - - - - ~ - - - - - -
I 'j -+---
I I I I I I I
I I J I I t I
- - - - - T - - - - - , - - - - - -1- - - - - - r - - - - - T - - - - - , - - - - - -, - - - - - -
I I I I I I I
I I I I I I I
I I ] I I I I
- - - - - J. - - - - - _I - - - - - _1- - - - - - 1- - - - - - ). - - - - - J - - - - - J - - - - - -
1 I t I I J I
I I I I I
I
I
4.26
i403 EA
-OC
,L,_I
10
5
o
o
100
350
150
200
250
300
50
Floor Dust Standard (pgift')
-------
Exhibit 4.9
Number of Homes Performing Interventions (Over Model Period)
~ ~iJ
E
o
::c
'015
iii
c
~
~ 1iJ
3iJ
Other :;iandards:
Soil: 2000 ppm
Flo,:or dl..r:ot: 50 fJ~1/fP
Pairrt damage Interior:
t','1alntenance: 5 fp
,iI,batemerrt: 20 ft2
e::denor:
t','laintenance: 20 fp
,iI,batemerrt: 100 fl'
-Total,l1,ffecied
- DI.~:;:t Cleaning
- Paint Irrter',m
-Sc,il Remo',,"al
1 000
Note The "Dust
Cleaning" vall.~es do
not include homes
witt-, cleanings done
accompanving soil
removal or interior
paint abatement.
'iC
":"'_,
,
,
,
I
----~---
,'..,
I I
, \
I L__-----.. r
I".!
- - - - 1- - - - -,1 - -
I II I
, I I I I
I I I I I
I t l I I I
- - - - t- - - - - t- _--::"...:...=.r - - - - -1 - - - '1 '1 '1 - - "'t - -
I I r, I I I I I I
I I 1''--,.- I I I I I I
I I I'~-..I I I I I
I I I j'--I---~I I I
- - - - 1- - - - -'- - - - _1- - - - -' - - - - .1 - - - - .1 - _a - - -~ - - - - ., - - - -"..,J. - - - -
I I I I I r I I .
I I I I I I I I
I I I I I I I I
I I I I I I I I
I I I I I I I I
- - - - ,- - - - - ,- - - - -,- - - - -, - - - - -. - - - - -, - - - - -I - - - - -I' - - - -
I I I I I I I I
I I I I I I I
~403 EA
4-27
I I I t I I I
I I I 1 I I 1
I I I I J I J
I I I I I I I
- - - _I - - - - ...J - - - - .J - - - - J - - - - .J - - - - .i - - - - l. - - - -
I I I I I I I
I I I I I I I
t I I I I I I
I I I I I I I
I I I I I I
- - - -I - - - - -I - - - - -I - - - - -I - - - - -, - - - - "1- - - - -
I I I 1 I
I 1 1 1
, I ,
~,
I
I
I
I
'I~----
,
,
o
o
'100
200
300
400
500
E;OO
700
8iJO
900
Window Sill Dust Standard (pg,ft';
-------
4.7.3 Number of Homes Performing Interventions for Alternative Soil Standards
The story for variations in soil standards is a little more complicated. As above, the number of paint
interventions that occur remains constant over the variations in soil standards. There are, however,
apparent tradeoffs between soil and dust interventions. As explained at the beginning of this section, soil
abatements include a dust cleaning. Some homes that would perform a soil abatement under a stringent
standard, may not perform the soil intervention but would perform a dust cleaning alone under a less
stringent soil standard. Thus, homes might switch from the soil intervention to the dust intervention
category.
In Exhibit 4-10, soil standards range from 150 ppm (the assumed post-intervention soil lead
concentration) to slightly over 7000 ppm (the highest pre-intervention soil lead level in the data set).
Homes are highly concentrated at low soil lead levels, with the number of homes performing soil
interventions dropping by about one-half as soil standards decrease from 150 ppm to about 500 ppm.
The decline in the number of homes continues only slightly less rapidly to about 900 ppm; it declines
more gently to about 3000 ppm.
While the number of homes performing soil abatements is declining rapidly as the soil standard becomes
less stringent, the number of homes performing dust cleaning are increasing at a substantial rate. The
result is that the total number of homes performing some form of intervention falls as soil standards
become less stringent in the 150 - 1000 ppm range. Then the total number of homes performing
interventions levels off, as the majority of homes that no longer warrant a soil intervention now require a
dust cleaning under the floor dust standards of 50 Ilg/ft2 and windowsill dust standards of 250 Ilg/ft2. In
this case, costs should fall more rapidly than the number of homes performing an intervention because
the relatively expensive soil interventions are being replaced with less expensive dust cleaning.
4.8 Likely Rates of Intervention and Their Impact on the Cost Estimates
As described in Chapter 3, alternative hazard standard candidates are evaluated in terms of the net
benefits they would generate under a set of specific assumptions about the behavior of residential
property owners and managers. The results of this normative analysis are intended to inform decision-
makers about the relative merits of the alternative standards by providing a set of comparable estimates.
The analysis, however, does not provide estimates of the likely rates of intervention, and thus the likely
costs of interventions, under these standards.
The main objective of this section is to provide estimates of the likely costs that would result from the
establishment of the proposed hazard standards. Data limitations preclude defining a baseline that
accurately reflects future intervention activity levels in the absence of ~403 hazard standards. Likewise,
data are not available on which to base an estimate of the effectiveness of the hazard standards in
changing behavior, especially since the behavior changes will largely depend on the effectiveness of the
information programs that will accompany the standards. Therefore, the costs estimated in this section
represent an estimate of total post-regulation costs, not the incremental costs due to the ~403 standards.
4-28
~403 EA
-------
Exhibit 4.10
Number 01 Homes Performing Interventions (Over Model Period)
~ 20
E
o
:J:
'5 '15
rJj
c:
~
~ '1D
30
',C
L--'
I I I I I I
I I I I I I
I I I 1 I I
I I I I I I t
- - - .L - - - - - .J - - - - - .J - - - - - _1- - - - - - L.. - - - - - .I. - - - - - .J. - - - - -
I I I I J I I
I I I I J I I
I I I I I I
I I I j I
.)tf",er' standards:
Floor dte! 50 jJglft'
Sill diEt 250 jJg/tF
Paint damage
interior:
t','1airrtemmce: 10 tF
j!.,bffiemerrt: 50 11'
exterior:
Mairrtenance 20 ft'
,l!.,bffiemerrt, 1 00 tF
-Tc,\;~1 ,.!1,ife,::ted
- Dust Cleaning
- Paint Irrtef'oin,
-Soil R:emo","al
8000
Note: The "Dus1
Cleaning" value~: do
not include homes
'fo.o'ith cleanlr,gs done
accompanying soil
removal,Jr' irrterior
pairrt abffiemerrt,
~,
,
- - - - - "t - - - - - -. - - - -I - - - - - -.- - - - - - 1- - - - - - 't - - - - - -, - - - - -
I I I I I I I
I I I I I I I
I I I I t
I ~~'I I I I I
- - - - +- - - ::-.~ - - - - - ~ - - - - - -1- - - - - - t- - - - - - t - - - - - -t - - - - -
(-...---- I I I I I I
_.j : : : : : : :
r-' I I I I I t I
__L_____J_---_J_---_J______L_____L_____~-----
I I I I I I I
I I I I I I I
I I I I I I I
I I I I I I I
I I I I I I I
'------1-----~------I------~-----T-----1-----
I I I I 1 I
I I I I I I
I I I 1 I
, I ,
~403 EA
4-29
o
o
1000
2000
3000
4000
5000
6000
7000
Soil Standard (ppm)
-------
The best data currently available on intervention rates are data from the state of Massachusetts. For
nearly ten years, Massachusetts has required that all residential lead inspections and interventions be
reported to the state. This is the only state with data on all such activities for an extended period of time.
In many ways, the Massachusetts regulations are very similar to the proposed ~403 hazard standards. In
addition, Massachusetts vigorously enforces state regulations that require that landlords abate lead-based
paint. State programs that provide long-term, interest-free loans to low-income households to pay for
lead abatements, a tax on real estate transfers that supports the state Childhood Lead-Based Paint
Protection Program, mandatory testing of children's blood for lead, and active local public health
programs further promote the removal of lead-based paint. Given all this support, it is unlikely that
national intervention rates will exceed those seen in Massachusetts.
Thus, the costs of interventions were estimated assuming that the proposed standards are in effect, but
using an intervention rate consistent with that found in Massachusetts, as opposed to the birth rate. Since
the Massachusetts program contains several factors that promote interventions that may not be present in
the federal program, a second intervention rate was also used to estimate costs. This second rate was set
at one-half of the Massachusetts rate. Exhibit 4.11 presents the total costs (over 50 years discounted at 3
percent) and the total number of homes with interventions for the three alternative rates of intervention
activity: the rate in the normative model, a rate equivalent to the Massachusetts rate, and one-half the
Massachusetts rate. If the rate of interventions were equivalent to the average rate in Massachusetts, and
the mix of interventions were consistent with the proposed ~403 standards, the total costs would be
approximately 55 percent of the costs estimated by the normative "birth-trigger" model. It is likely that
the actual rate, and thus the actual total costs, will fall between the Massachusetts rate and one-half the
Massachusetts rate. As shown in Exhibit 4.11, this would mean that between 6.4 and 12.8 million homes
would experience an intervention during the 50-year period, or an average of 128,000 to 254,000 homes
a year. The present value of the 50-year costs would range from $15 to $29 billion.
Exhibit 4.11
Costs Under Alternative Assumptions About Intervention Rates: Proposed Standards.
Intervention Rate Present Value of Total Costs over Number of Homes
50 Years, Discounted at 3 Percent With Interventions
($billion) (million)
$52.8 20.6
$28.9 12.8
$14.7 6.4
Normative Model used in Chapter 6
Equivalent to Massachusetts Rate
Equivalent to One-Half of Massachusetts Rate
* Proposed Standards:
Interior paint: 10 sq ft or more - repair, 50 sq ft or more - abate
Exterior paint: 20 sq ft or more - repair, 100 sq ft or more - abate
Window sill dust: 250 sq ft2
Floor dust: 50 IJg/ft2
Soil: 2,000 ppm
4-30
~403 EA
-------
Appendix 4.A: Estimating Soil Removal Costs
The following discussion describes the methodology used to estimate soil intervention costs for single-
family homes. As described in the risk assessment, soil removal occurs when the average of the
foundation and remote soil-lead concentrations exceed the hazard standard. When the hazard standard is
exceeded, the intervention strategies are determined by: 1) the average of the pre-intervention soil-lead
concentration in samples taken at the dripline and entryway sampling areas (i.e., the foundation); and 2)
the pre-intervention soil-lead concentration in samples taken at the remote areas. If the average pre-
intervention soil concentrations exceed the candidate hazard standard, one of the following three
scenarios will result:
.
The pre-intervention soil lead concentration at the foundation of the house is greater than the hazard
standard, thereby triggering a soil removal intervention at the foundation of the house;
The pre-intervention soil lead concentration in yard areas away from the foundation of the house is
greater than the hazard standard, thereby triggering a soil removal intervention in yard areas away
from the foundation; or
Both the pre-intervention soil lead concentrations at the foundation of the house and yard areas away
from the foundation exceed the hazard standard, thereby triggering a soil removal in both areas.
.
.
To account for each of these possible scenarios, a unit cost methodology was developed to estimate the
total cost of abating the foundation and remote areas of the yard both together and individually. The
following discussion describes the methodology used to estimate the unit costs and the size of the areas
likely to be treated, respectively.
Unit Cost Estimates
Unit costs for soil abatement were estimated based on standard reference manuals used in construction
cost estimation: 1) Hometech Remodeling and Renovation Cost Estimator (1996); and 2) R.S. Means'
Repair and Remodeling Cost Data (1996). The R.S. Means' Repair and Remodeling Cost Data book
is recommended for residential, commercial, and industrial repair and remodeling projects costing
between $10,000 and $1 million, and was therefore used in estimating multi-family costs. The Hometech
Remodeling and Renovation Cost Estimator was used to generate costs for a single-family home, as
suggested by Robert Santucci, cost consultant to the National Center for Lead-Safe Housing (NCLSH).
Table A.l presents the unit cost values used in this analysis.
Table A.1: Unit Cost Estimates for Soil Abatements
Abatement Activity Single-family Home Multi-family Building
Unit Costs Unit Costs
Soil removal using small backhoe- $6.08/yd3 $7.00/yd3
(+$525) (+$390)
Backfill using small backhoe- $0.20/tr $0.46/tr
Replacement soil cost $40.00/yd3 $40.00/yd3
Raking and seeding (by hand) $0.30/tr $0.24/tr
Soil transportation costs $22/ton $22/ton
Landfill (non-hazardous) $35/ton $35/ton
Interior dust cleanup $391 $7,852
Landfill (hazardous) $139/ton $139/ton
* A small backhoe is defined to have a Y2 cubic yard bucket.
9403 EA
4-31
-------
The fixed cost portion of a soil removal, shown in parentheses in the first row of Table A.l, is greater for
single-family jobs than for multi-family jobs. Presumably, this reflects the fact that larger jobs provide a
greater return on equipment use and thus the opportunity cost of the backhoe is less. Alternatively, the
larger per cubic yard cost for multi-family jobs may compensate for the lower fixed cost. Also, note that
the cost of the one-time dust clean-up constitutes a larger percent of total costs for multi-family jobs than
for single-family jobs. This fact is reasonable given that the amount of soil removed from a multi-family
site is proportionally much less than the number of units at the same site.
Area Calculations
In order to calculate the total cost of conducting a soil abatement, the unit cost estimates described above
were combined with estimates of the area likely to treated. The following sections describe the
methodology used to estimate these areas for single-family homes.
Area Calculations - Single Family Homes
As described above, soil lead concentrations are measured for two areas of the yard -- perimeter areas and
yard areas away from the foundation (i.e., remote areas) -- and are therefore potentially subject to a soil
removal. Because the HUD data used in the analysis does not include data on the size of either the home
nor yard, average areas and average costs must be calculated for use in the analysis. Data were not
available to directly generate national level estimates of the average perimeter area and remote area for
single-family homes in the United States; however, data were available from the American Housing
Survey (AHS) to estimate these values based on lot size, square feet within the house or apartment, and
the number of floors in the house or apartment.l1 To briefly summarize the methodology: 1) the house's
footprint (i.e., the amount of land covered by the building) was estimated by dividing the square footage
of the house by the number of floors in the home; 2) from the footprint of the house, a perimeter was
extrapolated (assuming a uniform shape for all homes); 3) from the perimeter of the home, an estimate of
the perimeter area extending three feet from the foundation of the home was calculated; and 4) the
remote area was calculated by subtracting the foundation and footprint areas from the total size of the
lot. The following sections describe each of these steps in greater detail, with each section building upon
the values calculated in the previous section.
Median Lot Size Calculation
Median lot sizes were calculated for both single-family attached and single-family detached homes and
three geographical categories (to be used later in the analysis) using data from the 1995 American
Housing Survey (AHS). Buildings with 2-4 units were also included in this analysis to maintain
consistency with other parts of the analysis and HUD's definition of a single-family home: a residence
with 1 to 4 units. The AHS does not report lot sizes for 2-4 unit buildings; therefore, the mid-point of the
single-family attached and single-family detached lot size range was used in the absence of any
alternative data. According to the AHS, units in buildings with 2-4 units make up only thirteen percent
of all units in buildings with 1-4 units. Median lot sizes are summarized in Table A.2 with response rates
II
The AHS, conducted by the Census Bureau, collects detailed data on the Nation's housing stock using a
sample of roughly 55,000 homes. Weights are assigned to each record in order to extrapolate values to the
Nation.
4-32
~403 EA
-------
indicated in parentheses. Response rates are not applicable to buildings containing 2-4 units because
these data were not available from the AHS.
Table A.2: Median Lot Sizes (sq. ft.) from the American Housing Survey
Central City Suburb Non-Metro
2-4 unit 2-4 unit 2-4 unit
S.F. Detached S.F. Attached building S.F. Detached S.F. Attached building S.F. Detached S.F. Attached building
8,400 3,000 5,700 15,000 5.500 10,250 40,000 11,000 25,500
(63 percent) (21 percent) (not applicable) (76 percent) (22 percent) (not applicable) (70 percent) (25 percent) (not applicable)
Geographic categories are defined as follows based on the categorization used in the American Housing Survey -- Central City Central
City; Suburb: Urbanized Suburb, Other Urban Suburb, Rural Suburb; Non-metro: Urbanized Non-Metro, Other Urban Non-Metro, Rural
Non-Metro.
Median Footprint Calculation
Median footprint values were calculated for both single-family attached and single-family detached
homes using data from the 1995 and 1985 AHS. Median footprint values were then subtracted from the
median lot size calculations (described above) to generate estimates of yard size.!2 As previously
discussed, buildings with 2-4 units were also included in this analysis. However, a different methodology
was used to estimate the median footprint of a building with 2-4 units.
The AHS does not report the footprint of a home; therefore, the value was derived for single-family
attached and single-family detached homes by dividing square footage values (extracted from the 1995
AHS) by the number of floors in the home (extracted from the 1985 or 1995 AHS). This methodology
requires the assumption that the square footage of a multi-story, single-family home is uniformly
distributed across all stories. Fifty percent of the single-family homes reporting floors data contain only
one story and are therefore unlikely to be affected by this assumption.
Data on the number of floors were extracted from both the 1985 and 1995 surveys because the Census
Bureau returns to the same housing units each year the survey is conducted and since 1985 has not
collected floors data when units had been visited during a previous survey year. Therefore limited floor
data were available from the 1995 survey alone, and the 1985 floor values were merged with the 1995
records using a unique ID number assigned to all housing units sampled by the AHS. All usable records
-- records with square footage and number of floors n were combined to fonn one data set.
A different methodology was used to calculate the footprint of a 2-4 unit building because the AHS
reports square footage values on a unit basis and does not report the square footage of an entire multi-
family building. The analysis assumes that all 2-4 unit buildings are configured vertically with equal
sized units one above the other (similar to a Boston triple-decker).!3 By making this assumption, the
square footage of an individual unit could be used to represent the entire footprint of the building. No
data were available from the AHS to test the strength of this assumption. Median square footage values
12
This methodology does not account for the presence of garages, porches, and paved areas that would not
be subject to a lead abatement.
13
This is equivalent to assuming that the number of floors in the building equaled the number of units, and
that each floor of a unit was the same size. Examples are side-by-side, two-story duplexes.
~403 EA
4-33
-------
were used to estimate the footprint for 2-4 unit buildings in each of the geographic areas considered in
this analysis (central city, suburb, and non-metro). Median footprint values are presented in Table A.3.
Table A.3: Median Footprint Sizes (sq. ft.) Derived from the American Housing Survey
Central City Suburb Non-Metro
2-4 unit 2-4 unit 2-4 unit
S.F. Detached S.P. Attached building S.P. Detached S.P. Attached building S.P. Detached S.F. Attached building
1,135 667 900 1,150 700 946 1,090 671 840
(83 percent) (46 percent) (46 percent) (81 percent) (60 percent) (57 percent) (86 percent) (65 percent) (59 percent)
Geographic categories are defined as follows based on the categorization used in the American Housing Survey -- Central City
Central City; Suburb: Urbanized Suburb, Other Urban Suburb, Rural Suburb; Non-metro: Urbanized Non-Metro, Other Urban Non-
Metro, Rural Non-Metro.
Perimeter Area Calculation
Calculating the perimeter area required three pieces of information: 1) the perimeter of the home, 2) the
configuration or shape of the home, and 3) the distance from the foundation that would likely be subject
to a soil removal. Santucci estimated that an abatement action would likely involve an area extending
three feet from the foundation of the home. This value, coupled with the median footprint sizes estimated
above, allowed us to extrapolate a measure of the home's perimeter, assuming a rectangular home with a
front to side ratio of 2:3. For example, to calculate the perimeter of a central city, single-family detached
home with a footprint of 1,135 square feet, the following formula was used to calculate the length ofthe
home's front and side:
FRONT
FOOTPRINT
1.5
where FOOTPRINT
1,135 sq. ft.;
SIDE =
FOOTPRINT
0.666
where FOOTPRINT = 1,135 sq. ft.
The perimeter areas calculated for single-family detached, single-family attached, and 2-4 unit buildings
are presented in Table AA. Based on Santucci, single-family attached homes are assumed to only have
soil present in the front or back of the home; therefore, perimeter areas were calculated based upon the
FRONT length only. Perimeter areas for 2-4 unit buildings were converted to per unit values assuming
three units per building.
Table A.4: Perimeter Area Sizes (sq. ft.) Derived from the American Housing Survey
Central City Suburb Non-Metro
2-4 unit 2-4 unit 2-4 unit
S.P. Detached S.P. Attached building S.F. Detached S.P. Attached building S.P. Detached S.F. Attached building
496 63 147 499 65 151 486 63 142
Geographic categories are defined as follows based on the categorization used in the American Housing Survey -- Central City
Central City; Suburb: Urbanized Suburb, Other Urban Suburb, Rural Suburb; Non-metro: Urbanized Non-Metro, Other Urban Non-
Metro, Rural Non-Metro.
4-34
~403 EA
-------
Remote Area Calculation
The remote area is defined by this analysis as the yard area away from the foundation of the home.
Estimates of remote areas for single-family attached, single-family detached, and 2-4 unit homes were
calculated by subtracting the median footprint values and perimeter areas from the median lot size values
for each of the geographic categories (central city, suburb, and non-metro). Calculations of perimeter
areas, footprint areas, and lot sizes are described above. Remote areas for 2-4 unit buildings were
converted to per unit values assuming three units per building. Remote area values for an single-family
building types are presented in Table A.5.
Table A.5: Remote Area Sizes (sq. ft.) Derived from the American Housing Survey
Central City Suburb Non-Metro
S.P. S.P. 2-4 unit S.P. S.F. 2-4 unit S.P. S.p. 2-4 unit
Detached Attached building Detached Attached building Detached Attached building
6,769 2,270 1,551 13,351 4,735 3,051 38,424 10,265 8,173
Geographic categories are defined as follows based on the categorization used in the American Housing Survey -- Central City
Central City; Suburb: Urbanized Suburb, Other Urban Suburb, Rural Suburb; Non-Metro: Urbanized Non-Metro, Other Urban Non-
Metro, Rural Non-Metro.
Bottom Line Remote Area and Perimeter Area Calculation - Single Family Home
A weighted average of the perimeter area and remote area was calculated for each of the three geographic
categories based on the prevalence of single-family attached and single-family detached homes as wen as
the prevalence of units in 2-4 unit buildings. Average perimeter areas and remote areas are summarized
in Table A.6. As the perimeter areas varied little across the three geographic locations, a straight average
of the three values was used in the final cost calculation: 417 square feet. Remote area values varied
significantly across the three geographic categories; therefore, the central city value was used assuming a
higher incidence of lead contaminated soil in central city areas. One-half the central city value (i.e., 2,571
square feet) was used in the final cost calculation because, where lots are very large, it is likely that soil
would be removed from only a portion of the entire yard. In addition, paved areas, which were not
accounted for in the area estimates, would not be subject to soil removal.
Table A.6: Weighted Average Perimeter Areas and Remote Areas By Geographic Location
Weighted Average Area (sq. ft.)
Abatement Area
Central City Suburb Non-Metro
Perimeter 373 432 447
Remote 5,142 11,700 35,220
Geographic categories are defined as follows based on the categorization used in the American Housing Survey -- Central City
Central City; Suburb: Urbanized Suburb, Other Urban Suburb, Rural Suburb; Non-Metro: Urbanized Non-Metro, Other Urban
Non-Metro, Rural Non-Metro.
~403 EA
4-35
-------
References
Abt Associates. 1995a. TSCA Title IV, Sections 402(a) and 404: Target Housing and Child-Occupied
Facilities Final Rule Regulatory Impact Analysis. Prepared for U.S. EP A, Office of Pollution
Prevention and Toxics, Regulatory Impacts Branch. September.
Abt Associates. 1995b. Exemption of Lead- Based Paint Architectural Debris from Subtitle C of RCRA,
Analysis of Costs, Benefits, and Risks. Prepared for Office of Solid Waste, U.S. EPA. September
11.
Battelle. 1997. Risk Assessment to Support Standards for Lead in Paint, Dust, and Soil. Prepared by
Battelle, for National Program Chemicals Division, Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency. EPA 747-R-97-006, December.
Elias, R. 1993. Personal communication between Rob Elias, U.S. Environmental Protection
Agency, Office of Research and Development and Alice Tome, Abt Associates Inc. March 16.
Flaherty, P. 1995, 1996. Personal communication between Pat Flaherty, Specialty Construction, Fridley,
MN and Chris Paciorek, Abt Associates Inc. October, 1995 and July, 1996.
Hometech. 1996. Remodeling and Renovation Cost Estimator.
Lead-Based Paint Hazard Reduction and Financing Task Force (Task Force). 1995. Putting the Pieces
Together: Controlling Lead Hazards in the Nation's Housing. HUD-1547-LBP. July.
Lim, B. 1996. Lead paint abatement and repair and maintenance study in Baltimore. Materials from the
Quality Assurance Project Plan sent by Benjamin Lim of U.S. EP A, Office of Pollution Prevention
and Toxics (OPPT), Chemical Management Division to Nishkam Agarwal, U.S. EPA, OPPT,
Regulatory Impacts Branch. February.
Means, R.S. 1996. Repair and Remodeling Cost Data.
National Center for Lead-Safe Housing (NCLSH). 1995. SpecMaster Database. Provided to Chris
Paciorek, Abt Associates by Jonathon Wilson, NCLSH. December.
National Center for Lead-Safe Housing (NCLSH). Undated. Creating a Lead-Based Paint Hazard
Control Policy: A Practical, Step-by-Step Approach for Nonprofit Housing Organizations.
Technical Assistance Bulletin 2.
Office of the President. 1996. Economic Report of the President.
Perket, C. 1994. Hazardous Waste Landfill Prices, 1989-1994. EI Digest, November, pp. 33-36.
Santucci, R. 1996. Personal communication between Robert Santucci, cost consultant to the National
Center for Lead-Safe Housing and president of Urban Restoration Corporation, North Carolina, and
Chris Paciorek, Abt Associates Inc. July.
4-36
~403 EA
-------
Spittler, T. 1993. Personal communication between Tom Spittler, U.S. EP A Region I, and Alice Tome,
Abt Associates Inc.
U.S. Department of Commerce (DOC) and U.S. Department of Housing and Urban Development
(HUD). 1993. American Housing Survey for the United States in 1993. U.S. Department of
Commerce, Economics and Statistics Administration and Bureau of the Census and U.S. Department
of Housing and Urban Development, Office of Policy Development and Research. Current Housing
Reports H150/93. Issued February 1995.
U.S. Department of Commerce (DOC) and U.S. Department of Housing and Urban Development
(HUD). 1995. American Housing Survey for the United States in 1995. U.S. Department of
Commerce, Economics and Statistics Administration and Bureau of the Census and U.S. Department
of Housing and Urban Development, Office of Policy Development and Research. Issued 1997.
U.S. Department of Housing and Urban Development (HUD). 1990. Comprehensive and Workable Plan
for the Abatement of Lead-Based Paint in Privately Owned Housing: Report to Congress.
Washington, DC. December.
U.S. Department of Housing and Urban Development (BUD). 1993. National Survey of Lead-Based
Paint in Housing: Documentation of Analytical Data Files. Prepared for U.S. Department of
Housing and Urban Development by Westat, Inc.. November 30.
U.S. Department of Housing and Urban Development (HUD). 1995. Guidelines for the Evaluation and
Control of Lead-Based Paint Hazards in Housing. Office of Lead-Based Paint Abatement and
Poisoning Prevention. June. Available from BUD USER (800-245-2691).
U.S. Department of Housing and Urban Development (BUD). 1996. Regulatory Impact Analysis of the
Proposed Rule on Lead-Based Paint: Requirements for Notification, Evaluation and Reduction of
Lead-Based Paint Hazards in Federally-Owned Residential Property and Housing Receiving Federal
Assistance. Prepared by ICF for Office of Lead-Based Paint Abatement and Poisoning Prevention,
U.S. Department of Housing and Urban Development.
U.S. Environmental Protection Agency (EPA). 1992. Applicability ofRCRA Disposal Requirements to
Lead-based Paint Abatement Wastes. Office of Pollution Prevention and Toxics. June.
U.S. Environmental Protection Agency (EPA). 1994. Estimating Costs for the Economic Benefits of
RCRA Noncompliance. Office of Regulatory Enforcement. Obtained by Abt Associates from
DPRA" Incorporated, St. Paul, MN.
U.S. Environmental Protection Agency (EPA). 1995a. Guidance on Identification of Lead-Based Paint
Hazards. Federal Register 60 (175): 47248-47256, September 11.
U.S. Environmental Protection Agency (EPA). 1995b. Lead-Based Paint Risk Assessment Model
Curriculum. Prepared for the Chemical Management Division by The National Center for Lead-Safe
Housing. June. A VI-0006700-1 and A VI-0006700-2.
~403 EA
4-37
-------
Westat, Inc. 1995. Report on the National Survey of Lead-Based Paint in Housing. Prepared for U.S.
EPA. March 17.
4-38
~403 EA
-------
5. Benefits
5.1
Introduction
Chapter 3 provided a description of the analytic approach used to conduct this benefit-cost analysis, and
in particular the linkage between the economic analysis and the risk assessment performed in support of
the ~403 standards (Battelle, 1997). For a thorough discussion of the health hazards associated with lead
focusing on those effects that are addressed in the benefits estimates provided here, the reader is referred
to Chapter 2 of the Battelle risk assessment document. The reader is also referred to Chapters 3 and 4 of
the risk assessment document for further details on the exposure modeling and dose-response modeling
incorporated into the estimation of the benefits of ~403 standards discussed both in this Chapter and
subsequently in Chapter 6
As described in Chapter 3, the benefit-cost analysis essentially compares alternative futures over a 50
year time horizon: a baseline or "no-action" alternative for which it is assumed that no changes are made
to current ambient lead exposure conditions, and a "post-action" alternative for which it is assumed that
the ambient lead exposure conditions are reduced in specific ways in response to the promulgation of
~403 standards.
The benefits of implementing the ~403 standards can be expressed in several ways. These include
estimates of:
.
The reduction in environmental lead levels (i.e., in paint, soil and dust) to which children are
exposed;
.
The number of children who experience lower exposure levels than they would have in the
baseline;
.
The reduction in blood lead levels in children resulting from lowered exposure levels;
.
The reduction in the incidence of specific adverse effects or consequences associated with
elevated blood lead levels in children (including IQ changes, medical interventions, remedial or
compensatory education); and
.
The monetary value associated with the reduction of the incidence of those adverse effects.
The remainder ofthis chapter presents these ~403 benefits in two sections. Section 5.2 addresses the
benefits as noted in the first four bullets above, which deal with reductions in environmental lead levels
and changes in the incidence of adverse health consequences associated with the exposure reductions.
Section 5.3 addresses the valuation or monetization of these benefits to allow for a comparison with the
costs of the ~403 standards.
~403 EA
5-1
-------
Note that within this chapter, the benefits are presented only for the specific ~403 standards that are
being proposed, namely:
Floor dust: 50 J.lg/ft2
Windowsill dust: 250 J.lg/ft2
Soil: 2000 ppm
Paint: Interior lead paint> I 0 ft2 of damage; Exterior lead paint >20 fe of damage
In Chapter 6, where benefits are compared directly with costs, the monetized value of the benefits are
presented for a range of ~403 alternatives in addition to those specifically being proposed.
5.2 Benefits as Reduced Exposure and Adverse Health Consequences
As noted above, this section provides estimates of the benefits of the proposed ~403 standards in terms
of reduced exposure to children and the associated reduction in adverse health consequences. In all cases,
these are reductions measured against a baseline of no changes in the ambient lead exposure conditions.
In other words, this is a marginal analysis with a baseline of no intervention.
As described in Chapter 3, the choice of the intervention strategy (i.e. the specific actions to be taken in a
housing unit to reduce lead levels) is a function of the particular lead hazard standard under analysis and
the lead conditions of that housing unit. Also, the timing of interventions is a function of the timing of
childbirths. In homes where hazard standards are exceeded, it is assumed that the interventions will be
carried out just prior to the arrival of a newborn child. Furthermore, they are repeated when necessary as
long as any child under age 6 is still present in the home.
Six intervention actions are considered in this analysis. These include two interior paint interventions
(high-intensity and low-intensity interior paint interventions); two exterior paint interventions (high-
intensity and low-intensity paint interventions); one soil intervention; and one dust intervention.
Depending on the conditions of the housing unit, various intervention elements were combined to form an
intervention strategy. The effectiveness and duration levels associated with interventions determine the
post-intervention conditions. These levels were discussed in Chapter 3 and at greater length in Battelle
(1997). The assumed post-intervention conditions for each alternative are summarized in Exhibit 5-1.
Chapter 4 provides estimates of the number of homes over the course of 50 years where these
interventions would take place in response to both the anticipated presence of a new child and the
existence of environmental lead levels exceeding alternative standards.
In the benefit-c~st model it is estimated that over the course of the 50 year period, approximately 173
million children will be born and occupy the approximately 98 million homes estimated to comprise the
current US housing stock. Of these, it is estimated that 131 million children will occupy target housing
stock built prior to 1979. Assuming interventions are performed as set forth in the model, approximately
43.8 million children, 33.4% of the 131 million children occupying target housing stock, will experience
a reduction in exposure to environmental lead levels from paint, soil, and/or dust as a result of the ~403
standards. Also, it is estimated that in the baseline, approximately 2.36 million children will experience
elevated blood lead levels due to direct ingestion of paint chips over the 50 year period; with the
5-2
t403 EA
-------
proposed 9403 standards it is estimated that only 1.09 million children will experience elevated blood
lead due to direct ingestion of paint chips.
Exhibit 5-1
Summary of Post-Intervention Conditions for Various Intervention Alternatives
(LBP refers to lead-based paint)
Abatement Alternative
Assumed Post-Intervention Conditions
High-Intensity Interior Paint
Low-Intensity Interior Paint
High-Intensity Exterior Paint
Low-Intensity Exterior Paint
Dust
. Paint: Deteriorated interior LBP made inaccessible for 20 years
. Dust: Floor dust lead loading level reduced to 40 ~g/fe
Window sill dust lead loading level reduced to 1 00 ~g/ft2
Lead concentration level reduced to 20% of pre-intervention
level for 20 years
. Paint: Deteriorated interior LBP made inaccessible for 4 years
. Dust: Lead concentration level reduced to 20% of pre-intervention
level for 4 years
. Paint: Deteriorated exterior LBP made inaccessible for 20 years
. Dust: Lead concentration level reduced to 20% of pre-intervention
level for 20 years
. Paint: Deteriorated exterior LBP made inaccessible for 4 years
. Dust: Lead concentration level reduced to 20% of pre-intervention
level for 4 years
. Dust: Floor dust lead loading level reduced to 40 ~g1ft2
Window sill dust lead loading level reduced to 1 00 ~g/ft2
Duration is 4 years in both cases.
Effect on dust lead concentration depends on other
interventions implemented (see Battelle 1997)
. Soil: Soil lead concentration permanently reduced to 150 ppm
areas where soil is removed
. Dust: Floor dust lead loading level reduced to 40 ~g/ft2.
Window sill dust lead loading level reduced to 1 00 ~g/ft2
Duration is permanent in both cases.
Effect on dust lead concentration depends on other
interventions implemented (see Battelle 1997)
Note: Lead levels remain constant in any case where starting levels are lower than assumed post-intervention ones.
Soil Removal
In the baseline analysis, it is assumed that the blood lead levels of children would continue to reflect
levels observed in the NHANES III, Phase 2 analysis characterized as a lognormal distribution with a
geometric mean (GM) of 3.14 f.lg/dl, a geometric standard deviation (GSD) of 2.09, and a resulting
expected value (arithmetic mean) of 4.12 f.lg/dl. As discussed previously, EPA has employed two blood
lead models (the IEUBK and the Empirical models) in the risk assessment to predict the effects of the
reduction in environmental exposure levels. Using the IEUBK model, it is estimated that the blood lead
GM will be 2.72 f.lg/dl with a GSD of 1.90 (arithmetic mean of 3.34 f.lg/dl). Using the Empirical model,
it is estimated that the blood lead GM will be 3.03 f.lg/dl with a GSD of 2.05 (arithmetic mean of 3.91
f.lg/dl).
~403 EA
5-3
-------
For the overall population of 173 million children for which these blood lead distributions apply, this is
an average reduction of 0.78 /lg/dl with the IEUBK model and 0.21 /lg/dl for the empirical model.
However, when it is recognized that these blood lead reductions are expected to occur only in the 43.8
million children noted above as being in target housing affected by these standards, it is estimated that
the average blood lead reductions among those children affected are 3.09 /lg/dl from the IEUBK model
and 0.83 /lg/dl from the Empirical model.
As noted in Chapter 2, the selected health end-points used in the assessment of the benefits of these
standards include several specific blood lead levels identified by the CDC (1991) as critical values above
which various levels of follow-up monitoring and/or specific medical interventions should be undertaken.
Key among these are blood lead levels of 10 /lg/dl and 20 /lg/dl. It is estimated that for the baseline
analysis approximately 10 million (of the 173 million children) born into current housing stock will have
blood lead levels above 10 /lg/dl, and 1 million will have blood lead levels above 20 Ilg/dl. With the
9403 standards, these numbers are reduced to 3.6 and 8.2 million exceeding 10 /lg/dl (for the IEUBK and
Empirical models, respectively) and 0.2 and 0.7 million exceeding 20 /lg/dl (again for the IEUBK and
Empirical models, respectively).
Critical components of the estimated benefits of reduced blood lead levels in children are the potential
improvement in IQ scores in general and the associated reduction in the incidence of low IQ scores «70)
in particular. As a result of the estimated reduction in average blood lead levels (coupled with the
relationship between blood lead and IQ scores discussed in Chapters 2 and 3), it is estimated that the
average improvement in IQ score among the 43.8 million children affected by the proposed standards
over the 50 year period would be 3.1 points based on the IEUBK model and 0.8 points based on the
Empirical model.
It is also estimated that over the 50 year modeling period, the number of children avoiding IQ scores
below 70 resulting from the 9403 standards ranges from approximately 26,000 children (from the
IEUBK model) to 9,000 children (from the Empirical model).
It is important to note that these are not the only benefits anticipated to result from the ~403 standards,
but rather are those considered the key benefits that are most directly measurable in this analysis. The
reader is referred to the Hazard Identification discussion in Chapter 2 of the risk assessment document
(Battelle, 1997) for a more thorough discussion of the potential benefits of reducing environmental lead
levels.
The following section of this chapter describes the approach to placing a monetary value on the benefits
of reduced blood lead levels in children, including those associated with medical interventions, education,
and IQ point changes.
5.3
Valuation of Benefits
This section describes the approach used to place a monetary value on the expected benefits. The
benefits are assigned monetary values to facilitate comparison with the costs of conducting interventions.
The approach used by this analysis defines benefits as avoided health damages and avoided elevated
blood lead levels. The main health effects considered are reductions in IQ and cognitive effects from lead
exposure. Available economic research provides little empirical data for society's willingness to pay to
5-4
~403 EA
-------
avoid decreases in IQ or adverse cognitive effects. To represent some portion of society's full willingness
to pay, alternative measures were calculated that considered three consequences of lead exposure for
children: decreased expected lifetime earnings, increased educational resources expended, and costs of
increased medical intervention. Foregone earnings are examined in Section 5.4.1. Increased educational
expenditures are addressed in Section 5.4.2. Costs due to increased medical intervention are presented in
Section 5.4.3. All estimates are presented in 1995 dollars. Exhibit 5-2 summarizes the components used
to provide values for the health effects considered by the benefits analysis.
5.3.1 Valuing Changes in IQ Points
The valuation of changes in IQ was completed in two steps. First, an estimate of the present value of the
earnings stream of an average newborn was calculated. Second, available economic literature was used
to estimate the percentage increase in lifetime earnings one would expect from a one point increase in IQ.
Average Earnings. To calculate the present value of the earnings stream of an average newborn, it was
assumed that at any given age the child will receive annual earnings in real terms equal to average
earnings "currently" received by persons of the same age. Data from the 1992 Current Population
Survey (CPS) were inflation adjusted to 1995 dollars to represent 1997 figures in this analysis. The
projected annual earnings stream was adjusted to take three factors into account. First, some real
increases in earnings were assumed to occur through general increases in productivity. Second, projected
earnings were lowered to take into account probabilities of survival. Finally, the lifetime earnings stream
was discounted to express the stream in present value terms.
A verage earnings calculations were performed for ages ranging from 18 to 64 using 1992 CPS data on
the average annual earnings, total persons, and the number of persons with earnings by gender, age, and
education group (US Department of Commerce, 1993). Average earnings were calculated for those in a
particular age group as a weighted average of average earnings in each gender and education sub-group.
The weights used were the fractions of the age group represented by each gender and education group.
Average annual earnings for those with earnings, total number of persons with earnings, and total number
of persons were typically reported by gender for various age groups (e.g., 18-24; 25-34; 35-44; 45-54;
55-64; and 18-24; 25-29; 30-34; 35-39; 40-44; 45-49; 50-54; 55-59; 60-64) and education groups (less
than 9th grade, 9-12 grade with no diploma, high school graduate, some college, associate degree,
bachelor's degree, master's degree, professional degree, and doctorate). Employment rates were
estimated based on estimates of total persons and of total persons with earnings. Estimates of zero
earnings for the unemployed were incorporated into the calculation of average earnings.
Several assumptions had to be made to calculate the average earnings stream because of limited data:
.
First, the CPS data only includes some information for those with professional degrees and
doctorates. In instances where numbers were not reported, earnings and employment rates were
inferred by comparing information on those with at least a BA to information on those with a BA
alone and those with an MA alone.
~403 EA
5-5
-------
C11
I
en
Exhibit 5-2
Summary of Benefits Analysis Estimate
Type of Effect
Source
Effect of a Single
Point Reduction in 10
Cost of Additional
Education
Total Effect of a
Single Point
Reduction in 10
Description
Sum of the direct and indirect effects on the
percent of earnings lost (2.39 percent) and
express the effect in terms of the present
value of average lifetime earnings
Sum of the direct costs ($316) and
opportunity costs ($627) of additional
education
Subtract the costs of additional education
from the effects on earnings lost
Special Education (10 Cost of special education beginning at age
less than 70 points) 7 and ending at age 18
Compensatory
Education (Blood-
lead greater than 20)
Medical
Intervention(for
several blood lead
ranges)
Cost of compensatory education beginning
at age 7 and ending at age 9
Cost of blood lead screening and medical
intervention for children less than six years
old (by blood lead risk group)
Estimate
$9,360 dollars
$1,114 in 1995 dollars
$8,346 in 1995 dollars
$53,836 in 1995 dollars
$15,298 in 1995 dollars
(All in $1995)
Risk Group 1:$58
R.G. IIA: $70
R.G. liB: $227
R.G. iliA: $417
R.G. IIIB: $678
R.G. IV: $9843
R.G. V: $9843
Product of the estimate of the present value of average lifetime earnings based on US
Department of Commerce (1993) ($359,391) and the assumed percentage loss of earnings
from a single point reduction in 10 of 2.39 percent (Salkever 1995)
Sum of the estimate of the direct and opportunity costs of additional education based on US
Department of Education (1993) data
Accounting for the cost of additional education was based on Salkever (1995)
Kakalik et al. (1981) estimate annual incremental regular classroom costs of $6,458 in 1995
dollars for special education. This estimate is the discounted value of such costs for age 7
through 18.
Kakalik et al. (1981) estimate annual incremental regular classroom costs of $6,458 in 1995
dollars for compensatory education. This estimate is the discounted value of such costs for
age 7 through 9.
Recommendations and actual practice based on information from CDC (1991), AAP (1995),
and medical practitioners. These estimates are the discounted costs per newborn
associated with each blood lead Risk Group.
em
.j:>
o
Co)
m
>
-------
.
Second, the CPS data reported total counts of persons in la-year age groups, but earnings
for those with earnings and counts of those with earnings in 5-year age groups. For the
purposes of this analysis, it was assumed that employment rates within the 5-year age
groups were equal to those of the to-year age groups. The CPS data, however, did not
include an estimate of the total persons, and thus employment rates, in each education group
for individuals in the 18-24 age group. Since employment rates for both men and women,
age 18-24, are available, this analysis assumes that the employment rate within each
education group of the 18-24 age group was equal to the overall employment rate for that
age group.
.
Third, the analysis assumes that the 1992 distributions of earnings and educational
attainment will hold constant over several decades (with some minor exceptions concerning
educational attainment) and are representative of those faced by children born in 1997. It
was assumed that the educational distribution remains the same as the current distribution
until those born in 1997 are older than age 49. After that age, the assumed educational
distribution is fixed at the distribution of those aged 45-49 in the CPS data. The data
beyond age 49 reveal significant declines in educational attainment and it is for this reason
that this age was selected as the cutoff. Because average years of schooling have tended to
rise over this century, older people often have fewer years of schooling than younger people.
If this assumption was not made, the model would have individuals losing years of
education over time. With these assumptions, it was possible to calculate the average
earnings for persons in the 18-24 age group and in the five year age groups ranging from 25-
29 through 60-64.
Present Value of Lifetime Earnings. The estimated average earnings were used to predict the present
value of future annual earnings over the lifetimes of those born in 1997, by using appropriate survival
rates, productivity increases, and discount rates. Survival rates (P) are the probability that a newborn
person will survive to a given age N. Survival rates are the multiplication of two probabilities: the
probability that a newborn survives to age N-l and the probability that the individual will survive from
age N -1 to age N. The US Department of Commerce (1992b) provided the probability that a person of a
particular age dies some time in that age year. This is sufficient information to calculate survival rates
for all ages as described above. Because the model is evaluating future populations, there were some
uncertainties associated with these probabilities. The model assumed that real earnings will increase by
one percent per year to reflect the fact that some portion of productivity increases (X) are reflected in
real earnings increases. The nation's productivity or output per capita tends to rise over time as the
capital stock rises. As explained in Chapter 3, the discount rate (r) used in this analysis is 3 percent.
The following formula estimates the present value of average earnings at age A, for a male or female
born in 1997. It is important to note that female and male earnings were calculated separately.
where:
PV =
A =
PVA = L ~~AY ~ N(1 +X)N-A+O.5j((1 +r)N-A+l)
present value of the total sum of earnings of a male or female received between ages
A and 64;
current age of male or female;
9403 EA
5-7
-------
N =
y =
P =
x =
successive ages of male or female in the future (A+ 1, ..., 64);
average annual earnings of male or female for a particular age (N);
survival rate of male or female for a particular age (N);
productivity rate of male or female assumed at the midpoint of age N; and
discount rate for the beginning of age N.
r
=
The present value average earnings are based on the 1992 CPS data, survival rates, an assumed discount
rate of 3 percent, and an assumed increase in productivity of 1 percent per year. The average earnings for
males was $485,946 and $250,797 for women. The average earnings for the entire population was
$366,021. This is a participation-weighted average obtained using the following equation:
A ve. Earnings for Pop. = % pop. male x male earnings + % pop. female x female earnings
There are several uncertainties associated with this approach to calculating the present value of the
average earnings stream:
First, uncertainties arise concerning the earnings distribution mainly because it is a
projection of lifetime earnings. Children born in 1997 and after will not enter the labor
market for several decades. The type of labor market that will exist and the distribution of
skills and education of this future labor force are both unknown. In addition, labor force
participation rates, a real wage growth rate of one percent, and year-to-year survival
probabilities are all assumed to stay the same until the year 2110. This includes the 64 year
full working life for children born in year 2046, the last year of the model run. Labor force
participation rates of women, the elderly, and other groups will most likely continue to
change over the next decades. Real earnings of women will probably continue to rise
relative to real earnings of men. Unpredictable fluctuations in the economy's growth rate
will probably affect labor force participation rates and real wage growth of all groups.
Medical advances will probably raise survival probabilities. Presently, the model uses
information on the 1992 distribution of education and earnings by age groups to characterize
the future labor market. This involves making the assumption that present trends will
continue in the future and it is unclear how this might bias results.
Second, this approach assumes that what individuals are paid in the market truly reflects
their marginal product as laborers. Earnings is used in place of marginal product in this
model because the latter value is a much more involved calculation. However, there is
concern that certain groups of people are discriminated against (e.g., women and minorities)
in the labor market such that they do not receive their true marginal product. For this reason,
the average earnings calculated here may be an underestimate of the true marginal product.
.
Third, the use of earnings is an incomplete measure of an individual's value to society. This
is particularly true for individuals who choose to not participate in the labor force for all of
their working years. If the opportunity cost of non-wage compensated work is assumed to be
the average wage earned by persons of the same sex, age, and education, the average lifetime
earnings estimates for these people would be significantly higher.
5-8
~403 EA
-------
.
Fourth, the CUITent model uses the earnings of all persons to determine average earnings. If
the exposed population are significantly different then the national population (i.e., minority
populations) then the CUITent model may be misrepresenting their future earnings streams.
For example, if the exposed population earn lower earnings relative to the national
population for reasons other than their lead exposure, then the average used by this analysis
may be an over estimate. Yet, as emphasized in the previous comment on discrimination, it
might be more appropriate to use the sample of the national population that best reflects the
marginal product of labor to assess the benefits of preventing IQ reductions.
Effects on Earnings from Changes in IQ. The second part of the benefits estimation for IQ changes
relies on the Salkever (1995) study. The value of avoiding a single IQ point loss was modeled as a loss
in expected lifetime earnings. Direct and indirect effects on earnings were considered in valuing lost IQ
points. The direct effect is the sum of the effects of IQ test scores on employment and earnings for
employed persons, with the years of schooling held constant. The indirect attributes are the effect of IQ
test scores on years of schooling attained, and the subsequent effect of years of schooling on the
probability of employment, and on earnings for employed people.
Salkever (1995) provides updated estimates of all the necessary parameters using the most recent
available data set, the National Longitudinal Survey of Youth (NLSY). Three regression equations
provide these parameters. The years of schooling regression shows the association between IQ scores
and the highest grade achieved, holding background variables constant. The employment regression
shows the association between IQ test scores, highest grade, and background variables on the probability
of receiving earned income. This regression provides an estimate of the effect of IQ score on
employment when schooling is held constant, and the effect of years of schooling on employment, when
IQ is held constant. The earnings regression shows the association between IQ test scores, highest grade
achieved, and background variables on earnings, for people with earned income.
These three regressions provide the parameters needed to estimate the total effect of a loss of an IQ point
on earnings. The direct effects of IQ on employment and earnings for employed persons, holding
schooling constant, come from the employment and earnings regressions. The indirect effect of IQ on
employment through schooling is the product of the effect of IQ on years of schooling, from the years of
schooling regression, and the effect of highest grade on employment, from the employment regression.
The indirect effects of IQ on earnings for employed persons through schooling is the product of the effect
of IQ on years of schooling, from the years of schooling regression, and the effect of highest grade
achieved on earnings for employed persons, from the earnings regression.
Based on the Salkever (1995) study, the most recent estimate of the effect of an IQ point loss is a
reduction in earnings of 1.93 percent for men and 3.22 percent for women, which is a participation-
weighted average of 2.39 percent.
There are numerous uncertainties associated with implementing this approach. Several assumptions were
necessary to estimate the foregone earnings associated with IQ reductions. First, it was assumed that IQ
decrements incUITed through lead exposure persist throughout the exposed child's lifetime. Second, it
was assumed that population changes in IQ have effects analogous to individual changes in IQ. This
assumption suggests that every unit decrease in IQ has an equal effect regardless of where an individual
~403 EA
5-9
-------
is in the IQ distribution and what the total magnitude ofthe person's IQ change is. Additional
uncertainties associated with specific components are discussed in the following sections.
Value of Foregone Earnings. The next step for determining the value of an IQ point involves
combining the percent earnings loss estimate with an estimate of the present value of expected lifetime
earnings. The present value average earnings of $336,021 is multiplied by the 2.39 percent earnings loss
to yield an average value of $8,708 per IQ point.
This IQ point value of $8,708, however, does not account for the costs associated with additional years of
education. The increase in lifetime earnings from additional education is the gross return to education.
The cost of the marginal education must be subtracted from the gross return in order to obtain the net
benefit per IQ point. There are two components of the cost of marginal education; the direct cost of the
education, and the opportunity cost of lost income during education. An estimate of the educational cost
component is obtained from the U.S. Department of Education (1993). The marginal cost of education
used in this analysis is assumed to be $5,500 per year. This figure is derived from the Department of
Education's reported ($5,532) average per-student annual expenditure (current plus capital expenditures)
in public primary and secondary schools in 1989-90. For comparison, the reported annual cost of college
education (tuition, room and board) in 4 year public institutions is $4,975, and $12,284 for private
institutions.
Salkever estimated 0.1007 reduced years of schooling per IQ point lost. Therefore, the estimated cost of
an additional 0.1007 years of education perIQ point where one year costs $5,500 is $554. Because this
marginal cost occurs at the end of formal education, it must be discounted to the time the exposure and
damage is modeled to occur (age zero). The average level of educational attainment in the population
over age 25 is 12.9 years (U.S. Department of Education, 1993). Therefore, the marginal educational
cost is assumed to occur at age 19, resulting in a discounted present value cost of $316.
The other component of the marginal cost of education is the opportunity cost of lost income while in
school. Income loss is frequently cited as a major factor considered in the decision to terminate
education, and must be subtracted from the gross returns to education. An estimate of the loss of income
is derived assuming that people in school are employed part time, but people out of school are employed
full-time. The opportunity cost of lost income is the difference between median annual full-time income
of $16,501, and $5,576 for part-time employment (U.S. Department of Commerce, 1993a). The lost
income associated with being in school an additional 0.1007 years is $1,100, which has a present
discounted value at age zero of $627.
The net benefit per IQ point in 1992 dollars is the gross value of an IQ point minus the costs of
education. This relationship is shown in the equation:
$8,708 - $316 - $627 = $7,765.
The GDP price inflator was used to adjust the 1992 dollars to 1995 dollars. This results in an IQ point
value of $8,346.
i403 EA
5-10
-------
5.3.2 Valuing Increased Educational Resources
Two categories of increased educational resources needed as a result of lead exposure are considered as
additional effects from lead exposure. First, lead exposure results in an increase in the number of
children with IQs less than 70, an indicator of mental handicap (Battelle 1997). During all their school
years, these children are likely to require special education tailored to the mentally handicapped. In
addition, some children whose blood lead is greater than 20 Ilg/dL may be less impaired but will still be
affected enough to need several years of compensatory education in addition to regular schooling. The
following sections describe the approaches used to obtain estimates of increased educational resources
needed. Exhibit 5-2 summarizes the data used to derive educational resource expenditure estimates.
Special Education for Children with IQs less than 70. To assign a value to the reduction in the
number of infants with IQs less than 70, an estimate of the reduction in education costs was calculated
using available data on educational expenditures. This approach is a clear underestimate of the total
benefits associated with avoiding cases ofIQ less than 70.1 Kakalik et al. (1981) used data from a study
prepared for the Department of Education's Office of Special Education Programs. They estimated that
part-time special education costs for children who remained in regular classrooms were $3,064 on an
annual basis in 1978 (in addition to regular class costs). Adjusting for changes in the GDP implicit price
deflator yields an estimate of $6,458 per child in 1995 dollars. This incremental estimate of the cost of
part-time special education was used to estimate the annual cost per child needing special education as a
result of impacts of lead on mental development. Costs were assumed to be incurred from ages 7 through
18. Discounting future expenses at a rate of 3 percent yields an expected present value cost of
approximately $53,836 per child. This is the benefit assigned for each case of IQ under 70 avoided. It is
an underestimate of the benefit since Kakalik et aI. (1981) measured the increased cost to educate
children attending regular school. The costs of attending a special education program were not
considered and are most likely considerably higher than those associated with regular schooling.
The total cost of special education is simply the reduction in the probability a child will have an IQ less
than 70 multiplied by the number of children born in a specified year and then multiplied by the cost of
special education. For example, if the baseline probability of an IQ less than 70 is 0.5% and the post-
intervention probability falls to 0.3% and there were a total of 1,000 children born, then the benefit
society accrues from reduced special education costs is $107,672 ([0.005-0.003]xl,000x$53,836 =
$107,672).
Compensating Education for Children with Blood Lead Levels Greater Than 20 IlgldL. When
calculating the cost of compensatory education, three relatively conservative assumptions were made.
First, it was assumed that no children with blood lead levels below 20 Ilg/dL would require compensatory
education later in life. This is conservative since many studies show cognitive effects at 15 Ilg/dL.
Second, it was assumed that only 20 percent of the children above 20 Ilg/dL would be severely affected
enough to require and receive some compensatory education. Third, it was assumed that each child who
needed compensatory education would require it for three years (age 7 through 9). Studies of the
persistence of cognitive effects indicate this is a conservative estimate and that effects often last longer
than three years.
The largest part of this benefit is the parents' willingness to pay to avoid having their child become
mentally handicapped, above and beyond the increased educational costs.
~403 EA
5-11
-------
Benefits were calculated by assuming that 20 percent of the children with blood lead levels greater than
201lg/dL received compensatory education for three years. After this time, no further blood lead related
educational expenditures were incurred by those children? The Kakalik et al. (1981) estimate of part-
time special education costs for children who remained in regular classrooms was also used to estimate
the cost of compensatory education for children suffering low-level cognitive damage. As indicated
above, adjusting for changes in the GDP implicit price deflator yields an estimate of $6,458 per child per
year in 1995 dollars. Discounting future costs at a rate of 3 percent annually to account for the age at
which costs are incurred yields a present value estimate of $15,298 in 1995 dollars.
The total value to society from a reduction in the probability of blood-lead concentrations greater than 20
Ilg/dl is the difference in the probability of having a blood-lead concentration greater than 20 Ilg/dl, times
20% of the number of children born in a specified year, times the present value per child. For example, if
the probability is reduced by .05% and there are 1,000 children born then the total value is $1,530
(0.0005xO.20xl,OOOx$15,298 = $1,530).
5.3.3 Valuing Increased Blood Lead Screening and Medical Treatment
Blood lead screening programs have been established in many public health programs because children
may remain asymptomatic even though blood lead levels are elevated. Screening programs attempt to
identify children who are at risk of developing lead exposure related illnesses. Once elevated blood lead
levels are detected, treatment costs are incurred to reduce blood lead to less serious levels, thereby
avoiding or mitigating adverse health effects. Follow-up blood lead tests are used to ensure that
intervention has accomplished the intended risk reduction.
The model calculates benefits as the reduction in these screening and medical intervention costs caused
by lead interventions.
The Centers for Disease Control (CDC), in their 1991 statement Preventing Lead Poisoning in Young
Children, has recommended protocols for blood lead screening and medical treatment for several
categories of blood lead levels, called risk groups (CDC, 1991; see Exhibit 5-3 for risk grouping). The
costs and assumptions used to determine the total cost per child are based primarily on treatment
protocols recommended in the 1991 CDC statement and the American Academy of Pediatrics (AAP,
1995). Recommended treatments may differ from typical treatments. However, the recommended
protocols are expected to serve as a reasonable approximation of typical treatment profiles. In cases
where information was lacking in the 1991 CDC recommendations, assumptions about the percent of
children treated in a given risk group were made based on actual practice.
2
See U.S. EPA (1986) for more detail on the data sources and the nature of the assumptions made to
quantify this benefit category.
5-12
i403 EA
-------
Exhibit 5-3
Risk Groups and Associated Screening and Medical Costs Per Child
Risk Group
I
IIA
liB
iliA
IIIB
IV
V
Blood-Lead Concentration
0-9 j.<.g/dL
10-14 j.<.g/dL
15-19 j.<.g/dL
20-24 j.<.g/dL
25-44 j.<.g/dL
45-69 j.<.g/dL
70 j.<.g/dL
Screening Costs
$58
$70
$169
$169
$325
$1 ,450
$1 ,450
Medical Costs
$0
$0
$58
$248
$353
$8,393
$8,393
In addition, although CDC does recommend all children younger than six should be screened, actual
practice suggests screening rates that are much lower. Therefore, this analysis assumes that the percent
of children screened is based on actual practice. Information from blood lead screening programs
indicates the level of screening in recent years (1994, 1995) is about 15 percent (MDDE, 1996; ILDPH,
1995). This level of screening represents the percent of children six years and younger who are screened
in a given year, whether they are being screened for the first time or are given repeat screening. This
value compares with an average screening rate of 17 percent from nine programs in 1983 (U.S. EP A,
1987). For the current analysis, a screening rate of 15 percent was assumed for both initial as well as
repeat screenings for risk groups I to lIIB.
Children are grouped into different risk groups based on their blood lead levels. The unit costs
determined for each risk group assume that each child is initially screened at age one, that elevated blood
lead levels are identified at the initial screening, and that children who are not medically treated remain in
the same risk group through age five. It was assumed that the total percent of children initially screened
were rescreened once a year from ages two through five, resulting in a total of five screenings per child.
Specific treatment elements include blood lead screening and other tests, medical treatment, and health
education. Where there were uncertainties regarding the percentage of children treated or the exact
treatment protocol used, conservatively low estimates for treatments and costs were made.
The medical costs listed in Exhibit 5-2 are based on costs of individual treatments presented in U.S. EPA
(1987), incorporate the probability that a child will be treated, and are updated from 1985 to 1995 dollars
using the average medical care cost index reported by the Bureau of Labor Statistics.3 A cost of $42 for
screening was taken from U.S. EP A (1987), updated to $81 in 1995 dollars. The cost of $42 includes the
full annual clinic fee of $20 per child presented in Table 4-2 of U.S. EP A (1987).4 The screening costs
The average medical care cost index increased from 113.5 in 1985 to 220.4 in 1990 (1982== 100) (CEA,
1996). As 220.4/113.5 == 1.94, a medical cost (1995$) == 1.94 x medical cost (1985$).
4
The analysis presented in U.S. EPA (1987) assumed that this annual fee was divided among Jour children.
However, because the current analysis assumes abatement decisions are based on blood lead levels of
children newly born into a household (rather than existing children in a household), it was assumed that
the full clinic fee may apply to one child only.
~403 EA
5-13
-------
encountered after the first year were discounted to 1995 dollars, using a 3% discount rate. Additional
information on the percent of children treated in different groups has been obtained from health care
practitioners and public health departments. The procedure used to determine these costs is presented in
Appendix 5.A.
The total benefit to society resulting from a reduction in the probability of being in a particular risk group
is simply the decrease in probability multiplied by the number of children born multiplied by the
reduction in screening costs. For example, if the reduction in the probability of being in risk group IIB is
0.2% and a total of 1,000 children are born, then the benefit to society from reduced screening and
medical costs is $454 (0.002 x 1,000 x (169 + 58) = $454).
5.4
Aggregation of Benefits
The values provide in Section 5.3 are the per-child benefits resulting from the reduction in household lead
levels. However, in order to calculate the net benefits of 9403 it is necessary to aggregate the benefits
accruing to all children born during the years 1997 to 2047.
In order to calculate the total present value of benefits one first calculates the value of benefits for each
cohort of children separately. A cohort of children is defined as all children born in a given year.
Benefits for each cohort are determined using information on the number of children born as well as
information on the per-child benefits from household lead reduction discussed in the previous section.
More specifically, the present value of total benefits for each cohort is calculated as the number of
children born in that cohort multiplied by the sum of the following benefit values: the value of foregone
earnings, the value of the decrease in the probability of an IQ less than 70, the value of the decrease in
probability of a blood-lead level greater than 20 f.lg/dl, and the value of a decrease in the probability of
falling into the high risk groupss. The value of benefits for each cohort is then discounted back to 1997
dollars. Finally the present value of total benefits for each individual cohort are summed together to
generate the total present value of benefits from the 9403 rule.
The aggregate monetary benefit of the specific 9403 standards being proposed are estimated to be $160
billion based on the blood lead changes predicted using the IEUBK model. Of this amount, over $157
billion is associated with the overall IQ point benefits related to future earnings.
Using the Empirical model, the estimated aggregate benefits of the 9403 standards being proposed are
$42.4 billion, of which $41.5 billion is associated with the overall IQ point benefits related to future
earmngs.
A more detailed series of aggregation equations can be found in Appendix 5.B of this chapter.
5-14
i403 EA
-------
Appendix 5.A Screening and Medical Costs for Risk Groups
This Appendix presents the screening costs and other medical costs, such as treatment for anemia,
neuropsychological evaluation, and chelation therapy for each of the blood lead level risk groups
described in Chapter 5. Where applicable, any assumptions that were made in calculating these costs are
discussed. Please note that the procedures for estimating the incidence of blood lead levels fa1ling in
various ranges used in the process of estimating screening and medical intervention costs were provided
in the risk assessment document (Batte1le, 1997).
Risk Group I (PbB 0 - 9 ~g/dL). CDC recommends a1l children younger than six years old be screened
for elevated blood lead levels. In addition, CDC recommends yearly repeat testing for children in this
risk group. In actual practice, universal screening is not being achieved. Therefore, the proportion of
children screened was multiplied by the screening cost and discounted to 1995 dollars. The resulting cost
per child for Risk Group I is $58.
Risk Group IIA (PbB 10 - 14 ~g/dL). As with Risk Group I, Risk Group IIA also requires repeated
blood lead screening. The CDC recommends that children in this risk group be screened more frequently
during the first year than children in Risk Group I. Specifically, children should be tested every 3 to 4
months. After three consecutive tests indicate that blood lead levels remain below 15 llg/dL, the child
should be tested again a year later (CDC, 1991). This analysis assumes that children are tested once
every four months during the first year, that blood lead levels remain below 15 llg/dL, and that children
are then tested once in each of the four subsequent years (at ages two through five.) Multiplying the
proportion of children expected to be screened, the cost of one test, and discounting to 1995 dollars
results in a cost per child for Risk Group IIA of $70.
Risk Group liB (PbB 15 -19 ~g/dL). Based on the CDC recommendation that children in this risk
group need repeat screenings every three or four months, this analysis assumes the frequency of screening
for Risk Group IIB is three times per year (for ages one through five).
An additional cost not estimated for Risk Group I and IIA includes the cost of health education. CDC
notes that it is prudent for parents to use simple interventions to decrease hazardous levels of lead in the
home. Education about the types of intervention that can be done may be achieved through a face-to-face
interview with the family (CDC, 1991). U.S. EPA (1987) presents the cost of a one-time personal
evaluation by a professional as $200. Updating this cost to $1995 results in health education costs of
$388.
The total cost for this risk group (incorporating the probability that a child will be screened and treated)
is $227 per child.
Risk Group iliA (PbB 20 - 24 ~g/dL). CDC recommends that children in this risk group be rescreened
every three to four months; the current analysis assumes that children are screened once every four
months. Health education costs are assumed to be the same as Risk GroupIIB.
In addition to the screening and education costs, CDC recommends that children with blood lead levels
greater than or equal to 20 Ilg/dL be tested for iron deficiency. A cost for this test of $20 (1985 dollars)
from U.S. EPA (1987) was updated to a value of $38 (1995 dollars). Based on results of the iron
~403 EA
5-15
-------
deficiency test, children in this group may need treatment for anemia. Medical practice indicates that
about half of children who are screened for elevated blood lead levels are treated for anemia (Shannon,
1996; McCord, 1996). This information does not specify whether a greater proportion (than 50 percent)
of children in higher blood lead level ranges require treatment for anemia; therefore, the current analysis
assumes that half of the children with blood lead levels of 20-24 ~g/dL are anemic and would require
treatment for anemia. A cost of $63 for anemia treatment was presented in U.S. EPA (1987). Updating
this cost to 1995 dollars results in a cost of $122.
In addition to these costs, CDC recommends that children with blood lead levels ~ 20 ~g/dL should have
a pediatric evaluation, with special attention given to neurologic examination and psychosocial and
language development (CDC, 1991). In this analysis, it is assumed that all children in this risk group
receive this neuropsychological evaluation. For children given chelation therapy, this examination may
be important both at the time of diagnosis and when the child approaches school age; however, this
analysis assumes only one neurological evaluation is performed. U.S. EP A (1987) provides a cost of
$600 ($1164 in 1995 dollars) for this evaluation.
Because only minimal data exist about chelating children with blood lead levels less than 20 ~g/dL, CDC
recommends that these children should not be chelated (CDC, 1991). In addition, AAP (1995) notes that
chelation treatment is not indicated in patients with blood lead levels less than 25 ~g/dL. Therefore, it is
assumed that all children in this risk group (and lower risk groups) are not chelated.
Incorporating probabilities of screening and anemia treatment, the total cost per child for this risk group
is $417.
Risk Group IIIB (PbB 25 - 44 ~gIdL). CDC recommends that children in this risk group be rescreened
every three to four months; health education costs, tests for iron deficiency for this risk group, and a
neuropsychological evaluation are also recommended. However, costs for this risk group differ
depending on whether or not a child is chelated. Depending on results of blood lead and further testing,
some children in this group may require chelation therapy to lower their blood lead levels to acceptable
levels. The costs and assumptions for blood lead screening, education, anemia treatment, and
neuropsychological evaluation are the same for nonchelated children in this group as for children in Risk
Group IlIA. Frequency of screening differs for chelated children compared with non-chelated children
(as described below). In addition, treatment for anemia is assumed to be included as part of the chelation
therapy, and is therefore not included as a separate cost for chelated children.
Children in this risk group must undergo an edetate disodium calcium (CaNa2 EDTA) provocation test to
determine whether they will respond to chelation therapy. The cost of performing this test on an
outpatient basis was presented as $150 in U.S. EPA (1987). This cost is used in the current analysis and
updated to 1995 dollars for a resulting value of $291.
There are varying recommendations about whether children in this risk group should be chelated. CDC
recommends that children in this risk group who have positive CaNa2 EDT A provocation test results
should undergo a 5-day course of chelation (CDC, 1991). Information suggests that about seventy-six
percent of children with blood lead levels 35-44 ~g/dL and 35 percent of children with blood lead levels
25 - 34 ~g/dL had positive provocation test results as noted in one source (CDC, 1991 as cited in
Markowitz and Rosen, 1991). Additional information indicates varying protocols of chelation
5-16
i403 EA
-------
requirements for this risk group. At Children's Hospital in Boston, Massachusetts all children in this risk
group are chelated (Shannon, 1996). At Children's Health Care of St. Paul, Minnesota where most
children in Minnesota with elevated blood lead levels are treated, chelation is not recommended for blood
lead levels lower than 40 J.lg/dL (McCord, 1996). Based on the information above, this analysis assumes
that fifty percent of children in this risk group require chelation.
Information suggests that an oral method of chelation performed on an outpatient basis can be used for
children in this Risk Group. Although CDC (1991) does not give a strong recommendation about
whether to chelate these children on an inpatient or outpatient basis or the type of chelating agent that
should be used, they do note that some practitioners use an oral chelating agent. The American Academy
of Pediatrics suggests that children in this risk group may benefit from oral therapy, which can be done
on an outpatient basis (AAP, 1995). Other information also suggests that chelation for blood lead levels
in this range may be performed on an outpatient basis (U.S. EP A, 1987; McCord, 1996) if the child can
be kept away from the source of exposure (U.S. EPA, 1987). Based on these recommendations, this
analysis uses an outpatient chelation cost from U.S. EP A (1987); the cost is $582 in 1995 dollars
(updated from $300 in 1985 dollars).
Children chelated once may need additional chelation treatment to bring blood lead down to acceptable
levels on a long-term basis (CDC, 1991). This analysis assumes that fifty percent of children who
receive a first chelation will require a second treatment based on information from U.S. EPA (1987).
Children who receive chelation therapy should be followed closely for at least a year or more to recheck
blood lead levels. CDC (1991) recommends retesting every other week for 6-8 weeks, and then once a
month for 4-6 months (or more often depending on the type of chelation performed.) Based on this
information, the current analysis assumes that seven repeat tests will be performed within the year
following chelation. Chelation is expected to decrease blood lead levels to below 25 J.lg/dL (CDC, 1991).
For this analysis, it was assumed that lead levels are decreased to below 25 J.lg/dL but remain at levels
above 15 J.lg/dL. Therefore, in addition to the seven repeat screenings in the year following chelation,
chelated children require screenings once every four months (after the first year) through age five as
suggested by CDC for children with blood lead levels greater than 15 J.lg/dL (CDC, 1991). Follow up
testing costs are the same as initial screening costs, at $81 in 1995 dollars.
The total cost per child for this risk group, incorporating probabilities of screening and treatment, is
$678.
Risk Group IV (PbB 45 - 69 ~g/dL). Only fifteen percent of children in lower risk groups are expected
to be screened. However, children in this group may present symptoms of lead poisoning, such as
lethargy, anorexia, vomiting, abdominal pain (CDC, 1991). For this analysis, it is assumed that all
children in this risk group will exhibit symptoms and will require follow up blood lead level testing.
CDC (1991) recommends that children in this group should not be given a provocation test, but should
be referred for appropriate chelation therapy immediately upon identification of this blood lead level.
Several sources suggest that chelation may be done on an inpatient or an outpatient basis for children in
this risk group. Although AAP (1995) suggests that children in this group may be orally chelated and
McCord (1996) notes that most chelations done in St. Paul are done on an outpatient basis, CDC (1991)
discusses a treatment regimen limited to CaNa2EDTA for this group of children because experience
~403 EA
5-17
-------
using other treatments is limited. AAP (1995) recommends against perfonning CaNa2EDT A on an
outpatient basis, and in addition, suggests that children in this group may need to be hospitalized for the
initiation of therapy. Based on this information, the current analysis assumes that chelation for this risk
group is performed using CaNa2EDTA on an inpatient basis. U.S. EPA (1987) lists a cost for inpatient
CaNa2EDTA therapy at $1,500. Based on CDC recommendations, increased frequency of follow up
blood lead testing after chelation is assumed to be the same as for Risk Group IIIB.
As with Risk Group IIIB, there may also be a need for repeat chelations. CDC (1991) indicates that a
second chelation may be needed, and perhaps a third chelation may be required if blood lead levels return
to levels greater than 45 /lg/dL. For children with elevated blood lead levels, U.S. EP A (1987) assumes
that fifty percent of children who have one chelation will require a second chelation, and that fifty percent
of those who receive a second chelation will require a third. The same assumptions are used in this
analysis.
As noted under Risk Group IIIB, chelation is assumed to decrease blood lead levels below 25 /lg/dL.
Therefore, after the first year of increased testing following chelation, testing is assumed to occur once
every four months for the subsequent years through age five (recommended for children with blood lead
levels greater than 15 /lg/dL).
Costs of health education, tests for iron deficiency, and neuropsychological evaluation are the same as for
Risk Group IIIB. As noted under Risk Group IIIB, the cost of anemia treatment is assumed to be
covered in the cost of chelation.
The total cost for a child in this risk group, incorporating probabilities of medical treatment, is $9,843.
Risk Group V (PbB ~ 70 ~g/dL). The lead poisoning symptoms listed for Risk Group IV are most
commonly associated with blood lead levels of 70 /lg/dL and greater. In addition, encephalopathy may
be associated with blood lead levels as low as 70 /lg/dL (CDC, 1991). It is assumed that for this
analysis, children in this category would exhibit symptoms, and therefore all children in this risk group
would be screened.
Children in this risk group represent a medical emergency and should be hospitalized and chelated
immediately (CDC, 1991). Therefore, children in this group would require inpatient chelation. A cost of
$2,000 for inpatient chelation was listed in U.S. EPA (1987); the equivalent cost updated to 1995
dollars is $3880 and is used in the current analysis. The same assumptions about the need for repeated
che1ations and follow up screenings after chelation are used for this group as for Risk Group IV.
Children in this risk group also require an iron deficiency test, health education and neuropsychological
evaluations, as indicated for Risk Groups IIIB and IV (CDC, 1991). Also, as with the previous two risk
groups, costs for anemia treatments are not included because the protocol of chelation treatment is
expected to alleviate iron deficiency (U.S. EPA, 1987).
The total cost for a child in this risk group, incorporating probabilities of medical treatment, is $9,843.
Use of Costs per Child in Benefits Analysis. The above monetary valuesdetennined for each risk
group are used to estimate benefits associated with the change in the number of children in each risk
5-18
~403 EA
-------
group resulting from a change in geometric mean blood lead level. The monetary values of the avoided
health effects used in the benefits estimation are summarized in Exhibit 5-2. The results of this benefits
analysis are presented in Chapter 6.
~403 EA
5-19
-------
Appendix 5.B Aggregating Benefits from Environmental Lead
Reduction
As stated previously, this study assumes that after g403 is implemented, all children born into homes
exceeding the standard will have their homes appropriately treated to insure the children's protection
during the first six years of life. Accordingly, all children born post-g403 will be protected at least up to
the level mandated by the standard. (Not all childbirths will trigger interventions, however, because
many children will be born into homes that already meet the standard.)
The benefits analysis is based on calculating benefits for all children born in the same year, referred to as
a cohort of children. Aggregating the benefits of all children born between 1997 and 2046 involves
summing the total present value of benefits for all cohorts. The expected blood lead distribution must be
calculated separately for each cohort (based on either the IEUBK or empirical model), because in each
year different home types are demolished at different rates. As a result, a changing mix of homes causes
the posH403 blood-lead distributions to vary slightly from one year to the next. Basically, the
measurement of health benefits is repeated 50 times -- once for each cohort of children. The expected
benefits from g403 will consequently differ by a small amount for each group of children born between
1997 and 2046, even before discounting is taken into account.
To illustrate how benefits are determined for each cohort, equations are shown for children born in 1997,
1998, and 2046 (years 1,2, and 50, respectively). The following variables are used in each equation:
let
TB(Cohort t) = total benefits for the cohort born in year t (Cohort t)
N = total number of households in 1997
Pt = probability that a child is born into a household in year t
Et = number of homes demolished from 1997 to year t
MQt = predicted increase in average IQ (post-g403 IQ minus pre-g403 IQ) for Cohort t
Ll70t = predicted decrease in the probability of IQ less than 70 for Cohort t
Ll20t = predicted decrease in the probability of blood-lead greater than 20 Ilg/dl for Cohort t
LlTlp ..., and LlT7t = the predicted change in the probability of belonging to risk group I, ...,
and V, respectively for Cohort t.
PV1Q = present value of an increase in IQ of one point per child in 1995 dollars
PV70 = present value of part-time special education costs for years 7-18 per child in 1995
dollars
PVzo = present value of part-time special education costs for years 7-9 per child in 1995 dollars
PV Tl, ..., and PV T7 = the present value of screening and medical treatment in risk group I, ...,
and V, respectively (per child in 1995 dollars)
r = discount rate (3%)
Note that the variable for housing demolition was based on information provided in Battelle (1996) and
that the variables for changes in IQ and blood lead were based on information provided in Battelle (1996)
and Battelle (1997).
The first example shows the calculation for benefits to children born in 1997 (Cohort 1). These benefits
are represented in Equation 1:
5-20
~403 EA
-------
Equation 1:
TB(Cohort 1)=
PjX(N-Ej)X((AIQj)x(PVIQ) + (A70j)x(PV70) + 0.20x(A20j)x(PV 20) + (LlT1 t )X(PV Tj) + ... + (Ll T7)x(PV TI))
This equation simply takes the number of children born in 1997 (Pj X(N-Ej)) and assigns to them values
from four benefit categories. A reduction in number of homes from attrition (Ej) is included even in 1997
because the analysis assumes that attrition takes place at the beginning of each year. The four benefit
categories are the following:
1.
2.
3.
4.
Foregone earnings regained because of reduced loss of IQ points;
Decrease in the probability of an IQ score less than 70;
Decrease in the probability of a blood-lead level greater than 20 /-lg/dL; and
Decreases in the probabilities of falling into CDC's high blood lead risk groups.
Again, these four benefit calculations are based on the blood lead distribution changes calculated using
the IEUBK or empirical model analysis associated with Cohort 1.
Total benefits to subsequent cohorts are slightly more involved because the benefits must be discounted
back to 1997. The total benefits for Cohort 2 (children born in 1998) are shown in equation 2:
Equation 2:
TB(Cohort 2)=
.e2x(N-E2)X((AI02)x(PVIQ) + (A70?)x(PV70) + 0.20x(A202)x(PV20) + (ATl, )x(PVn) +... + (AT7,)x(PVnll
(1 +r)
Again, this equation states that the number of children born into pre-1997 housing (Px(N -E2)) are
assigned benefits from an increase in IQ, a reduction in the probability of an IQ less than 70, a reduction
in the probability of a blood-lead concentration greater than 20 /-lg/d1, and the reduced probability of
belonging to a high risk group. These benefits accrue when the children are born in 1998 so they are
discounted back 1 year to 1997.
The total benefits for Cohort 50 (children born in 2046) are shown in equation 3:
Equation 3:
TB(Cohort 50) =
.eSOX(N-ESO)x((AIOso)x(PVIQ) + (A70so)x(PV70) + 0.20x(A20so)x(PV20) + (ATlt )x(PVTj) +... + (ATIt)x(PVnll
(1 +r)49
The number of children born into pre-1997 housing (PX(N-Eso)) are assigned benefits. The benefits
occur when the children are born in the year 2046 so they are discounted back 49 years to 1997.
~403 EA
5-21
-------
The final step involves summing the present value of total benefits for all 50 cohorts:
Total Benefits = TB(Cohort 1)+ TB(Cohort 2)+ . . . + TB(Cohort 50).
5-22
~403 EA
-------
References
American Academy of Pediatrics (AAP). 1995. Treatment Guidelines for Lead Exposure in Children.
Pediatrics. 96(1): 155-160.
Ashenfelter, O. and J. Ham. 1979. Education, Unemployment, and Earnings. Journal of Political
Economy 87(5):99-131.
Barnett, W.S. 1992. Benefits of Compensatory Preschool Education. The Journal of Human Resources
27(2): 279-312.
Battelle. 1996. Procedures and Results for Input to the Economic Analysis for Section 403. Prepared
for TCB, CMD, OPPT, US EPA, May 13.
Battelle. 1997. Risk Assessment to Support Standards for Lead in Paint, Dust, and Soil. Prepared by
Battelle, for National Program Chemicals Division, Office of Pollution Prevention and Toxics,
U.S. Environmental Protection Agency. EPA 747-R-97-006, December.
Bishop,1. 1992. The Impact of Academic Competencies on Wages, Unemployment, and Job
Performance. Carnegie-Rochester Series on Public Policy 37: 127-194.
Blackburn, M.L. and D. Neumark. 1993. Omitted-Ability Bias and the Increase in Return to Schooling.
Journal of Labor Economics 11(3): 521-544.
Blackburn, M.L. and D. Neumark. 1992. Unobserved Ability, Efficiency Wages, and Interindustry
Wage Differentials. The Quarterly Journal of Economics, 1421-1436.
Centers for Disease Control (CDC). 1991. Preventing Lead Poisoning in Young Children. U.S.
Department of Health and Human Services, Public Health Service. Atlanta, GA. October.
Centers for Disease Control (CDC). 1991. Strategic Plan for the Elimination of Childhood Lead
Poisoning, February.
Chamberlain, G. and Z. Griliches. 1976. More on Brothers. Harvard Institute of Economic Research
Discussion Paper No. 469. Cambridge, MA.
Council of Economic Advisors (CEA). 1996. Economic Report of the President. Government Printing
Office. Washington, D.C. February.
Cropper and Krupnick. 1989. The Social Costs of Chronic Heart and Lung Disease. Washington, DC:
Resources for the Future.
Ganderton, P. and P. Griffin. 1993. Impact of Child Quality on Earnings and the Productivity of
Schooling Hypothesis. Contemporary Policy Issues 11(July): 39-47.
~403 EA
5-23
-------
Griliches, Z. 1977. Estimating the Returns to Schooling: Some Economic Problems. Econometrics
45:1-22.
Grogger, J. and E. Eide. 1995. Changes in College Skills and the Rise in the College Wage Premium.
The Journal of Human Resources 30(2): 280-310.
Hanushek, E.A., Rivkin, S.G., and D.T. Jamison. 1992. Improving Educational Outcomes While
Controlling Costs. The Carnegie-Rochester Series on Public Policy 37: 205-238.
Illinois Department of Public Health (ILDPH). 1995. Get the Lead Out: Illinois Childhood Lead
Poisoning Surveillance Report. ILDPH, Childhood Lead Poisoning Prevention Program.
Springfield, IL.
Kakalik, l et al. 1981. The Cost Of Special Education. Rand Corporation Report N-1791-ED.
Krupnick, A.l and M.L. Cropper. 1989. Valuing Chronic Morbidity Damages: Medical Costs and
Labor Market Effects. Draft Final Report to the US Environmental Protection Agency, OPPE.
June 26, 1989.
Markowitz, M.E. and J.P. Rosen. 1991. Need for the Lead Mobilization Test in Children with Lead
Poisoning. Journal of Pediatrics. 119: 305-310. [As cited in CDC, 1991.]
Maryland Department of the Environment (MDDE). 1996. Facsimile from Environmental Analysis
Division to Amy Benson of Abt Associates Inc. Childhood Lead Registry (CLR) Occurrence of
Lead Poisoning in Children 0-72 Months; Reportfor 1/94 - 12/94. Sent November 15.
McCord, James. 1996. Personal Communication between Dr. James McCord, Children's Health Care of
St. Paul, MN and Amy Benson of Abt Associates Inc. November 25.
McGee and Gordon. 1976. The Results of the Framingham Study Applied to Four Other U.S.-based
Epidemiologic Studies of Coronary Heart Disease. The Framingham Study: An
Epidemiological Investigation of Cardiovascular Disease, Section 31, April.
Murnane, RJ., Willet, lB., and F. Levy. 1995. The Growing Importance of Cognitive Skills in Wage
Determination. The Review of Economics and Statistics, 77(2): 251-266.
Needleman, H.J., Schell, A., Bellinger, D., Leviton, A., and E. Allred. 1990. The Long-Term
Effects of Exposure to Low Doses of Lead in Childhood: An 11- Y ear Follow-up Report. The
New England Journal of Medicine 322(2): 83-88.
Piomelli et al. 1984. Management of Childhood Lead Poisoning. Pediatrics 4: 105.
Pirkle, J.L., J. Schwartz, lR Landis, and W.R Harlan. 1985. The Relationship between Blood Lead
Levels and Blood Pressure and its Cardiovascular Risk Implications. American Journal of
Epidemiology, 121: 246-258.
5-24
~403 EA
-------
Pockock, SJ., Smith, M., and P. Bagrhurst. 1994. Environmental Lead and Children's Intelligence: A
Systematic Review of the Epidemiological Evidence. BMJ 309: 1189-1197.
Pooling Project Research Group. 1978. Relationship of blood pressure, serum cholesterol, smoking
habit, relative weight and ECG abnormalities to incidence of major coronary events: final report
ofthe Pooling Project. Journal of Chronic Disease, 31.
Salkever, D. 1995. Updated Estimates of Earnings Benefits from Reduced Exposure of Children to
Environmental Lead. Environmental Research 70: 1-6.
Schwartz, J. 1993. Beyond LOELs, p Values and Vote Counting: Methods for Looking at the Shapes
and Strengths of Association. Neurotoxicity, 14(2-3),237-246, SummerlFall.
Schwartz, J. 1994. Societal Benefits of Reducing Lead Exposure. Environmental Research 66:
105-24.
Shannon, M. 1996. Personal Communication between Dr. Michael Shannon, Children's Hospital in
Boston, MA and Amy Benson of Abt Associates Inc. November 25.
Shurtleff, D. 1974. Some Characteristics Related to the Incidence of Cardiovascular Disease and Death.
The Framingham Study: An Epidemiological Investigation of Cardiovascular Disease, Section
30, February.
Silbergeld, E.K., J. Schwartz, and K. Mahaffey. 1988. Lead and osteoporosis: mobilization oflead from
bone in postmenopausal women. Environmental Research, 47, 79-94.
U.S. Department of Commerce. 1993. Money Income of Households, Families, and Persons in the
United States: 1992. Current Population Reports, Consumer Income, Series P60-184.
U.S. Department of Commerce. 1992. Statistical Abstract of the United States. US Department of
Commerce.
U.S. Department of Education. 1993. Digest of Education Statistics. US Department of Commerce.
U.S. Environmental Protection Agency. 1984. A Survey of Literature Regarding the Relationship
Between Measures of IQ and Income. Prepared by ICF, Inc., Report to US EPA, Office of
Policy Analysis, June.
U.S. Environmental Protection Agency. 1986. Reducing Lead in Drinking Water: A Benefit Analysis.
Office of Policy Planning and Evaluation, Draft Final Report, December.
U.S. Environmental Protection Agency. 1987. Methodology for Valuing Health Risks of Ambient Lead
Exposure. Prepared by: Mathtech, Inc. Prepared for: Ambient Standards Branch, Office of Air
Quality Planning and Standards, under EPA Contract No. 68-02-4323. December.
~403 EA
5-25
-------
u.s. Environmental Protection Agency. 1990. Review of the National Ambient Air Quality Standards
for Lead: Assessment of Scientific and Technical Information. OAQPS Staff Paper, Air
Quality Management Division, Research Triangle Park, NC., December.
U.S. Office of Management and Budget. 1992. Guidelines and Discount Rates for Benefit-Cost
Analysis of Federal Programs, Circular No A-94, Revised. October 29, 1992.
5-26
S403 EA
-------
6.
Net Benefits
The model described in Chapters 3, 4, and 5 was developed to provide estimates of the costs, benefits,
and net benefits, under a set of specific assumptions about the behavior of residential property owners
and managers, for alternative definitions of lead hazard standards addressing paint, floor dust, window
sill dust, and soil. The estimates obtained from the model are intended to inform the decision-makers
about the relative merits of the alternative standards from a benefit-cost perspective. Since both the costs
and the benefits of compliance increase as hazard standards become more stringent (i.e. more protective),
neither costs nor benefits by themselves provide a sufficient means for evaluating the relative merits of
alternative standards. Net benefits, however, which are calculated as the difference between the benefits
to society and the corresponding costs to society of compliance with a particular standard, can serve as a
measure of the degree to which society is better or worse off due to compliance with alternative hazard
standards, and thus better inform the decision.
By estimating net benefits for a broad range of alternative hazard standards, the analysis can identify one
(or more) combination of paint, dust and soil standards that maximizes net benefits (i.e. is the most
efficient standard in economic terms). The net benefits analysis can also measure the relative degree to
which society is made worse off from a standard that is less costly, but also less protective, than the one
generating maximum net benefits. Conversely, it can also indicate when a set of standards that is more
protective than the one that maximizes net benefit yields that greater degree of protection at a cost that
exceeds the value of the additional benefits.
Because the objective of Title X is to protect children, and the objective of 9403 is to provide guidance to
parents and property owners to that end, the benefit-cost model focuses on how to maximize net benefits
to children. In other words, the intent of 9403 is to inform people about what they should do -- and when
-- in the arena of protecting their children from the adverse affects of lead exposure. The consideration of
when actions are taken is important in the calculation of net benefits. The model assumes that
intervention actions are timed to happen just before a newborn child is introduced into the home. If
interventions were to occur later, the child would experience some exposure to lead levels that exceed the
standards. This would reduce the benefits that the child would otherwise receive. If interventions were to
occur well before the appearance of the infant, money would be spent with no immediate benefits to the
child, thus increasing the costs relative to the benefits. For these reasons, the model assumes that hazard
testing and intervention will occur just before the appearance of the newborn. Although it is recognized
that this is not necessarily how individuals will behave with respect to these standards, structuring the
analysis in this way provides a systematic approach to estimating what the net benefits of these standards
might be if affected parties do behave in a manner that both maximizes the potential benefits and
minimizes the potential costs of acting to reduce lead exposure to children. Therefore, all of the net
benefits estimates presented in this chapter must be viewed in the context of the above fixed assumptions
concerning the timing of actions taken.
The remainder of this chapter addresses how the costs and benefits vary across alternative candidate
hazard standards, identifies the candidate standards that maximize net benefits, and compares the
maximum net benefits to the values estimated for the particular set of standards being proposed by EP A,
presented in the table below.
~403 EA
6-1
-------
Media Proposed Standards
Paint Condition 1995 HUD Guidelines for deteriorated
paint*
Floor Dust 50 IJg/ff
Window Sill Dust 250 IJg/ft2
Soil 2000 ppm
* For the analysis the following values were used:
Damaged Paint: Interior: 10 sq.ft. or more -- Repair
50 sq.ft. or more -- Abatement
Damaged Paint: Exterior: 20 sq.ft. or more -- Repair
100 sq.ft. or more --
Abatement
Sections 6.1 through 6.3 examine the trends in costs, benefits, and net benefits when the numerical
standards for floor dust, window sill dust, and soil are varied individually, with the remaining standards
being fixed at their respective proposed values. As explained in Chapter 3, the analysis assumes that
homes that receive any lead intervention will receive interventions for all media (floor dust, window sill
dust, paint and soil) that exceed the standards. This, combined with the fact that there are interactions
among the interventions for both the costs and the benefits, means that standards can not be accurately
evaluated one at a time. Instead, the standards for a single medium must be evaluated in the context of
specified standards for all other media. To allow for this, the analysis calculated costs, benefits and net
benefits for all possible combinations of standards.! For simplicity of presentation, however, this chapter
presents the results of the analysis varying the standard for one medium at a time, while holding the
standards for the other media constant at the level of their proposed standard. The purpose of these
sections is to provide the reader with some insights regarding the relationship between reducing exposure
from these media and the resulting net benefits.
It is also important to note that the net benefit results are presented separately here for the two different
blood lead models used by EP A in the risk assessment, and the differences in the results obtained from
these two models are explored. In general, the analysis using the IEUBK model generates higher
estimates of blood-lead changes, suggests larger net benefits at any particular set of standards, and
points toward more stringent standards as maximizing net benefits, than suggested by the Empirical
model.
The differences ~n the estimates resulting from the two models are due to several factors including the
incorporation of different variables and very different functional forms to relate environmental lead levels
to children's blood lead levels. Notably, the functional form of the IEUBK Model is such that it is much
more sensitive to changes in environmental lead than the Empirical Model. Also, the IEUBK Model uses
For each medium, the alternative standards were defined in terms of incremental changes in the levels of
lead. For example, floor dust standards varied by increments of 10 flg/ft2 (e.g., 40 flg/ftZ, 50 flg/ft2, 60
flglft2, 70 flg/ft2, etc.). Likewise, soil standards were analyzed in increments of 50 ppm (150 ppm. 200
ppm. 250 ppm, 300 ppm, etc.). All combinations of standards were analyzed.
6-2
~403 EA
-------
lead dust concentrations and the Empirical Model uses dust lead loadings as input variables. Since dust
lead concentrations and loadings are not well correlated in the actual housing unit data collected by HUD
and used for this analysis, these differences in input variables result in differences in estimated blood-lead
changes and thus benefits. Loading and concentration are not necessarily correlated: loading means the
mass of lead in dust per unit area, and concentration means the mass of lead per mass of dust. In a home
with a very low dust load, dust lead loading must also be small; but it is quite possible for the
concentration of lead in the dust present to be high.
A third major cause of the difference between IEUBK and Empirical Model-based results are the changes
in dust contamination that result from paint interventions. Based on the risk assessment, dramatic
reductions in dust lead concentration accompany all paint interventions, while reductions in dust lead
loadings accompany only the interior paint abatements. These interior paint abatements are relatively
rare occurrences. Consequently, paint interventions lead to very large benefits under the IEUBK Model,
whereas they lead to negligible benefits under the Empirical Model.
The reader is referred to Chapters 3 and 5 and to the Risk Assessment document prepared for this rule
(Battelle 1997) for further discussion of these two blood lead models.
6.1 Costs, Benefits and Net Benefits for Various Candidate Floor Dust Hazard Standards
The 9403 standards will define lead-based paint hazards, as well as levels of lead in floor dust, window
sill dust, and soil considered to be hazards. As described in Chapters 3 and 4, when floor and/or window
sill dust lead levels exceed their standards, the assumed intervention is a lead-specific dust cleaning. In
addition, the same type of dust cleaning is assumed to be performed in conjunction with soil interventions
and interior paint abatements (but not paint repair nor exterior paint abatements). These soil and paint
related dust cleanings are performed regardless of the pre-intervention lead levels in the dust. Therefore,
interventions in response to the floor and window sill dust standards occur only in cases where there is no
interior paint abatement nor soil abatements. Under most of the candidate hazard standards analyzed,
there will be a relatively large number of these "stand-alone" dust cleanings.
As shown in Exhibit 4.8 of Chapter 4, the number of homes performing a dust cleaning intervention is
not only relatively large, but also relatively insensitive to the floor dust hazard standards. The number of
homes predicted by the model to perform a dust cleaning declines slowly for floor dust standards between
40 J.lg/ft2 and 90 J.lg/ft2, dipping at a standard of 100 J.lg/fe, and declining very slowly after that.
While the number of homes receiving "stand-alone" dust cleanings is far larger than either the number of
homes receiving paint interventions or soil interventions, dust cleanings are much less costly than either
paint or soil interventions. Thus costs change very little as floor dust standards vary. The table labeled
Exhibit 6.1 presents the costs for a representati ve selection of floor dust standards, assuming that the
paint, window sill dust and soil standards are set at the proposed standards. The two graphs (Exhibits
6.2 and 6.3) present these same costs for a wider range of floor dust standards. As shown in the table,
the costs for all interventions range from about $51.3 billion, when floor dust standards are set at their
least stringent levels, up to about $53.0 billion for the most stringent (40 J.lg/ft2). The most stringent
level is set at the assumed post-intervention lead level, the level of lead assumed to be present in floor
dust after the dust intervention occurs. The least stringent standard shown in the table is at the level
where costs have leveled off, they do not drop below this level within the data set (see Exhibits 6.2 and
!403 EA
6-3
-------
6.3). Note that the highest cost is only about 3 percent larger than the lowest cost. Thus there is little
basis for choosing among floor dust standards on the basis of cost alone.
Exhibit 6.1
Costs, Benefits and Net Benefits for Alternative Floor Dust Standards,
W.th 0 h M d' S t t P d St d d *
I t er e la e a ropose an ar s
IEUBK Model Empirical Model
Floor Dust Total Cost of Total Benefits of Net Benefits of Total Benefits of Net Benefits of
Standards All Interventions Allinterventions Allinterventions Allinterventions Allinterventions
(lJgltf) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
40 53.0 161.5 108.5 42.5 -10.5
60 52.6 158.1 105.5 42.2 -10.4
80 52.4 154.4 102.0 41.9 -10.4
100 51.6 147.6 96.1 40.6 -11.0
120 51.4 128.7 77.3 40.2 -11.2
140 51.4 128.4 77.0 40.1 -11.3
160 51.4 128.4 77.0 40.1 -11.3
180 51.4 128.4 77.0 40.1 -11.3
200 51.3 126.1 74.8 40.0 -11.3
220 51.3 126.1 74.8 40.0 -11.3
* Proposed Standards:
Interior paint: 10 sq ft or more - repair, 50 sq ft or more - abate
Exterior paint: 20 sq ft or more - repair, 100 sq ft or more - abate
Window sill dust: 250 IJg/tf
Soil: 2,000 ppm
Exhibits 6.1, 6.2 and 6.3 also present the estimated benefits and net benefits for alternative floor dust
standards. Total benefits, as estimated using the IEUBK Model, range from about $126.1 billion to
$161.5 billion with increasing stringency in floor dust standards. The highest level of benefits is about
28 percent larger than the lowest benefits, with benefit levels about constant up to a standard of 120
Ilg/fe and increasing with increasing stringency after that. Under this risk model, the benefits are still
increasing much more rapidly than the costs at the most stringent levels. Therefore, net benefits are
maximized at the most stringent floor dust standard of 40 f-lg/ft2. Net benefits at this set of standards
are estimated to be $108.5 billion.
The Empirical Model generates lower estimates of benefits, and the benefit estimates change less with
changes in the floor dust standards. Using the Empirical Model, total benefits range from about $40.0
billion to $42.5 billion with changes in the floor dust standards. The highest level benefits are only about
6 percent higher than the lowest benefit level. The Empirical Model provides lower estimates of benefits,
and the net benefits corresponding with each floor standard are negative (i.e. costs exceed benefits).
Nevertheless, these results can be used to identify the floor dust standard that generates the maximum net
benefits (i.e. the least negative net benefits). Under this risk model, net benefit levels are about constant
up to 100 f-lg/ft2, and increase as the floor dust standard becomes more stringent up to the standards of 80
and 60 Ilg/ft2 Net benefits are nearly identical for these two standards. Net benefits are slightly worse
for the 40 f-lg/fe level.
6-4
~403 EA
-------
Exhibit 6.2: IEUBK-based Model Costs, Benefits, and Net Benefits
o
c
~ 150
i:i5
-------
Exhibit 6.3: Empirical-based Model Costs, Benefits, and Net Benefits
100
~
c
~ 50
as
~
0
-100
::00
Other s1andarcjs:
Soil: 2000 ppm
Sill dust: 250 1J';'11ft'
Paint (jamage
interior:
Maintenance: 10ft'
'Ł!,batemerrt: 50 ft'
exterior:
Maintenance: 20 ft'
,II,I:)atemerrt: 1 00 ft'
- Total Cosh
- T atOll Benefits
- Net Benefits
400
6-6
~403 EA
150
I I I I I I I
I I I I I J I
I I I I I J I
I I I I I I r
I I I I I I I
- - - - - .L - - - - - .J. - - - - - .J - - - - - ..J - - - - - _1- - - - - _1- - - - - - L - - - - -
I I I I I I I
I I I I I I I
I I I I I I I
I I I I I I I
I I J I I I I
,I. - - - - - ), - - - - - _I - - - - - _I - - - - - _,- - - - - _,- - - - - - ,- - - - - -
I t I I , I I
I I I I I J I
I I I I I I I
I J I I I I I
I I I I I t
---1- I I I I I I
T "T - - - - - , - - .., - - - ,- - - - - -,- - - r - - - - -
...............~--...J(' I I I I I
I I I t I J I
I I I I I I I
I I I I I I I
I I I I 1 I I
----_.L_---_.J._---_.J_---_..J______I______I______L____-
I I I I I I I
I I I I I I ,
I I [ r I t I
1 1 I I I r I
I I I I I I I
I I I t I I I
- - - - - "I - - - - - -( - - - - - -I - - - - - -1 - - - - - -I - - - - - -,- - - - - - 1- - - - - -
I I t I I I I
I I J I I I I
I I I I I I I
I I I I I I I
I I I I I I
-50
o
50
"100
150
200
250
300
350
Floor Dust Standard (pg,ft')
-------
Comparing the results of the two blood-lead models, the floor dust standard that maximizes net benefits
(when the other standards are set at their proposed levels) appears to be in the range of
40 - 80 ~g/ft2, with about a 1.3 percent difference in costs over this range.
6.2 Costs, Benefits and Net Benefits for Various Candidate Window Sill Dust Hazard Standards
Dust cleaning interventions occur when either the floor dust or the window sill dust standard is exceeded.
Unlike the floor dust standards, however, the number of homes performing a dust cleaning in response to
the window sill dust standards is quite sensitive to the window sill standards. As shown in Exhibit 4.9,
the number of interventions falls rapidly as window sill dust standards are reduced in stringency from
100 to 220 ~g/ft2 and then falls less rapidly until about 450 ~g/ft2, after which the number remains about
constant.
Because of the relatively low cost of a dust cleaning, the percentage change in total costs of all
interventions is much smaller than the percentage change in the total number of homes receiving a dust
intervention. The table presented in Exhibit 6.4, and the graphs presented in Exhibits 6.5 and 6.6, give
the total costs of all interventions for a range of window sill dust standards, with the other media set at
the proposed standards. As shown in Exhibit 6.4, the costs for all interventions range from about $49.5
billion for the least stringent window sill dust standards up to about $59.8 billion for the most stringent
(1 00 ~g/ft2). The most stringent level is set at the assumed post-intervention lead level, the level of lead
assumed to be present in window sill dust after the dust intervention occurs. The least stringent standard
shown in the table is at the level where costs have leveled off (see Exhibits 6.5 and 6.6). Note that the
highest cost is almost 21 percent larger than the lowest cost. This range in costs is larger than the range
associated with floor dust standards in part because the range of window sill dust lead levels in the data is
much wider than the range of floor dust lead levels in the data. In addition, a much larger proportion of
homes in the data have window sill dust lead levels above the minimum standards than have floor dust
levels above the minimum floor dust standard.
Exhibits 6.4, 6.5 and 6.6 also present the estimated benefits and net benefits corresponding with
alternative window sill dust standards. Total benefits, as estimated using the IEUBK Model, range from
about $150.6 billion to $178.4 billion with changes in window sill dust standards. Benefit levels increase
steadily with increasing stringency, with a jump at the most stringent window sill dust standard. The
highest level of benefits is about 18 percent larger than the lowest benefits. Under this risk model, net
benefits are maximized at the most stringent window sill dust standard of 100 ~g/ft2 Net benefits at this
set of standards are estimated to be $118.5 billion.
The Empirical Model generates lower estimates of benefits than the IEUBK Model. Using the Empirical
Model, total benefits range from about $38.1 billion to $46.3 billion with changes in the window sill dust
standards. As with the IEUBK results, benefits increase fairly steadily with increases in stringency. The
highest level benefits are about 21 percent higher than the lowest benefit level. The Empirical Model
provides lower estimates of benefits, and the net benefits corresponding with each window sill standard
are negative (i.e. costs exceed benefits). Nevertheless, these results can be used to identify the window
sill dust standard that generates the maximum net benefits (i.e. the least negative net benefits). Under
this risk model, net benefits increase as the window sill dust standard becomes more stringent up to the
standard of 310 ~g/ft2 and then decline as window sill dust standards become increasingly stringent.
~403 EA
6-7
-------
The two blood-lead models generate more divergent estimates of the window sill dust standard that
maximizes net benefits (when the other standards are set at their proposed levels) than was true for the
floor dust standards. Nevertheless, the two estimates of 100 f.lg/ft2 and 310 f.lg/ft2 are both toward the
lower end of the entire range of window sill dust lead levels present in the HUD sample of homes. The
majority of homes with window sill dust lead levels above 100 f.lg/ft2 , however, are within this range.
Thus, the selection of a standard within this range can affect a great many homes. This is shown in
Exhibit 4.9 and by the range of costs represented. The cost for a window sill dust standard of 100 f.lg/ft2
is 16 percent higher than the cost at 310 f.lg/ft 2.
Exhibit 6.4
Costs, Benefits and Net Benefits for Alternative Window Sill Dust Standards,
W.th Oth M d' S t t P d S d d *
I er e la e a ropose tan ar s
IEUBK Model Empirical Model
Window Sill Total Cost of Total Benefits of Net Benefits of Total Benefits of Net Benefits of
Dust Standards Allinterventions Allinterventions Allinterventions Allinterventions Allinterventions
(~gItf) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
100 59.8 178.4 118.5 46.3 -13.6
150 57.0 169.6 112.7 45.1 -11.9
200 56.2 168.5 112.3 44.4 -11.8
250 52.8 160.1 107.2 42.4 -10.5
300 52.4 159.7 107.3 41.9 -10.5
310 51.5 158.0 106.5 41.2 -10.3
350 50.9 156.3 105.4 40.4 -10.4
400 50.9 156.3 105.4 40.4 -10.4
450 50.4 154.3 103.9 39.7 -10.7
500 50.3 154.1 103.9 39.5 -10.8
550 49.9 153.5 103.6 39.0 -10.9
600 49.9 153.5 103.6 39.0 -10.9
650 49.7 153.2 103.6 38.5 -11.1
700 49.5 150.6 101.1 38.1 -11.4
750 49.5 150.6 101.1 38.1 -11.4
* Proposed Standards:
Interior paint: 10 sq ft or more - repair, 50 sq ft or more - abate
Exterior paint: 20 sq ft or more - repair, 100 sq ft or more - abate
Floor dust: 50 IJg/ft2
Soil: 2,000 ppm
6-8
~403 EA
-------
Exhibit 6.5: IEUBK-based Model Costs, Benefits, and Net Benefits for Alternative
Window Sill Dust Standards
fII
c
~150
as
.,.
300
Other siandard:;.:
Soil: 2000 ppm
Floor dust 50 I-IglW
Paint damage
irrterior
r'lIalrrtenance: 10ft'
,l!.,baternerrt: 50 ft'
exterior.
Maintemmce: 20 ft'
,!!.,baternerrt: 100 ft'
- Total Benefits'
- Net Benefits:
- T crtal Co:;1:;:
1000
i403 EA
6-9
250
" I
" ,
" I
" I
I I I I
- - - - - - - - _1- - - - - - - - - -1- - - - - - - - - ...J - - - - - - - - - .J - - - - - - - - -
I I I I
I I J I
I I t I
I I t I
I I I I
- - - - - - - - _1- - - - - - - - - _1- - - - - - - - - _1- - - - - - - - - _I - - - - - - - - -
t I r I
I I r I
I I I I
~"~-+" : : :
I I " I I
- - - - - - - - -,- - - - - - - - - -I - - - -, - '---------.,..---
I I
, I
I ,
I I
I
200
'100
------ - I_-
I
,
,
I
--'
.'
- - - - - - - - -1- - -
I
I
I
I
I
- - 1- - -
1
I
I
I
I
I
I
I
I
I
I
I
50
o
o
600
:::00
200
400
Window Sill Standard (pg,oft')
-------
Exhibit 6.6: Empirical-based Model Costs, Benefits, and Net Benefits for Alternative
Window Sill Dust Standards
100
f/I
c:
~ 50
i:i5
-
0
200
IJlher standards:
Soil: 2000 ppm
Floor dus1: 50 j.Jgfft'
Paint damage
irrterior:
Maintenance: 10ft'
,l!.,baternent: 50 ft'
erler"ior:
Maintenance: 20 W
,l!..baternent: 1 00 W
- Total Co:::1::::
- Total Benefits
- Net Benefits:
1000
6-10
~403 EA
150
I I J I
I I I I
I I I I
I I I I
I I I I
- - - - - - - - - ,- - - - - - - - - .J - - - - - - - - - .J. - - - - - - - - - L - - - - - - - - -
I I I I
I I I I
] I I I
I I I I
1 I I I
I I I I
- - - - - - - - -1- - - - - - - - - -, - - - - - - - - - ~I - - - - - - - - - 1- - - - - - - - - -
, I I I
I I I I
I I I I
I I I I
---1--. I I I
- - - - =_::-=-::...:1.. - -,-
I --"-"--"'-"I-....n..._.._~...
I I I I
I I I I
I I I I
I I I I
- - - - - - - - - 1- - - - - - - - - .J - - - - - - - - - .J. - - - - - - - - - L.. - - - - - - - - -
I I I I
I I
I I I I
I 1 I I
I I I I
I I I I
- - - - - - - - -,- - - - - - - - - -I - - - - - - - - - -( - - - - - - - - - ,- - - - - - - - - -
I I I I
I J I I
I I I I
I I I I
I I I 1
-50
-100
o
600
800
200
400
Window Sill Standard (lIglfP)
-------
6.3 Costs, Benefits and Net benefits for Various Candidate Soil Hazard Standards
While floor dust and window sill dust standards are related to each other by the fact that the same
intervention action is taken in response to either one, changes in the floor or window sill dust standards
do not induce changes in the number of paint or soil interventions performed. Soil dust standards,
however, do affect the number of dust interventions performed. As explained in chapter 4, when soil
standards become less stringent, the number of homes performing a soil abatement declines. Some of
those homes, however, will have floor and/or window sill dust lead levels that exceed the dust standards
and so will continue to do a dust cleaning. In other words, with the decline in the soil standards, some
homes will switch from the soil intervention (with dust cleaning) category to the dust only intervention
category. As shown in Exhibit 4.10, when the stringency of the soil standard declines from 150 ppm to
about 900 ppm, the number of homes performing soil interventions declines very rapidly. From there,
the number performing soil interventions declines less rapidly until the standard reaches about 3000 ppm,
at which point the number levels off. Over the entire range, the number of homes performing soil
abatements falls from nearly 20 million to nearly zero. At the same time, assuming the other media are
set at their proposed standards, the total number of homes performing any kind of intervention falls only
until the soil standard reaches about 700 ppm, after which it changes very little. The total number of
homes performing any type of intervention falls from about 28 million to 21 million with changes in the
stringency of the soil standards. These differences between soil interventions and total interventions are
due to the homes that switch from soil to dust only interventions.
Because soil interventions are much more expensive than dust only interventions, costs also fall fairly
rapidly as soil standards become less stringent. The table presented in Exhibit 6.7, and the graphs
presented in Exhibits 6.8 and 6.9, give the total costs of all interventions for a range of soil standards,
with the other media set at the proposed standards. As shown in Exhibit 6.7, the costs for all
interventions range from about $45.6 billion, when soil standards are set at their least stringent level, up
to about $94.3 billion for the most stringent (250 ppm). The most stringent level is set just above the
assumed post-intervention soil lead level of 150 ppm, the level of lead assumed to be in the replacement
soil used in the soil intervention. The least stringent standard shown in the table is at the level where
costs have leveled off (see Exhibits 6.5 and 6.6). Note that the highest cost is over twice the lowest cost.
This range is so large both because soil abatements are relatively expensive and because the number of
homes that exceed the standards increases very rapidly as standards become more stringent than the 1500
ppm level. Note, in particular, that the number of homes that exceed the soil standard doubles as soil
standards increase in stringency from 500 ppm to 150 ppm.
Exhibits 6.7, 6.8 and 6.9 also present the estimated benefits and net benefits for alternative soil
standards. Total benefits, as estimated using the IEUBK Model, range from about $128.1 billion to
$256.0 billion with changes in soil standards. The highest level of benefits is almost twice the lowest
level. Benefits increase slowly with increasing stringency in the soil standards up to about 3,000 ppm.
After that point, benefits increase more rapidly, with a sizeable jump between 1,000 and 500 ppm.
Under this risk model, net benefits are maximized at a very stringent standard; in this case a soil standard
of 250 ppm. Net benefits at this set of standards are estimated to be $161.8 billion.
~403 EA
6-11
-------
Exhibit 6.7
Costs, Benefits and Net Benefits for Alternative Soil Standards,
W' h 0 h M d' S t t P d St d d *
It t er e la e a ropose an ar s
IEUBK Model Empirical Model
Soil Total Cost of Total Benefits of Net Benefits of Total Benefits of Net Benefits of
Standards Allinterventions Allinterventions All Interventions Allinterventions Allinterventions
(ppm) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
250 94.3 256.0 161.8 63.4 -30.8
500 84.0 236.0 152.0 58.4 -25.6
1000 71.1 196.2 125.1 50.5 -20.7
1500 59.4 174.1 114.7 45.0 -14.4
2000 52.8 160.1 107.2 42.4 -10.5
2500 50.4 147.8 97.4 39.6 -10.8
3000 49.4 138.8 89.3 38.2 -11.3
3500 49.4 138.8 89.3 38.2 -11.3
4000 49.4 138.8 89.3 38.2 -11.3
4350 45.8 129.6 83.8 36.7 -9.1
4500 45.8 129.6 83.8 36.7 -9.1
5000 45.6 128.1 82.5 36.5 -9.1
* Proposed Standards:
Interior paint: 10 sq ft or more - repair, 50 sq ft or more - abate
Exterior paint: 20 sq ft or more - repair, 100 sq ft or more - abate
Window sill dust: 250 sq ff
Floor dust: 50 1J9/ft2
As was the case with the other media, the Empirical Model generates lower estimates of benefits. Using
the Empirical Model, total benefits range from about $36.5 billion to $63.4 billion with changes in the
soil standards. The highest level benefits are about 74 percent higher than the lowest benefit level.
Similar to the IEUBK Model results, benefits increase steadily with increases in soil standard stringency,
with an additional jump between 1,000 ppm and 500 ppm. Because the Empirical Model provides lower
estimates of benefits, the net benefits corresponding with each soil standard are negative (i.e. costs
exceed benefits). Nevertheless, these results can be used to identify the soil standard that generates the
maximum net benefits (i.e. the smallest in absolute value terms). Under this risk model, net benefits are
maximized at a soil standard of 4350-4500 ppm. As the soil standards increase in stringency from that
point, net benefits decline somewhat, then increase to a secondary net benefit maximum at a soil standard
of 2000 ppm. At standards more stringent than 1500 ppm, net benefits decline very sharply.
The two blood-lead models generate very divergent estimates of the soil standard that maximizes net
benefits (when the other standards are set at their proposed levels). The two estimates are either a very
stringent soil standard of 250 ppm (using the IEUBK Model) or a not-stringent soil standard of 4350
ppm (using the Empirical Model). The analysis using the Empirical Model indicates that net benefits at a
soil standard of 2000 ppm would be about 15 percent lower than the maximum net benefits.
6-12
~403 EA
-------
Exhibit 6.8: IEUBK-based Model Costs, Benefits, and Net Benefits for Alternative Soil
Standards
o
c:
~150
i:i5
~
300
250
I I I I I I I
I I I I I I I
I I I I I I I
" I I I I I I I
\ 1 I I I I I I
-~___L_____~-----~-----J______I______L_____L____-
"I t I f I I I
'I I I I I I I I
I I I I I I I I
'! I I I I I I I
I I I I I I I
- - - - :., L - - - - - .I. - - - - - _I - - - - - _I - - - - - -'- - - - - - '- - - - - - 1- - - - - -
-,..~> I I I I I I
I I I I I I I
I I I I I I I
I I I I I I I
I ~~----! I I I I I
I ~ I r I I I
- - r - - - - - ~- , - - - - - -, - - - - - -,- - - - - - r - - - - - r - - - - -
I I j I I I I
I I I I \........- I I I
I I t I I t
I I I I I I
I I I I I
~----_J______I___-_-L---__L____-
I I I I I
I I
I I
I I
I I
I I
I 1''='''= - -,- - - - - -,- - - - - - ,- - - - - -
I I
I I
I I
I I
I I
C~rler standar(j:: 0
Floor dus.t 50 IJglft'
Sill dus.t: 250 IJglft'
Paint damage
Irrterior:
Maintenance: 10ft'
,l!.,batemerrt: 50 rt'
e ,terior:
Maintenance: 20 ft'
,l!.,batemerrt: 01 00 rt'
000.000_0"'- Total Benefit::::
- Net Benefits
- Total (:cds
:3000
~403 EA
6-13
200
100
50
---I I
L...-__~-.,-, :
I" I
I ........,,-----\
-----T------,
I I
I I
I I
I I
I I
o
o
7000
1000
2000
3000
4000
5000
6000
Soil Standard (ppm)
-------
Exhibit 6.9: Empirical-based Model Costs, Benefits, and Net Benefits for Alternative Soil
Standards
100
'II
c:
~ 50
as
-
0
-100
200
150
I I I I I J
I I I I I I
I I J I I t
I I I I I J
1 I I I I I I
- - - - - 1. - - - - - l. - - - - - ..1 - - - - - ..1 - - - - - ...J - - - - - ...J - - - - - _1- - - - - -
] I I I I I I
I I I I I I I
t I I I I I I
I I I I I I I I
I, I I I I I I I
~ - - - - .I. - - - - - .I. - - - - - } - - - - - -'. - - - - - _I - - - - - _I - - - - - _1- - - - - -
I,. I I I I I I
\- I I ] I I I
-'1 I I I I I f
"".'-"-~ I I I I r
._-""" I I I I I I
".~'- I -'I--.J I I I I I
- - - :.~-~~~,- - ~. I - - - ~ - - - - - ~ - - - - - -~.- - - - - -
I I I I I I I
I I J I I I I
I I I I I I I
I I J I I I I
- - - - - 1. - - - - - l. - - - - - .J. - - - - - ..1 - - - - - ...J - - - - - ...J - - - - - _1- - - - - -
I I I I
I I I I I
I I I I I
I I I I I I
I I I I I I
1 I I I I I I
- - - - or - - - - - T - - - - - -i - - - - - Ąi - - - - - -,- - - - - -1- - - - - -1- - - - - -
J I I I I I I
J I I I I I I
I I I I I I I
I I I 1 I I J
I I I I I I I
Other standards:
Floor dust: 50 1J9lfP
Sill dust: 250 IJgfft'
Palrrt damage
interior:
Maintenance: 10ft'
,.!I,batement: 50 ft'
exterior:
Mairrtenance: 20 ft'
,8,batemerrt: 1 00 W
- Total "::OS1:5:
- Total Benefits
- Net Benefits
8000
6-14
~403 EA
-50
o
4000
7000
5000
6000
1000
2000
3000
Soil Standard (ppm)
-------
6.4 Hazard Standards that Maximize Net Benefits
The prior sections of this chapter have discussed the standards for each of three media, in each case
assuming that the other media would be held constant at the proposed standards. In addition to floor and
window sill dust and soil, these proposed standards include paint standards of:
Medium Proposed Standard Intervention Activity
Deteriorated Interior Lead- Based Paint 10 sq.ft. or more Repair
50 sq.ft. or more Abate
Deteriorated Exterior Lead- Based Paint: 20 sq.ft. or more Repair
100 sq.ft. or more Abate
The economic analysis did not consider costs and benefits of alternative paint standards in terms of the
amount of deteriorated interior or exterior paint. Data limitations prevented EPA from quantifying the
health risks (and therefore the benefits associated with risk reduction) as an explicit function of the extent
of interior and exterior lead paint deterioration. Data limitations also hindered the estimation of costs as
a function of the extent of deterioration because EP A focused its paint standard definitions on the area of
deterioration on individual components (similar to the 1995 HUD Guidelines) while the available data
from the HUD national survey addresses the amount of deterioration for an entire residence. The
deterioration amounts used in the paint standard values shown in the above table were provided by EP A
to approximate the actual proposed standards.
This final section of Chapter 6 presents the standards that maximize net benefits overall (i.e., lets the
floor dust, window sill dust and soil standards vary independently). These standards are compared to the
proposed standards.
As described in the earlier sections of this chapter, and shown in Exhibit 6.10, the two blood-lead models
generate different benefit estimates for any given combination of standards. In addition, the benefit
estimates change at different rates under the two models and thus the set of standards that maximize net
benefits is different under the two models. The proposed standards fall in between the standards that
maximize net benefits under the two models. The three standards presented are:
Floor Dust Window Sill Dust
Standard Standard Soil Standard
Standards that Maximize Net
Benefits under IEUBK Model 40 ~g/ft2 80 ~g1ft2 250 ppm
Proposed Standards 50 ~g/ft2 250 ~g/ft2 2000 ppm
Standards that Maximize Net
Benefits under Empirical Model 1 00 ~g/ft2 31 0 ~g1ft2 4350 ppm
9403 EA
6-15
-------
For each of the two blood-lead models, Exhibit 6.10 presents the set of standards that maximize net
benefits, along with the costs, benefits and net benefits for that standard. In addition, the exhibit presents
the proposed standard with its cost, benefit and net benefits. The proposed standard is shown twice, once
with benefits calculated using the IEUBK Model and once using the Empirical Model. The top half of
the table presents results using the IEUBK Model. The IEUBK net benefit maximizing standards are
more stringent than the proposed standard. Because of the large number of homes in the lower range of
environmental lead levels, the IEUBK standards would cost nearly twice as much as the proposed
standard and the benefits would be nearly 1.7 times those of the proposed standard. In addition, the net
benefits, at $173 billion, would be substantially higher than the net benefits of the proposed standard, at
$107 billion.
The Empirical Model net benefit maximizing standards, on the other hand, are less stringent than the
proposed standard. There are many fewer homes in these ranges of environmental lead. The Empirical
Model net benefits maximizing set of standards would cost less ($44 billion as compared to $53 billion)
and would produce smaller benefits ($35 billion as compared to $42 billion) than the proposed standard.
However, its net benefits are somewhat larger than those of the proposed standard, while still negative.
Appendix A to this chapter presents the costs, benefits, and net benefits for alternative candidate hazard
standards, when the costs and benefits of paint interventions and testing costs are excluded from the
estimates.
6-16
~403 EA
-------
Exhibit 6.10
C
ompanson of Standards Under Alternative Risk Assessment Models
IEUBK Model Results
Standards that ~
Maximize Net Benefits ~ Proposed Standards
Floor Dust Standard 40 ~g/fe 50 ~g/ft2
Window Sill Dust Standard 1 00 ~g/fe 250 ~g/fe
Soil Standard 250 ppm 2,000 ppm
Total Cost $100.4 billion $52.8 billion
Total Benefit $273.6 billion $160.1 billion
Net Benefit $173.2 billion $107.2 billion
Empirical Model Results
Standards that i
Maximize Net Benefits i Proposed Standards
Floor Dust Standard 80 ~g/ft2 50 ~g/ft2
Window Sill Dust Standard 31 0 ~g/ft2 250 ~g/fe
Soil Standard 4,350 ppm 2,000 ppm
Total Cost $44.0 billion $52.8 billion
Total Benefit $35.1 billion $42.4 billion
Net Benefit -$8.9 billion -$10.5 billion
~403 EA
6-17
-------
Appendix 6A. Net Benefits with Paint Intervention and Testing
Excluded
As explained in the beginning of Chapter 6, the hazard standard candidates cannot be accurately
evaluated one medium at a time. Because of various interaction effects, standards for all of the media
(dust, paint, and soil) must be analyzed jointly. Nevertheless, some readers may be interested in the
contribution to the total costs and total benefits that are made by each of the media separately. While
there are significant limitations to estimating benefits or costs for a single medium, it is possible to
develop approximate values. These estimates, and the limitations, are presented in this appendix.
When trying to exclude paint intervention and testing, estimating the benefits due to dust and soil
interventions is much more problematic than estimating their costs. The benefits are not a simple
summation of the benefits from reductions in each medium, and the benefits are not linear with reductions
in exposure. Benefits are calculated as changes in the population distribution of blood lead levels, and
the "first" reductions provide the largest benefits.
There are two approaches to estimating benefits that exclude the benefits from paint interventions. One
estimates the population benefits assuming no paint intervention occur. This approach tends to
overestimate the benefits and the costs from dust and soil interventions. First, all of the "first reduction"
benefits get assigned to dust and soil. In addition, some homes will now receive a dust only cleaning
(with attendant costs and benefits), which they would not if paint abatements were occurring. The second
approach estimates benefits and costs assuming that paint interventions are the only things occurring, and
then subtract these from the total benefits and costs. This approach tends to overestimate the paint-
related benefits (i.e., paint gets all the "first reduction" benefits) and thus underestimates the dust and soil
benefits. In either approach, the model assumes no testing occurs, so there are no testing costs.
Since net benefits are of primary interest, the results shown in this appendix use the first approach. In
this approach, some of the overestimates in benefits are offset by the overestimates in costs. With testing
costs of about $14 billion (over 50 years discounted at 3 percent) and paint costs of $20 billion (over 50
years discounted at 3 percent), costs drop by $33-$34 billion when testing and paint costs are excluded.2
Exhibit 6.A.1 presents costs, benefits, and net benefits for three of the floor dust standards. Because the
combined costs of testing and paint intervention are so much larger than the paint intervention benefits,
excluding these costs and benefits results in positive net benefits under either blood lead model. Using
either the IEUBK or the Empirical Model to estimate benefits, floor dust standards of 50 flg/ft2 maximize
net benefits.
Exhibit 6.A.2 presents the costs, benefits, and net benefits for four of the windowsill dust standards. As
with Exhibit 6.A.1, all the net benefits are positive. Comparing these results to those presented in
Exhibit 6.4, the results are very similar in terms of which window sill standards yield the maximum net
benefits: 100 flg/ft2 if benefits are estimated using the IEUBK model and 250 flg/fe if benefits are
estimated using the Empirical Model.
They do not appear to drop by the full amount because of the slight increase in number of dust
interventions described above.
6-18
~403 EA
-------
Exhibit 6.A.3 presents the results for alternative soil standards. Again, excluding the costs and benefits
of paint interventions and testing costs results in positive net benefits for each case. Excluding these
costs and benefits tends to emphasize the differences in the two blood lead models. When benefits are
estimated using the IEUBK model, net benefits are maximized when soil standards are at their most
stringent. Alternatively, when benefits are estimated using the Empirical Model, net benefits are
maximized when standards are at their least stringent level.
Exhibit 6.A.1
Costs, Benefits and Net Benefits for Alternative Floor Dust Standards,
With Other Media Set at Proposed Standards. (Testing costs and ~ aint interventions excluded)
IEUBK Model Empirical Model
Floor Dust Total Cost of Total Benefits of : Net Benefits of Total Benefits of : Net Benefits of
Standards All Interventions Allinterventions : Allinterventions Allinterventions : Allinterventions
(IJ g/ff) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
50 19 108 89 39 19
100 18 95 77 37 19
150 18 74 57 36 18
* Proposed Standards:
Window sill dust: 250 sq tf
Soil: 2,000 ppm
Exhibit 6.A.2
Costs, Benefits and Net Benefits for Alternative Window Sill Dust Standards,
With Other Media Set at Proposed Standards. (Testing costs and I aint interventions excluded)
IEUBK Model Empirical Model
Window Sill Dust Total Cost of Total Benefits of ; Net Benefits of Total Benefits of ; Net Benefits of
Standards All Interventions Allinterventions : Allinterventions Allinterventions ; Allinterventions
(lJg/ff) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
100 27 126 100 43 16
250 19 108 89 39 19
500 17 102 85 36 19
1000 15 97 81 33 18
* Proposed Standards:
Floor dust: 50 1J9/fe
Soil: 2,000 ppm
Exhibit 6.A.3
Costs, Benefits and Net Benefits for Alternative Soil Standards,
With Other Media Set at Proposed Standards" (Testing costs and ~ aint interventions excluded)
IEUBK Model Empirical Model
Soil Total Cost of Total Benefits of : Net Benefits of Total Benefits of ; Net Benefits of
Standards All Interventions Allinterventions : Allinterventions Allinterventions : Allinterventions
(ppm) ($ Billion) ($ Billion) ($ Billion) ($ Billion) ($ Billion)
500 50 193 143 55 5
1000 38 150 112 47 9
1500 26 124 98 42 16
2000 19 108 89 39 19
3000 16 86 70 35 19
5000 12 73 61 33 21
* Proposed Standards:
Window sill dust: 250 sq tf
Floor dust: 50 1J9/ft2
!403 EA
6-19
-------
Reference
Battelle. 1997. Risk Assessment to Support Standards for Lead in Paint, Dust, and Soil. Prepared by
Battelle, for National Program Chemicals Division, Office of Pollution Prevention and Toxics,
U.S. Environmental Protection Agency. EPA 747-R-97-006, December.
6-20
9403 EA
-------
7. Sensitivity Analysis
7.1
Introduction
The estimation of the impacts of the ~403 standards presented in the preceding chapters of this report
includes a large number of inputs and assumptions concerning various aspects of both the potential costs
and benefits of these standards. Most of these inputs and assumptions carry with them some degree of
uncertainty. In some cases, the alternative values or approaches to modeling the impacts ofthese rules
could conceivably lead to results that are different from those presented in Chapters 4 through 6. In some
cases it is possible to perform alternative calculations of the impacts using different assumptions to
quantify the magnitude of the difference in outcome. In other cases, it is only possible to address the
uncertainties in a qualitative manner and provide some indication based on judgment of the likely effect
of those uncertainties on the impact estimates.
This chapter focuses primarily on the results of several specific sensitivity analyses that were conducted
to measure the effect of particular aspects of the model on model results. 1 "Sensitivity" is an uncertainty
measure that reflects the rate or degree to which the results of the analysis change relative to changes
made in a particular input variable or assumption. Sensitivity analyses are not necessarily intended to
provide a measure of the uncertainty in the input variable itself (that is, how well that value is known) but
rather to assess how important that variable is with respect to the outcome obtained and by extension
how important it is to have a particular degree of confidence in the value used for that variable in the
main analysis.
As is evident from the description of the benefit-cost model presented previously (as well as the risk
assessment model incorporated into it), there are numerous model elements that could have been selected
for sensitivity analyses. The particular elements of the model chosen to be included in the sensitivity
analysis presented here reflect those identified by EPA as likely to have a significant effect on the results
or for which there was a particular interest in determining what the potential effects might be. Six
particular elements were chosen for these sensitivity analyses2:
.
Discount rate
Monetary value of an IQ point loss/gain
Inclusion of hazardous waste disposal costs for some soil removal
Exclusion of small IQ point changes
Real estate transactions rather than pending birth as the intervention trigger
Considering dust and soil impacts independent of each other and paint impacts
.
.
.
.
.
It should be noted that the use of both the IEUBK and the Empirical models for predicting children's blood
lead levels incorporated into the main part of the benefit -cost analysis is in effect a form of uncertainty
analysis as well.
2
Several sensitivity analyses, mostly relating to the characterization of population blood lead levels, are
included in the ~403 risk assessment document prepared by Battelle (1997).
~403 EA
7-1
-------
The first three elements for which sensitivity analyses were conducted are parametric inputs for which
alternative values are used. The second three elements involve changes in the modeling procedures used
in the main benefit-cost analysis.
The sensitivity analyses consider the effect of changes in these elements on two outcomes of the benefit-
cost modeling. The first is consideration of the effect of alternative specifications on the costs and
benefits of the proposed ~403 standards. The second is consideration of the effect of these alternative
specifications on the determination of the set of ~403 standards that produce maximum net benefits. In
both cases, the analyses are conducted separately using the IEUBK and the Empirical blood lead models.
In addition to these six specific sensitivity analyses, this chapter also provides a more qualitative
summary assessment of the uncertainty in various components of the benefit-cost model and the potential
impact those uncertainties might have in the outcomes of the analysis.
7.2
Analyses Involving Parameter Changes
This section of the sensitivity analysis focuses on three parameters: value of the discount rate, value of
each IQ point, and the cost of disposing of soil removed during soil interventions. Each of these
parameters is of interest for a different reason, but in each case the question is what is the appropriate
value ofthe parameter, not how is it used in the analysis. Because the model estimates costs and benefits
over a 50-year period, and the resulting benefit streams stretch even further into the future, the results
may be very sensitive to the discount rate used in the analysis. The second parameter, value of each IQ
point, is likely to have a substantial impact on the benefits estimation because changes in population IQ
levels account for the great majority of monetary benefits in this analysis (over 98 percent at the option
selected). While the third parameter is not likely to have as large an impact on results, EP A may be
making a change in the hazardous waste disposal regulations that would affect the costs of soil
interventions. The third parameter sensitivity analysis looks at the potential impact of such a change. In
each case, the costs, benefits and net benefits for the proposed standards are used to demonstrate the
impact of the alternative parameter values. Furthermore, net benefit-maximizing standards are shown for
each alternative.
7.2.1 Discount rate
A 3 percent discount rate has been adopted as the most appropriate rate for use in this analysis, based on
a rationale presented in section 3.5.3 of this report. However, OMB frequently recommends the use of a
7 percent discount rate in benefit-cost analyses for government regulations. This section presents results
using 7 percent and compares them against costs, benefits, and net benefits in the baseline (3 percent)
analysis. Exhibit 7-1a gives these results for the proposed standards.
Using a 7 percent discount rate reduces the present value of both total costs and total benefits, with the
reduction in benefits relatively greater than the reduction in costs. This relative difference in declines is
due to the differences in the timing of costs and benefits -- the benefits occur further in the future than
their related costs, thus the higher discount rate has a bigger impact on benefits. The difference is
greatest for soil removals, where the interventions are permanent (i.e. costs incurred today generate
benefits for all future cohorts), and lowest for dust interventions and paint repairs, where the
interventions last only four years (costs incurred today generate benefits for children present over the next
four years).
7-2
~403 EA
-------
Exhibit 7-1a
Effects on Costs and Benefits of Proposed Standards due to Changing Discount Rate
Assumption
Costs
($ billion)
Benefits
($ billion)
Net Benefits
($ billion)
Base
(3% Discount Rate)
~ Empirical IEUBK
$52.8 $52.8
Alternative
(7% Discount Rate)
~ Empirical IEUBK
$34.1 $34.1
$42.4
$160.1
$5.2
$19.2
-$10.5
$1 07.2
-$29
-$14.9
n/a
n/a
Alternative as % of
Base
~ Empirical
65%
IEUBK
65%
12%
12%
The reductions in net benefits under both the IEUBK and Empirical models is a function of the reductions
in costs and benefits under each model. While the relative changes in costs and in benefits are the same
across models, the relative changes in net benefits are different, because of the different magnitudes of
costs and benefits for each model. The negative net benefits under the Empirical Model more than
double in absolute value, while the net benefits under the IEUBK Model shift from strongly positive to
slightly negative.
Exhibit 7-1b compares the net benefit-maximizing standards assuming a 7 percent discount rate, versus
the same for a 3 percent discount rate. Since the analysis identifies standards for each of three media,
using two different risk assessment models, there are six cases to be considered in this comparison. In
five out of six, standards are less stringent under a 7 percent regime; in the other case, standards are the
same between scenarios. This is the expected trend given that when standards are fixed at a constant
level (e.g. the option selected), the cost-to-benefit ratio for interventions is higher with a 7 percent rate
than a 3 percent rate. Interventions will lead to positive net benefits only in homes with very high levels
of contamination, where large improvements in conditions and in expected occupant IQ take place
following intervention. The one standard that does not change between discount rates -- floor dust under
the IEUBK model n is associated with very high positive net benefits to begin with.
Exhibit 7-1b
Effects on Net Benefit-Maximizing Standards due to Changing Discount Rate Assumption
Standards Values ($ billion)
Scenario
Model
Base .,
(3% Discount
Rate) IEUBK
Alternative
(7% Discount
Rate) IEUBK
*Net benefits are maximized when no interventions are triggered through the standard in
question.
Sill
Dust
(~gIft2)
310
100
Floor
Dust
(~g/ft2)
80
40
Soil
(ppm)
4350
250
Empirical
Costs
44.0
100.4
25.4
Empirical
none*
none*
none*
40
4350
27.1
none*
Benefits
35.1
Net
Benefits
-8.9
273.6
3.0
173.2
-22.4
14.8
-12.3
~403 EA
7-3
-------
7.2.2 Value of an IQ Point
The economic analysis presented in chapters 3, 4, 5 and 6, uses the most recent data and approach for
assessing the value of an IQ point. This value, in 1995 dollars, is $8,346 (assuming a 3 percent discount
rate -- see chapter 5). However, earlier EPA analyses have used an alternative, and lower, value of
$6,847, in 1995 dollars. (USEPA 1985, 1986) Exhibit 7-2a compares costs and benefits of the analysis
for these two different IQ point values.
Exhibit 7-2a
Effects on Costs and Benefits of Proposed Standards due to Changing 10 Valuation
Assumption
Base Alternative Alternative as % of
(10 Value = $8,346) (10 Value= $6,847) Base
~ Empirical IEUBK ~ Empirical IEUBK ~ Empirical IEUBK
I I I
Costs $52.8 $52.8 $52.8 $52.8 100% 100%
($ billion)
Benefits $42.4 $160.1 $34.9 $131.8 82% 82%
($ billion)
Net Benefits -$10.5 $107.2 -$17.9 $79.0 n/a 74%
($ billion)
Costs are not affected by a change in IQ point value, because unit costs are not affected nor are the
number or timing of interventions. In turn, the number of children protected and post-intervention blood
lead levels remain unchanged. Benefits, however, are reduced by essentially the same percentage as the
reduction in IQ value ($6,847 is 82 percent of $8,346). While there are several categories of monetized
benefits in this analysis which are not directly linked to changes in IQ, these categories combined make
up under two percent of benefits for the option selected, under both the IEUBK and empirical models.
Thus, the relative reduction in benefits is nearly the same as the relative reduction in IQ point value, for
both models. Net benefits reduce as a function of the change in benefits.
Exhibit 7-2b compares the net benefit-maximizing standards under both IQ point value assumptions.
Standards that maximize net benefits are set at the margin. In other words, given one standard, the
standard which is one unit more stringent is preferable if the additional homes affected by that standard
yield greater marginal benefits than costs.3 In this sensitivity analysis on IQ point value, all marginal
benefits are reduced by a nearly uniform factor, 18 percent, while marginal costs are not affected. As a
result, when the lower IQ point value is used, the net benefit-maximizing standards should be equal to or
less stringent than they are in the base analysis. The results in Exhibit 7-2b match this prediction. Two
out of three standards become slightly less stringent under the Empirical model, and no standards change
Given the structure of the economic analysis, benefits cannot be directly calculated for any specific home
or group of homes smaller than the entire set. Total benefits are calculated based on the blood lead
distribution as aggregated across all homes in the analysis. However, the concept of marginal benefits is
very useful for understanding numerous model results. It can be construed in the following way. The
marginal benefit generated by a new intervention X in a home Y is the total benefits under this scenario,
minus the total benefits under a scenario identical in every way except that intervention X does not take
place in home Y.
7-4
~403 EA
-------
under the IEUBK model. The latter result may appear surprising, but is a reflection of the fact that
marginal homes with relatively low dust lead loads and soil lead concentrations generally experience
strongly positive net benefits when the IEUBK model is used.
Exhibit 7-2b
Effects on Net Benefit-Maximizing Standards due to Changing 10 Valuation Assumption
Standards Values ($ billion)
Scenario Model
: Floor Sill
: Dust Dust Soil Net
: (jJg/ft2) (jJ glft2) (ppm) Costs Benefits Benefits
Base Empirical 80 310 4350 44.0 35.1 -8.9
(10 Value =
$8,346) IEUBK 40 100 250 100.4 273.6 173.2
Alternative Empirical 80 340 4650 43.1 28.1 -15.0
(10 Value=
$6,847) IEUBK 40 100 250 100.4 225.0 124.7
7.2.3 Hazardous Waste Disposal of Soil
This sensitivity analysis assumes that there is no cost premium for the disposal of soil with very high
levels of lead. The base analysis assumes that soil with a lead concentration greater than 2000 ppm must
be disposed of as hazardous waste, at a great supplement to the standard cost of soil disposal (see chapter
4 for itemization of costs). Exhibit 7-3a gives a comparison of results under each scenario.
Exhibit 7-3a
Effects on Costs and Benefits of Proposed Standards due to Changing Assumptions
Regarding Whether Removed Soil Must Ever Be Treated as Hazardous Waste
: Base ~ Alternative :
j (Soil with >2000 ppm i (No Soil Disposed of j
i Lead is Disposed of as i as Hazardous Waste, :
~ Hazardous : thus Reducing Costs:
Waste) of Disposal) :
Alternative as % of
Base
! Empirical IEUBK ~ Empirical IEUBK ~ Empirical IEUBK
I I I
Costs $52.8 $52.8 $48.7 $48.7 92% 92%
($ billion)
Benefits $42.4 $160.1 $42.4 $160.1 100% 100%
($ billion)
Net Benefits -$10.5 $107.2 -$6.3 $111.4 60% 104%
($ billion)
Since this change in assumption does not affect the number of homes getting interventions nor the
effectiveness of those interventions, benefits remain constant. Total costs decrease, however, because
costs for soil removal decrease. For the option selected, with a standard of 2000 ppm for soil removal,
~403 EA
7-5
-------
all two mil1ion homes perfonning soil interventions are affected by the relaxed soil disposal requirements
of the sensitivity analysis and costs decline by $4 million national1y.
Exhibit 7-3b compares the net benefit-maximizing standards assuming no hazardous waste disposal of
soil, versus the base analysis. Since, relative to the base case, some unit soil costs decrease, and benefits
are unaffected at each standard, the expected consequence of this exercise is that soil standards should
become more stringent if they change at al1. Analysis with the Empirical model yields this result;
however, the IEUBK model produces the opposite pattern.
Exhibit 7-3b
Effects on Net Benefit-Maximizing Standards due to Changing Assumptions Regarding
Whether Removed Soil Must Ever Be Treated as Hazardous Waste
Standards Values ($ billion)
Scenario Model
! Floor Sill
! Dust Dust Soil Net
~ (1J9/ft2) (lJg/ft2) (ppm) Costs Benefits Benefits
Base Empirical 80 310 4350 44.0 35.1 -8.9
(as 7-3a) IEUBK 40 100 250 100.4 273.6 173.2
Alternative Empirical 80 310 1650 47.2 41.2 -6.0
(as 7-3a) IEUBK 40 100 300 88.1 266.3 178.2
The IEUBK result suggests that when no soil is treated as hazardous waste, soil removal at homes where
the yard average concentration is between 250 ppm and 300 ppm produces negative net benefits on the
whole. Similarly, when soil above 2000 ppm is treated as hazardous waste, it would appear that this
col1ection of homes produces positive net benefits from soil removal. However, these two statements
cannot both be true, because no home in the HUD survey with a yard average soil lead concentration in
the 250-300 ppm range contains any soil above 2000 ppm. Thus neither costs nor benefits stemming
from this group of homes should be affected by the changed assumption in this analysis. This case is a
special exception to the rule that net benefit-maximizing standards are set at the margin.
The reason for the paradoxical finding under the IEUBK model has to do with the possibility of mixing
soil to avoid hazardous waste disposal costs (in the case that there is a definition for soil as hazardous
waste). As described in chapter 4 (section 6.3), special disposal of soil removed from a yard is not
required in two cases. These are that either none of the soil removed exceeds the hazardous waste
definition, or the following conditions are met:
.
Soil is removed from both the home perimeter and remote areas;
Soil from one of the two areas exceeds the hazardous waste definition; and
Mixed together, the soil removed from both areas does not exceed the hazardous waste
definition.
.
.
Based on the HUD survey, there are 1.1 million homes where one area of the yard exceeds 2000 ppm, but
the other falls between 250 and 300 ppm. They all exceed the basic soil standards in question. Of these,
931 thousand perform interventions during the course of the 50-year analysis. This means that if the
7-6
i403 EA
-------
hazardous waste definition of soil is 2000 ppm, these homes must incur the major expense of hazardous
waste disposal when the soil standard is 300 ppm. However, when the standard is 250 ppm, soil is
removed from both yard areas and may be mixed to avoid this extra cost. The supplemental cost of
removing a greater volume of soil is small in comparison; overall savings for this set of homes is $2.6
billion.
Thus it appears that in the base analysis, the net benefit connected with removing soil from homes with
soil lead concentrations between 250 and 300 ppm is somewhat negative. There are 2.3 million homes in
this category, of which 2.0 million would perform soil interventions at a cost of $7.7 billion. The benefit
associated with these interventions is, in fact, slightly less -- $7.2 billion (as calculated by the technique
described in footnote 3). However, this deficit is more than offset by the $2.6 billion reduction in the cost
of soil removal in 931 thousand homes, precipitated by changing the soil standard from 300 ppm to 250
ppm. These interrelations explain why net benefits are maximized at a standard of 250 ppm in the base
analysis, but at 300 ppm when there is no definition for hazardous waste disposal of soil.
7.3
Analyses Involving Changes in Modeling Procedure
In addition to alternative assumptions about certain parameter values, three sets of sensitivity analyses
were performed to investigate the impact of making certain structural changes in the model. Each of
these changes addresses an important assumption in the analysis. The first question is how much of the
benefits are the result of very small changes in IQ levels. While the value assigned to a given change in
IQ can be altered simply by changing a single parameter, as in 7.2.2, eliminating small IQ changes from
the estimation of benefits requires a basic restructuring in methodology. The second structural sensitivity
analysis proposes an alternative "trigger" for intervention events. The model used in the analysis
presented up to this point assumes that any and all interventions needed to bring the housing unit into
compliance with the standards occurs just before the arrival of a newborn child in that unit. There is
evidence, however, that interventions do not necessarily occur then, and do occur at other times. In
particular, another type of event that frequently triggers interventions is property transaction. A
"transaction trigger" model was constructed, therefore, to compare to the "birth trigger" model used in
the baseline analysis. The third set of structural sensitivity analyses attempt to investigate each medium
one by one, completely independent of the standards for the other media.
7.3.1 Benefits from Small IQ Changes
The core analysis assumes that a difference in average blood lead levels between two populations, no
matter how small that difference is and regardless of the magnitude of blood lead levels involved, is
associated with a corresponding difference in average IQ scores. The cost-benefit analysis performed for
these standards is essentially a comparison of the blood-lead distributions that would occur between two
populations: one with the 403 standards versus one without the 403 standards. Furthermore, the
analysis relies on the empirical finding that a difference in average IQ scores between two populations,
again no matter how small, is associated with a difference in average lifetime earnings. Note that it is not
possible to say that for any pair of individuals that a difference in blood lead will necessarily reflect a
difference in IQ scores or lifetime earnings. The available research, however, does demonstrate that such
differences do occur on the average for groups of individuals.
Notwithstanding the fact that the risk assessment and benefit-cost analysis were constrained to address
population average changes, it was recognized that there might be an interest in considering the
~403 EA
7-7
-------
contribution to those population average changes made by subgroups in the population whose particular
blood lead and IQ point improvements might be considered small.
This analysis poses special problems for procedure. In the normal calculation of benefits, a blood lead
distribution for each entire cohort born is calculated under baseline and post-intervention scenarios.
Most children are born into homes that meet standards and thus experience no interventions. These
children have the same blood lead levels between scenarios. Since this group is included in the
calculation of aggregate blood lead distributions, the difference between baseline and post-intervention
mean blood lead levels for each cohort tends to be quite small. It is always much smaller than the
average blood lead change for children in homes where interventions do take place. Therefore, the blood
lead-difference "screen" cannot be applied to the population average difference.
One way around this would be to break up the population distribution into one hundred percentiles, and
then take the difference between matching percentiles for baseline and post-intervention scenarios. These
differences could then be compared against the screen. The problem with this approach is that children
do not remain in the same percentile groups between scenarios, so the differences are not meaningful.
Children in lead-contaminated home types will be in the upper percentiles of the baseline scenario, but in
the post-intervention scenario, due to the effectiveness of interventions, they may exhibit lower blood
lead levels than children in home types that never exceeded any standards and performed no
interventions.
The approach adopted was to consider each home type individually, and split its baseline and post-
intervention blood lead distribution into one hundred percentiles.4 It is more reasonable to assume these
percentile groups stay in order. The blood lead changes in each percentile group were "scaled" so that in
the aggregate (that is, across all percentiles and all housing groups) the overall average blood lead change
matched the average changed obtained in the baseline analysis for the aggregate population. Then, a
"screen" was applied to these scaled blood changes observed in each percentile group so that only those
percentile groups where a blood lead change of 3.89 flg/dL were used to estimate the IQ point
improvement benefits.
The final results of this sensitivity analysis are presented in Exhibit 7-4, alongside the core analysis
results. Costs are not affected by the IQ point difference screen, because unit costs are not affected, and
neither are the number or timing of interventions. Benefits are reduced, as a natural consequence of the
fact that the screen excludes benefits from many children with small changes in IQ point reduction, who
contribute to the core analysis benefits total.
In homes with damaged lead-based paint, children are divided into separate groups according to the
presence and extent of pica behavior exhibited. These pica groups are then each split into one hundred
percentiles. Otherwise, children would not plausibly remain in the same percentile group from baseline to
post-intervention: children exhibiting pica would move from the highest blood lead percentiles to the
middle of the range, after paint ingestion exposures were eliminated.
7-8
~403 EA
-------
Exhibit 7-4
Effects on Costs and Benefits of Not Counting Benefits
from Individual 10 Point Changes of Less than One
Base (No 10 Screen) 10 Difference Screen
: Empirical IEUBK: Empirical IEUBK
$52.8 $52.8 $52.8 $52.8
% of Base
Costs
($ billion)
Benefits
($ billion)
Net Benefits
($ billion)
: Empirical
100%
IEUBK
100%
$42.4
$160.1
$5.9
$151.7
12%
70%
-$10.5
$107.2
-$46.9
$98.9
n/a
56%
Another striking feature of the results is that a substantially greater portion of benefits is lost when the
empirical model is used, as opposed to the IEUBK model. This is because given the same interventions,
the changes in blood lead distributions for each home type are much greater under the IEUBK than the
empirical model (see chapter 6 and Battelle (1997)). Thus, the Empirical model generates a much higher
proportion of cases of small blood lead or IQ point changes that do not exceed the screen, and are not
counted.
To clarify why this difference between models exists, it is helpful to take the example of an actual home
type from the HUD dataset -- for example, home ID number 411207, which performs a dust intervention
only. Using the IEUBK model, the baseline blood lead geometric mean for children living in this home
type is 15.83 Ilg/dL, and the geometric standard deviation is 1.6. The post-intervention figures are 10.03
Ilg/dL and 1.6. The analysis assumes that these distributions can each be approximately scaled to
become compatible with NHANES data, and divided into one hundred percentiles. These percentiles
represent small groups of children with identical blood lead levels, and the percentile groups are assumed
to stay in the same rank order from the baseline to the post-intervention scenario. Values for a small
number of the percentile groups are shown in Exhibit 7-5. Thus, the baseline/post-intervention
difference for the first percentile is 1.05 Ilg/dL, and for the 100th percentile is 12.89 Ilg/dL. Children
represented by the first percentile through the 56th percentile are excluded from IQ-related benefits,
because their blood lead change falls under the screen of 3.89 Ilg/dL.
Using the empirical model, the blood lead figures are much smaller. In the baseline, the geometric mean
is 4.47 Ilg/dL and the geometric standard deviation is 1.6. The geometric mean reduces to 4.06Ilg/dL in
the post-intervention scenario. Percentile breakdowns are given in Exhibit 7-5. Children represented by
all percentiles are excluded from IQ-related benefits -- a much greater portion than was the case under the
IEUBK model.
~403 EA
7-9
-------
Exhibit 7-5
Baseline/Post-Intervention Blood Lead Difference by
Percentile Group for HUD Home 411207
Difference Between Baseline and Post-
Intervention Blood Lead Levels (lJg/dL)
Percentile
1
25
50
56
57
75
100
IEUBK Model
1.05
2.63
3.62
3.886
3.93
4.96
12.89
Empirical Model
0.15
0.37
0.51
0.55
0.55
0.70
1.81
New net benefit-maximizing standards were not determined in this analysis due to its high degree of
complexity and computational intensiveness. However, based on results at the option selected,
qualitative predictions are possible. Under the Empirical model, eliminating small IQ changes reduces
benefits as sharply as changing the discount rate to 7 percent (by 88 percent). The discount rate change
also reduces costs, but the small IQ adjustments do not. It is reasonable to expect, therefore, that when
small IQ benefits are not counted, the net benefit-maximizing standards under the Empirical model
should be less stringent than they are in the 7 percent discount rate analysis. In other words, the only
medium where any standard at all may be favorable is window sill dust, and that standard is likely to be
very lenient.
Under the IEUBK model, the reduction in benefits is not nearly so great when small IQ changes are
screened out (by 30 percent). It is moderately larger than the benefits reduction when a smaller IQ value
is used (section 7.2.2 -- by 18 percent). Neither sensitivity analysis affects costs. Thus it is reasonable to
expect that when small IQ benefits are not counted, the net benefit-maximizing standards under the
IEUBK model should be somewhat less stringent than they are in the low IQ value analysis, where they
do not change at all. In other words, there should be little or no change from the optimal standards in the
baseline analysis.
7.3.2 Transaction Trigger for Interventions
The base analysis assumes that the birth of a child triggers interventions in homes that exceed section
403 standards. This assumption results in maximum efficiency: each intervention performed is matched
with a child to benefit from it, and is implemented at the last possible moment for maximum overlap with
the child's development, and for least present value of cost (due to discounting). Furthermore, all
children born into homes exceeding standards are protected. If a child under six is still present when the
effectiveness of an intervention lapses, the intervention is repeated.
An alternative way to imagine response to section 403 standards is that interventions will be performed at
times of real estate transaction. These may be particularly convenient times to intervene because homes
are likely to be unoccupied, and other renovations may be taking place as well. Additionally, Section
1018 of Title X, "Disclosure of Information Concerning Lead upon Transfer of Residential Property,"
7-10
S403 EA
-------
provides a new incentive for lead abatement during real estate transaction. Section 1018 requires that
home sellers or lessors must tell home buyers or renters everything already known about the presence of
lead in the home. The future occupant must also be granted ten days to conduct a risk assessment or
inspection. Thus, it is reasonable to imagine that all parties will be aware of possible lead risks at the
time oftransaction, and may be likely to perform interventions to increase home saleability or safety.
Like the birth trigger model, the transaction trigger model operates on the assumption that if a child under
six is present when the effectiveness of an intervention lapses, the intervention is repeated.
Exhibit 7-6a compares results between the birth trigger model (the base analysis) and the transaction
trigger model (the sensitivity analysis). Costs are greater in the transaction trigger model, and benefits
are lower, using either the IEUBK or empirical model for predicting blood lead levels.
Costs ($bil)
Benefits ($bil)
Net Benefits
Exhibit 7-6a
Effects on Costs and Benefits due to Changing Assumption about Intervention Trigger
Base (Birth Trigger) Transaction Trigger % of Base
i Empirical IEUBK j Empirical IEUBK! Empirical IEUBK
$52.8 $52.8 $74.8 $74.8 141% 141%
$42.4 $160.1 $19.7 $81.2 46% 51%
-$10.5 $107.2 -$55.1 $6.3 n/a 6%
Costs increase because interventions occur at a faster rate in the transaction trigger model than in the
birth trigger model. Under the transaction trigger, they occur whenever a property changes hands or the
tenant moves out: 8.15 percent a year for single-family homes and 28.45 percent a year for multi-family
housing units (USDOC and HUD 1989). By contrast, the birth rate is projected to be less than four
percent per household every year of the model run (Battelle 1996). As a result, more total interventions
take place in the transaction trigger model (by 19 percent), and they are more crowded toward the early
years. They therefore receive little discounting compared to the more spread-out costs of the birth trigger
model.
At the same time, benefits are lower under the transaction trigger model than the birth trigger model.
This is because, with the former, many children are born into homes which exceed section 403 standards,
but which have not had a recent transaction. These children receive no benefits, whereas they would be
protected in the birth trigger model, in which all children born into homes exceeding standards receive
protection.
Exhibit 7-6b compares the net benefit-maximizing standards assuming a transaction trigger, versus the
same for the birth trigger. For the Empirical model, standards are considerably less stringent across the
board with the transaction trigger. This is the expected pattern, since for any set of interventions at any
set of homes, costs are higher and benefits lower assuming a transaction trigger. For the IEUBK model,
the window sill dust standard is less stringent, while the floor dust and soil standards remain constant.
Marginal net benefits in the latter two standards are strongly positive under the birth trigger, large enough
to remain positive despite the switch to a transaction trigger.
~403 EA
7-11
-------
Exhibit 7-6b
Effects on Net Benefit-Maximizing Standards due to Changing Assumption about
Intervention Trigger
Standards Values ($ billion)
Scenario Model
~ Floor Sill
! Dust Dust Soil Net
i (~glft2) (~g/ft2) (ppm) Costs Benefits Benefits
Base (Birth Empirical 80 310 4350 44.0 35.1 -8.9
Trigger) IEUBK 40 100 250 100.4 273.6 173.2
Alternative Empirical 130 900 4650 56.5 11.5 -45.0
(Transaction
Trigger) IEUBK 40 650 250 127.6 163.5 35.9
*Net benefits are maximized when no interventions are triggered through the standard in
question.
7.3.3 Single Medium Analysis
This section presents an alternative method of determining which standards among many may maximize
net benefits -- a method that is different than the technique used in chapter 6. In both chapter 6 and this
analysis, the standards for paint remain fixed, but standards for lead content in floor dust, window sill
dust, and soil may vary.
In chapter 6, alternative standards for each medium are explored independently. However, while one
medium's standard is varied, the other media also have standards in effect, which go into model
calculations. This sensitivity analysis explores what happens when the standard for one medium is
varied, but no other standards are in effect. In other words, no interventions take place except for those
triggered by the single medium standard being considered. Exhibit 7 - 7 summarizes the results.
Exhibit 7-7
Net Benefit-Maximizing Standards in a Single Medium Analysis
Model
Medium
Net Benefit-Maximizing Standard
Floor Dust
Base Analysis* Single Medium
Analysis
40 ~g/ft2 40 ~g/ft2
1 00 ~ g/ft2 1 00 ~g/ft2
250 ppm 250 ppm
80 ~ g/ft2 40 ~g/ft2
31 0 ~g/ft2 31 0 ~g/ft2
4350 ppm 1650 ppm
IEUBK
Window Sill Dust
Soil
Floor Dust
Empirical
Window Sill Dust
Soil
7-12
~403 EA
-------
As a general rule, net benefit-maximizing standards for the single medium analysis are expected to be
equally or more stringent than standards in the base analysis. This is because of partial redundancy
between different standards or intervention types. When there is no window sill dust standard, for
instance, under the Empirical model, the optimal floor dust standard may become more stringent if
moderately contaminated floors are associated with highly contaminated window sills in some homes.
Empirical model results
This is, in fact, the case. The optimal floor dust standard drops from 80 to 40 ~g/ft2 between the base
and single medium analyses under the empirical model. Based on the HUD dataset, 3.8 million homes
have floor dust lead loads between 40 and 80 ~g/ft2. Of these, 2.5 million (65.2 percent) have window
sill dust lead loads over 250 ~g/ft2. This set (set A) receives dust cleanings under both scenarios, the
base analysis (where the floor dust standard is 80 ~g/ft2 and the sill dust standard is 250 ~g/ft2), and the
single medium analysis (where the floor dust standard is 40 ~g/ft2). For set A, the marginal benefits of
cleaning outweigh the marginal costs. However, for the other homes with floor dust lead levels between
40 and 80 ~g/ft2, that have low sill dust lead levels (the remaining 34.8 percent n set B), the marginal
costs of cleaning outweigh the benefits. This is why the net benefit-maximizing floor dust standard in the
base analysis is 80 ~glft2. The optimal standard in the single medium analysis is 40 ~g/ft2 because the
positive marginal net benefits from set A outweigh the negative marginal net benefits from set B. In
other words, window sill dust contamination drives the choice between floor dust standards in the two
scenarios described. There is no interaction with soil contamination, and little with damaged lead-based
paint, in the homes considered here.
Finally, the difference between net benefit-maximizing soil standards under the Empirical model stems
from the fact that 80.1 percent of the 2.1 million homes with soil lead concentrations between 1650 ppm
and 4350 ppm have higWy contaminated window sill dust, with lead loads well above 250 ~g/ft2. Soil
removals are accompanied by dust cleanings, which result in reduced dust lead loads and concentrations.
The soil-related benefits for homes in the 1650-4350 ppm range do not match the costs of soil removal,
but when dust-related benefits are added in, marginal net benefits are positive for the group. Dust
benefits are added at the margin in the single medium analysis, and that is why the optimum soil standard
is 1650 ppm in that analysis. In the base analysis, however, the window sill standard is 250 ~g/ft2, so
dust cleaning takes place in the group of homes characterized by 1650-4350 ppm soil lead and high
window sill lead, even in the absence of removing soil. Since dust cleaning is less expensive than soil
removal, it is more efficient in a multimedia scenario for the soil standard to remain high, at 4350 ppm,
and for the dust standards to trigger dust cleaning in the homes that need it.
IEUBK model results
In contrast to the Empirical model results, all net benefit-maximizing standards remain the same under
the IEUBK model, whether they are calculated using the base methodology or the single medium analysis.
The reason the single medium approach does not make either dust standard more stringent is elementary:
each is already at its most stringent possible value based on the baseline methodology. 40 ~glft2 is the
minimum allowed standard for floor dust because it is the assumed post-intervention floor dust lead load
in the risk assessment. Similarly, 100 ~g/ft2 is the assumed post-intervention window sill dust lead load
(Battelle 1997).
~403 EA
7-13
-------
The soil standard, 250 ppm, is also very close to its minimum value, 150 ppm, the assumed lead
concentration in replacement soil (Battelle 1997). However, some further explanation can be offered as
to why it does not drop further in the single medium analysis.
In homes with soil lead concentrations in the vicinity of 250 ppm, very little benefit can be realized
directly from soil removal. However, substantial benefits may accrue following the associated dust
cleaning, from reductions in dust lead concentrations. The average dust lead concentration for HUD
survey homes with soil lead concentrations between 250 and 300 ppm is 562 ppm. This high value helps
to explain why the net benefit-maximizing soil standard is so low in the baseline analysis. However, for
homes with soil between 200 and 250 ppm, the dust average is only 213 ppm, and it is just slightly
greater for homes with soil between 150 and 200 ppm. Therefore, it is not surprising that the single
medium analysis does not generate a lower soil standard than 250 ppm.
Finally, lowering the soil standard may be beneficial because it can lead to soil mixing and the
elimination of soil hazardous waste disposal costs. This kind of cost reduction is substantial when the
soil standard changes from 300 ppm to 250 ppm (see section 7.2.3); however, no further such
advantages accrue when the soil standard drops further, as far as 150 ppm. Additionally, any advantages
which might have existed, would have been equal between the single medium and baseline analysis. In
sum, there is no reason why the soil standard should drop any lower than 250 ppm in the single medium
analysis.
Combining single medium analyses
This section has focussed on the effect that a single medium analysis has on net benefit-maximizing
standards chosen. Costs and benefits have not been presented because they may be misleading. They
cannot be combined across media in most cases. For instance, many homes which incur costs for
repeated dust interventions in a single medium analysis of floor dust, may not incur these costs in a
multimedia situation because a soil removal takes place. Also, benefits cannot be added among different
analyses, because they are generated based on population-wide blood lead distributions calculated from
conditions in all homes. Even if there were no overlap in which home types receive interventions in
different single medium analyses, it would not be appropriate to add benefits across analyses because
benefits cannot be directly assigned to specific homes.
7.4
Additional Elements of Uncertainty
This section presents a qualitative assessment of additional elements of uncertainty associated with
inputs to the economic analysis of ~403. In many cases, the analysis is limited to a qualitative
assessment because data are not available on which to base a quantitative sensitivity analysis. In other
cases, the complexity of the analysis precludes it being undertaken at this time. The inputs investigated
in this qualitative assessment include the unit costs of interventions, the valuation of different types of
benefits, and the design of the overall analysis. These were selected because they are unique to the
benefit-cost analysis, whereas other uncertainties stem from the risk assessment. In this section, first the
sources of uncertainty are presented, and then an assessment of their likely impact on the estimations of
costs, benefits, net benefits, and net benefit-maximizing standards are discussed.
7-14
~403 EA
-------
7.4.1 Unit Costs of Interventions
Sources of Uncertainty
There are three basic sources of uncertainty with regards to the unit costs of interventions. First,
inaccurate estimation of any of the inputs used to develop unit costs, as described in chapter 4, would
lead to the underestimation or overestimation of those costs. There is one special case where the bias is
known. The unit costs for paint abatement used in the analysis do not reflect the possibility that home
occupants may need to move out temporarily during intervention. This would result in increased costs.
Based on the HUD survey data, however, very few homes require paint abatements. Therefore, this
source of uncertainty may not have a significant effect on total rule costs even if temporary relocation of
families proves to be common during paint abatement.
Second, it is not known how future economic forces will affect unit costs. Section 403 standards are
likely to result in an increased demand for intervention services: will this drive their prices up, or lead to
innovation and cost reduction?
Third, it may not be appropriate to assign average unit costs to all single family, or multi-family homes.
The unit costs for soil removal were calculated to reflect the fact that soil contamination is systematically
and positively associated with smaller yards. However, no such adjustments were made with regards to
paint or dust interventions. For instance, if dust contamination is associated primarily with the oldest
homes, and if very old homes are typically larger than newer pre-78 ones, then the dust intervention unit
cost should reflect the need to clean a larger home than the national average size, or different unit costs
should be assigned to homes in different age classes. A similar situation may arise if the geographic
regions where homes are most likely to exceed the standards are also the regions with the highest
intervention costs.
Effect of Uncertainty on Benefits Estimates
For any given set of standards -- for example, the option selected -- changes in unit costs of intervention
will have no effect on benefits estimates.
Effect of Uncertainty on Cost Estimates
For any given set of standards, changes in unit costs of intervention will have simple, predictable effects
on total cost estimates. For example, if the estimated unit cost of exterior paint maintenance were to be
raised by 30 percent, then the portion of the total estimated rule cost associated with exterior paint
maintenance would increase by 30 percent. The relative increase of total costs would depend on the
portion of total costs made up by exterior paint maintenance costs.
Effect of Uncertainty on Net Benefits and Net Benefit-Maximizing Standards
For any given set of standards, the effect of unit cost uncertainty on net benefits will be a direct function
of its effect on cost estimates. Net benefits will decline by the amount that costs increase.
The effect on net benefit-maximizing standards is more difficult to predict. To the extent that the unit
cost for an intervention type increases, however, the net benefit-maximizing standard for that intervention
type will tend to become more lenient; to the extent that the unit cost for an intervention type declines, the
related standard will tend to become more stringent. This is because net benefit maximizing standards
are set at the margin, with the cost of each marginal intervention compared against the associated benefit,
which is unaffected by any uncertainty in the costs. Uncertainty in unit costs of paint intervention will
~403 EA
7-15
-------
have no effect on standards that maximize net benefits in this analysis because alternative paint standards
are not considered.
7.4.2 Valuation of Benefits
Sources of Uncertainty
There are two basic sources of uncertainty in the monetary values associated with reduced incidence of
adverse health effects considered in this analysis. First, inaccurate estimation of any of the inputs used to
develop these values, as described in chapter 5, would lead to their incorrect estimation. Second, it is
impossible to know how future economic forces will affect the inputs. For instance, will expected
lifetime earnings change in the future? This would affect the valuation of an IQ point.
In addition, the benefits are underestimated because certain benefits categories are not included in this
analysis, such as benefits to children age six and older, to other children spending time at homes with
interventions, to adults, and to ecosystem health. The size of these excluded benefits, however, is not
known so the degree of underestimation is uncertain.
Effect of Uncertainty on Benefits Estimates
The potential effects of changes in the valuation of IQ have already been discussed in depth earlier in this
chapter. Anything less than a major change in the other values used to calculate benefits -- the expenses
assigned to special or remedial education, or to medical treatment -- should have a very small effect on
benefits estimates. This is because only a small fraction of the population receives special education or
medical intervention, and therefore benefits from reduction in the associated expenses, between scenarios,
are low. At the option selected, these benefits account for under two percent of total monetized benefits.
The addition of benefits not previously counted in the analysis would clearly have the effect of increasing
total benefits estimates, potentially by a significant amount, because the population age six and older is
much larger than the population under six. However, per-individual damage from lead exposure is
believed to be the greatest in young children, whose nervous system is still developing. (Battelle 1997)
Effect of Uncertainty on Cost Estimates
For any given set of standards, changes in the valuation of benefits will have no effect on cost estimates.
Effect of Uncertainty on Net Benefits and Net Benefit-Maximizing Standards
For any given set of standards, the effect of uncertainty in benefits valuation on net benefits will be a
direct function of its effect on total benefits estimates. Thus, net benefits are not likely to change
significantly due to changes in the cost of special education or medical intervention; however, they could
increase substantially if new benefits categories were to be added.
Likewise, the effect of uncertainty on net benefit-maximizing standards is likely to be negligible with
regards to education and medical costs, but could be significant with regards to new benefits categories.
In the latter case, increased benefits may lead to more stringent standards for dust and soil.
7.4.3 Other Modeling Issues
Several analytic components in addition to cost and benefit inputs contain important elements of
uncertainty. These uncertainties include the appropriate time frarne of analysis, the probability of future
interventions taking place even in the absence of national standards for household lead hazards, and the
7-16
i403 EA
-------
likely rate of interventions after standards are issued. Each area of uncertainty is addressed briefly in
turn.
Appropriate Time Frame of Analysis
The economic analysis considers costs and benefits relating to cohorts of children born over the fifty year
period, 1997 to 2046. This choice of time frame has effects on total costs, benefits, and net benefits, and
possibly also on the standards that maximize net benefits. A shorter time frame would result in smaller
absolute magnitudes of costs, benefits, and net benefits, because fewer interventions would take place,
and fewer cohorts of children would benefit from them. Similarly, a longer time frame would lead to
greater magnitudes.
The effect of time frame changes on net benefit-maximizing standards is less clear. When the analysis
period changes, even though cost and benefit totals move in the same direction, their ratio changes. The
longer the frame of analysis, the greater the ratio of benefits to the total cost of interventions. This is in
large part because soil removals are assumed to have permanent effectiveness. Thus, one intervention
paid for and performed in 1997 conveys benefits to a child born at the same home in 2040. Under such
circumstances, the longer the period considered, the better the investment in soil removal appears.
Longer time frames, then, will tend to favor more stringent soil standards as the standards that maximize
net benefits; and shorter time frames will lead to less stringent soil standards. Dust standards may move
in the opposite direction, because dust interventions have short durations, and soil interventions include
dust cleanings (thus preempting dust interventions that would have been triggered by dust standards).
For example, if soil standards become less stringent, then homes that no longer perform soil interventions
no longer receive the benefits from associated dust cleaning. This may result in an increase in the
stringency of dust standards to capture dust-related benefits from these same homes.
Using a long time frame for the economic analysis has the advantage of favoring an appropriate balance
between short-term and permanent interventions. However, it also carries the increased general
uncertainty which comes with projections made far into the future.
Interventions without Section 403
The baseline for the economic analysis assumes that no interventions will take place in the absence of
national standards. This assumption makes possible its corollary, that the NHANES III Phase 2 national
blood lead distribution will remain constant over the entire analysis duration in the baseline scenario of
"no action."s In turn, this corollary is critical to the methodology used for projecting future national
blood lead distributions in scenarios with interventions, and thus, for calculating benefits.
However, interventions to remove lead hazards are already taking place in the pre-standards world. How
would model results change if some interventions were included as part of the baseline scenario?
Because fewer interventions would be occurring as a result of 403, the costs, benefits, and net benefits
would decrease.
The distribution will remain constant within the housing stock considered in the analysis: homes built
before 1997.
~403 EA
7-17
-------
The effect that baseline interventions might have on net benefit-maximizing standards is not as clear,
especially considering the limited information available on where interventions currently do take place.
However, there is reason to believe that the effect should be small or none. Standards that maximize net
benefits are set at the margin. Thus, unless baseline interventions are disproportionately concentrated
among homes with contamination levels near the current net benefit-maximizing standards, these
standards should not be perturbed.
Intervention Rates
Two different triggers for interventions in homes that exceed standards have been presented -- births or
real estate transactions. As the model is constructed, however, each trigger always leads to intervention
in a home exceeding standards at the time of the trigger event. This results in modeled national
intervention rates which are substantially greater than current known regional rates, a discrepancy which
does not appear realistic even in the aftermath of the issuance of national standards. It is not the purpose
of this analysis to model actual projected intervention rates, but it is useful to consider briefly how costs,
benefits, net benefits, and net benefit-maximizing standards would be affected by an assumed lower rate
of intervention.
Clearly, costs, benefits, and net benefits would all decrease in rough proportion to the decrease in
interventions, since interventions are what engender both costs and benefits. The effect on net benefit-
maximizing standards, however, is more difficult to assess. Especially if the decrease in intervention
rates were applied uniformly across different home types, there is no clear reason to suspect that these
standards should indeed change from the current analysis results.
7-18
~403 EA
-------
References
Battelle. (1996). Procedures and Results for Input to the Economic Analysis for Section 403. Prepared
for TCB, CMD, OPPT, US EPA, June 17.
Battelle. (1997.) Risk Assessment for the Section 403 Rulemaking. Draft Report. Volume 1, Chapters
1 to 7. Prepared by Battelle, Columbus, Ohio for U.S. EP A, Office of Pollution Prevention and
Toxics, Chemical Management Division. July 1.
Schwartz, J. (1994.) Low-Level Lead Exposure and Children's IQ: A Meta-Analysis and Search for a
Threshold. Environmental Research, Vol 65: 42-55.
U.S. Department of Commerce and US. Department of Housing and Urban Development. (1989.)
American Housing Survey for the United States in 1989. DOC, Economic and Statistics
Administration and Bureau of the Census and HUD, Office of Policy Development and
Research.
US. Environmental Protection Agency. (1985.) Costs and Benefits of Reducing Lead in Gasoline:
Final Regulatory Impact Analysis. Prepared by the U.S. Environmental Protection Agency,
Office of Policy Analysis, Economic Analysis Division. February.
U. S. Environmental Protection Agency. (1986.) Reducing Lead in Drinking Water: A Benefit
Analysis. Prepared by US. Environmental Protection Agency, Office of Policy Planning and
Evaluation, Draft Final Report. December.
~403 EA
7.19
-------
7.20
M03 EA
-------
8. Supplementary Analyses
In addition to the benefit -cost analysis, several other types of impacts are important to consider in
evaluating a regulation. This chapter presents analyses that measure the impact of ~403 of Title IV of
the Toxic Substances Control Act (TSCA) on small entities (Section 8.1), minority and low-income
groups (Section 8.4), children (Section 8.5), and the burden placed on state, local and tribal governments
and the private sector (Section 8.2). In addition, the costs of complying with paperwork requirements for
~403 were considered (Section 8.3).
8.1
The Regulatory Flexibility Act (RFA) and Small Business Regulatory Enforcement
Fairness Act (SBREFA)
As described in the Preamble and earlier chapters of this report, the ~403 standards do not require or
mandate any actions by homeowners, landlords, or personnel performing lead-based paint identifications
and interventions. Instead, ~403 standards inform decision-makers about what conditions constitute a
hazard and recommend potential actions. As a result, EPA is not required to conduct a Regulatory
Flexibility Analysis under the Regulatory Flexibility Act (RF A), as amended by the Small Business
Regulatory Enforcement Fairness Act (SBREFA).
The RFA requires analysis of a rule's economic impact on the small entities that will be subject to the
rule's requirements. It requires that the analysis identify the types, and estimate the numbers, of small
entities "to which the proposed [or final] rule will apply," and describe the rule "requirements" to which
small entities "will be subject" and any regulatory alternatives, induding exemptions and deferrals, which
would lessen the rule's burden on small entities. (Sections 603 and 604 ofthe RFA.) Rules that do not
establish requirements applicable to small entities (e.g., rules establishing or revising national ambient air
quality standards under the CAA or water quality standards under the Clean Water Act) are thus not
susceptible to RFA analysis and may be certified as not having a significant economic impact on a
substantial number of small entities. This is particularly true when the national standards do not
themselves require any particular action, as is the case with ~403.
Nevertheless, EP A has conducted a more limited analysis of the potential impact on small entities of
these standards as they work within the market. Two groups of entities are considered: lead-based paint
inspection and abatement firms, and landlords. The small entity impacts of ~403 on the lead testing and
abatement sector are presented in Section 8.1.1 and the small entity impacts on the real estate sector are
presented in Section 8.1.2.
8.1.1 Impact of *403 on the Lead Testing and Abatement Industry
The impact of ~403 on small lead testing and abatement firms is ambiguous. In general, it is expected
that the information dissemination facilitated by ~403 will result in additional household lead
interventionsl However, there is no reliable method for determining how many additional interventions
It is possible, although unlikely, that ~403 will result in fewer household lead interventions. This would
occur if households are currently intervening at lead levels below those outlined in the ~403 standards and
if those interventions stopped occurring once the ~403 standards were distributed. Given the current low
rate of interventions and the persistence of lead levels in excess of those warranting intervention under the
standards, it is expected that interventions will not decrease after promulgation.
~403 EA
8-1
-------
will occur. Even if these numbers were known, it is not known whether the additional interventions
would result in relatively more business for small testing and abatement firms. Data on the size
distribution of firms in this industry can allow for some informed speculation regarding the likely impact
of 9403 on small businesses.
8.1.1.1 Major Findings
There is no Standard Industrial Classification (SIC) code that uniquely corresponds to the lead testing
and intervention industry. Based on the types of activities performed, however, most of the firms
affected by this regulation are likely to be part of two SIC groups:
.
SIC 1799 Construction: Special Trade Contractors, Not Elsewhere Categorized, or
.
SIC 8734 Engineering, Accounting, Research, Management, and Related Services: Testing
Laboratories
Inferences about the potential impact of 9403 on the lead testing and intervention industry can be made
based on data available from the Census on these two SIC codes.
The Small Business Administration (SBA) defines the small business threshold for the construction
trades (SIC 1799) as firms earning less than $7 million per year, while the small business threshold for
the testing industry (SIC 8734) is defined as firms earning less than $5 million. Exhibit 8.1 provides
data from the Census of Construction (1992) on the size distribution of firms in these two industries.
Notice that 99.02% of the firms in the construction industry (SIC 1799) are small businesses and 96.36%
of the firms in the testing industry (SIC 1834) are small businesses. To the degree that 9403 will
increase demand for lead testing and abatement services, the fact that almost all finns in these two
industries are small businesses implies that the impact on small businesses will be positive and
potentially substantial.
8.1.2 Impact of *403 on the Rental Real Estate Sector
The analysis of the impact of 9403 on the real estate sector is restricted to owners of multi-family
residential properties. Even though 9403 does not mandate any intervention activity and, as a result,
carries no direct legal mechanism to ensure that homes exceeding the standard are abated, these standards
will become part of Federal mortgage programs administered by the U.S. Department of Housing and
Urban Development. In addition, it is likely that an indirect legal enforcement mechanism will develop
through the threat of tort law liability suits. While 9403 was developed to provide guidance for
homeowners to determine when a lead intervention is warranted, it can also serve as guidance for the
courts in determining when a property owner's decision to not intervene is an act of negligence for which
the owner can be held financially liable.
8-2
~403 EA
-------
Exhibit 8.1
Characteristics of Establishments
Number of Percent of Total
Establishments Establ ishments
SIC 1799: Small Businesses
9,392 37.17%
5,664 22.41%
3,859 15.27%
3,220 12.74%
2,151 8.51%
641 2.54%
95 0.37%
25,022 99.02%
Average Sales ($)
Less than $1 OOk
$100 to $249k
$250 to $499k
$500 to $999k
$1 to $2.49 mil
$2.5 to $4.9 mil
$5 to $7 mil
Total or Weighted
Average-Small
Average Number
Employees
42,773
161,763
354,328
703,325
1,525,144
3,435,056
5,424,933
442,920
1.6
3.5
6.2
10.5
20.6
41.4
64.0
6.8
$7 to $9.9 mil
$10+ mil
Total or Weighted
Average-Small
SIC 1799: Large Businesses
144 0.57%
104 0.41%
248 0.98%
7,704,316
20,235,067
12,404,476
90.9
213.2
135.7
$5 to $9.9 mil
$10+ mil
Total or Weighted
Average-Small
Source: As reported in TSCA Title IV, Sections 402(a) and 404: Target Housing and Child Occupied Facilities
Final Rule, Regulatory Impact Analysis, August 1996, U.S. Environmental Protection Agency
Less than $1 OOk
$100 to $249k
$250 to $499k
$500 to $999k
$1 to $2.49 mil
$2.5 to $4.9 mil
Total or Weighted
Average-Small
SIC 1834:
532
903
782
745
753
286
4,001
Small Businesses
12.81%
21.75%
18.83%
17.94%
18.14%
6.89%
96.36%
SIC 1834:
113
38
151
Large Businesses
2.72%
0.92%
3.64%
57,195
171,025
358,313
711,117
1,587,911
3,395,601
790,224
1.9
3.6
6.5
11.5
24.3
48.2
12.5
6,811,283
19,885,132
10,101,391
87.5
247.3
127.7
Furthennore, mortgage lenders are likely to be more hesitant to fund property acquisitions if those
properties exceed the 9403 standards. This reluctance stems from the mortgage lenders desire to ensure
that they are not held liable for any adverse health impacts that lead levels in excess of the standard may
induce2. The combination of tort liability suits and mortgage lending requirements indicates that
landlords are the group most likely to follow 9403 to the letter, intervening whenever the standard is
exceeded and not intervening when lead levels are deemed "acceptable" by 9403.
The presumption the mortgage lenders may be hesitant to lend for purchases of property with lead
standards in excess of ~403 standards stems from the experience with the asbestos regulations. After
promulgation of the asbestos regulations, mortgage lenders made asbestos abatement a condition of
lending. While this conditional lending has declined over time it is anticipated that a similar initial
response will result from the potential liability issues implied by the ~403 standards.
~403 EA
8-3
-------
8.1.2.1 Definition of Small Entity
The focus in this section is on the cost to multi-family residential property owners of complying with
9403 and any potential difference in the cost burden likely to fall on owners of smaller rental businesses.
To do this it is necessary to define what constitutes a small rental business. The Small Business
Administration defines the small business threshold for an "owner or operator of rental apartment units"
at $5 million in rental revenue.
Data from the Property Owners and Managers Survey (POMS) is used to determine if property meets the
small business criteria and whether owners are able to absorb intervention costs associated with 9403 out
of their rent streams. POMS is a national survey of rental units conducted from November 1995 to June
1996 by the U.S. Census Bureau. The sample consists of 16,300 rental units in 5,754 properties which
were stratified and assigned weights to reflect the national stock ofrental units. Publicly owned, military,
owner-occupied and vacation units were excluded from the study.3
Total rental revenue per owner was determined by multiplying the rent of the unit surveyed by the
number of units in all properties owned by the landlord. Using the SBA definition of a small rental
business, nearly all of the properties (99.6%) surveyed by POMS were owned by small businesses.
8.1.2.2 Expected Impacts
The ratio of annual compliance costs to annual rent streams gives an indication of the ability of a rental
business to comply with the 9403 standards. In order to calculate this ratio one needs data on the lead
levels in the property (this determines what interventions, if any, must be performed), the number of units
in the building, the rate at which interventions occur, and the total rent from all units in this property.
Unfortunately no single data source contains all of this information. The HUD data set described in
Chapter 3 provides data on household lead levels for 63 multi-family properties. The POMS data set
provides data on the rent for each rental unit sampled, as well as the number of units in the property.
Using these data, the analysis calculates the annual costs each landlord would incur for testing and
intervention, and compares this to the landlord's annual rent. The ratio of costs to rents is then compared
against standard benchmarks to evaluate whether or not the impact would be characterized as significant.
In order to make combined use of the HUD and POMS data, the analysis exploited the fact that each
sample was representative of multi-family housing nationwide. Hence, the frequency of interventions
predicted based on the HUD data set reflect nationwide frequency, and these frequencies could then be
applied to the properties found in the POMS data set. Exhibit 8.2 below gives the percentage of multi-
family properties that require various lead interventions according to the HUD data, the frequency with
which those interventions need to be repeated to insure that a child is protected for six years, and the cost
of each intervention.
As with the cost and benefit estimation in Chapters 4 and 5, the cost of compliance calculated here
assumes that units are tested and interventions are performed whenever a child is about to be born into a
unit. The birth rate of 3.8% determines the frequency of testing and the birth rate combined with the
A separate survey, asking a different set of questions, was perfonned for single-family rental units. Since
the number of single family rental units is very small, the analysis here relies solely on multi-family rental
units.
8-4
~403 EA
-------
probability of requiring different lead interventions given in exhibit 8.2 determine the frequency of lead
interventions within a property.
Low-intensity Interior Paint
High-Intensity Interior Paint
Low-Intensity Exterior Paint
High-Intensity Exterior Paint
Dust
Soil
Testing
Exhibit 8.2
Frequency of Lead Interventions in Multi-Family Housing
Probability of Frequency of
occurrence in multi- intervention required to
family housing protect child for 6 years
0.8% 2
0.0% 1
1.6% 2
0% 1
17.7% 2
0.0% 1
100.0% 1
Cost of each
intervention
$437
$4,687
$182
$2,275
$262
$2,077
$235
For example, a dust intervention must be repeated every four years, and hence, 2 dust interventions are
required per child born (on average).4 Using a birth rate of 3.8% and a probability of occurrence of
17.7%, the annual number of dust interventions in a 100 unit building is 1.34 units
(100*.177*.038*2=1.34) at a cost of $351 0.34*$262=$351). Similar calculations are performed for
low-intensity interior and exterior paint interventions and testing. The annual costs for each type of
intervention are summed together with the testing costs to determine the total annual compliance cost.
This calculation is performed for each property in the POMS data set based on the number of units in the
property.
The annual rent stream for each property is calculated as the rental revenue from the single unit surveyed
in the POMS study multiplied by the number of units in the building. This calculation assumes that the
unit surveyed in the POMS study is representative of the other units in the building. There is no means of
determining whether this assumptions leads to an over or underestimation of the rent stream.
The ratio of annual compliance costs to annual rent payments is equivalent to the commonly used ratio of
compliance cost to sales, and it determines the degree to which the property owner will be capable of
complying with the ~403 standards. For the purposes of determining small business impacts it is
assumed that business can accept a compliance cost to rent ratio less than 3%. Exhibit 8.3 provides the
number of property owners experiencing a compliance cost to rent ratio greater than 3%, between 3% and
1 %, and less than 1 %.
Notice that no property owners experience a cost to rent ratio larger than 3%. No large businesses
experience a compliance cost to rent ratio greater than 1 %. Just over 22,000 small rental businesses are
expected to have an annual compliance cost to rent ratio greater than 1 % but less than 3%. While 22,000
may appear to be a large number of businesses, there are over 2.2 million small rental businesses in
An exception to this occurs if any given unit has more than one child under 6. For example if a couple has
2 children 2 years apart then only 2 interventions will be performed and both children will be protected
for 6 years. However, for the small business analysis, we assume that 2 interventions are performed for
each child born. This leads to an overestimation of compliance costs for property owners.
~403 EA
8-5
-------
existence. Thus, only 1 % of all the small rental businesses experience a cost to rent ratio greater than
1 %. Given the relatively small impact on the rental real estate sector in general, and small rental
businesses in particular, the ~403 rule will not have significant economic impact on a substantial number
of small entities.
Exhibit 8.3:
Ratio of Annual Compliance Costs to Annual Rent Payments, by size of business
Comparative Ratios Large Businesses Small Businesses
Annual Compliance Cost 15,060 2,192,394
< 1%
Annual Rent Payments
Annual Compliance Cost 22,191
1% <3% 0
AnnualRentPaymen~
Annual Compliance Cost 0 0
3%
Annual Rent Payments
Total Number of Businesses 15,060 2,214,585
8.2 Unfunded Mandates Reform Act (UMRA)
Under Title II of the Unfunded Mandates Reform Act, the cost to state, local and tribal government or the
private sector of compliance with federal regulations must be calculated and considered during the
regulatory process. Because ~403 is a regulation which provides information to consumers about
household lead safety and does not require households or public entities to take any action with respect to
that information, no costs are imposed on state, local and tribal governments or the private sector. As
such, this action is not subject to the requirements of sections 202 and 205 of (UMRA) because this
action does not contain any "federal mandates." Similarly this regulation contains no regulatory
requirements that might significantly or uniquely affect small governments, so no action is needed under
Section 203 of UMRA.
8.3 Paperwork Reduction Act (PRA)
The Paperwork Reduction Act (PRA) requires EP A to prepare an Information Collection Request (ICR),
which estimates the reporting and recordkeeping burden imposed by their regulations. Under the PRA,
"burden" means the total time, effort, or financial resources expended by persons to generate, maintain,
retain, or disclose or provide information to or for a Federal agency. This includes the time needed to
review instructions; develop, acquire, install, and utilize technology and systems for the purposes of
collecting, validating, and verifying information, processing and maintaining information, and disclosing
and providing information; adjust the existing ways to comply with any previously applicable
instructions and requirements; train personnel to be able to respond to a collection of information; search
data sources; complete and review the collection of information; and transmit or otherwise disclose the
information.
8-6
~403 EA
-------
Section 403 contains no reporting or recordkeeping requirements, and thus no ICR is necessary for this
rule. However, an ICR was prepared and filed for the promulgation of regulations for TSCA ~402(a) and
404, and these burden estimates were based on estimates of the number of lead-based paint identification
and intervention activities anticipated. EPA re-examined the ~402(a) and 404 ICR and determined that
these estimates would not change due to the standards being proposed for ~403. The ~402(a)/404 RIA
and ICR estimated lead-based paint identification and intervention rates based on activity levels in
Massachusetts. Massachusetts standards are similar to those being proposed by EP A, and Massachusetts
has a very aggressive enforcement program coupled with state loan programs to encourage abatements in
units occupied by low-income families. Therefore, national rates under ~403 are not likely to be higher
than these. In addition to number of events, the reporting and recordkeeping burden is affected by the
number of people trained and filing for certification, and the number of firms offering training. Again,
the ~402(a)/404 RIA and ICR based these numbers on Massachusetts estimates. The analysis
determined, and state officials confirmed, that there was significant overcapacity in the state. Both
because the number of events and the number of persons and firms were overestimates for ~402(a)/404,
and because the ~403 standards are similar to the Massachusetts standards, EPA determined that the
~402(a)/404 ICR did not need to be revised to reflect the ~403 standards.
8.4 Executive Order 12898 Federal Actions to Address Environmental Justice
Increasingly questions of equity are playing a role in crafting environmental regulation. Of particular
concern is the relationship between who receives the benefits of regulation and who bears the costs.
Initially one might assert that the voluntary nature of this rule ensures that those who bear the costs of
~403 also receive the benefits and, hence, the distribution of costs and benefits across any demographic
or socioeconomic group would be identical. However, two households performing the same intervention
with the same costs may receive dramatically different benefit levels. For example, a reduction of soil
lead from 5000 ppm to 150 ppm will yield greater benefits than a reduction from 2100 ppm to 150 ppm
despite the fact that the cost of the soil interventions are the same. This section seeks to determine how
the costs and benefits of ~403 are distributed across race and income lines.
Data on race and income were collected during the HUD survey along with information on lead levels in
interior and exterior paint, dust, and soil for 284 households. This sample was stratified and weighted to
represent the housing characteristics of the nationwide housing stock. Data on race were combined to
form 3 major race categories: non-Hispanic white, African-American, and other, non-white. Other, non-
white includes Hispanic (of any race), Asian, and American-Indian households. Data from the HUD
survey on income were limited to whether the household had an annual income of more or less than
$30,000. While households earning more than $30,000 are referred to as upper-income households and
those earning less than $30,000 are referred to as lower-income households throughout this chapter,
$30,000 is not a typical definition of upper-income. The income classification should really be thought
of as a indicator, where households earning less than $30,000 are below the median income and those
earning more are above the median.
The data from the HUD survey were not designed to be used for an environmental equity analysis. As a
result three major assumptions were made which allow the data to be used in this analysis. A fourth
assumption was required in order to quantify expected costs for each demographic group. These
assumptions are:
~403 EA
8-7
-------
.
The HUD sample is demographically representative of the national population: Because
demographic analysis was not the purpose for which the data were originally collected, the
sample was not stratified to represent the demographic make-up of the nationwide population
but rather to reflect the characteristics of the national housing stock. Thus, the first assumption
made in the equity analysis is that the HUD sample is not only a reasonable representation of the
national housing characteristics but also a reasonable representation of the population living in
those housing units. This is a strong assumption because in the HUD survey there are
approximately 7.2 million African-American households while in actuality the figure is closer to
10.2 million. This discrepancy can be overcome somewhat by focusing strictly on percentage of
households or per-household measures of equity.
.
The demographic make-up of households does not change between 1997 and 2046: In
this analysis, the costs of compliance with 9403 occur between 1997 and 2046. In order to
aggregate those costs and assign the aggregate to different race and income groups it was
assumed that the demographic make-up of the household does not change over those 50 years.
.
Birth rates do not vary across racial and income groups: The costs and benefits calculated
below are based on the "birth-trigger" model which assumes that interventions occur when a
child is born into a housing unit with lead levels that exceed the 9403 standards. No
differentiation in birth rates across race or income is incorporated.
.
All households exceeding the ~403 standard perform the appropriate interventions: This
assumption has been used throughout the cost-benefit analysis of 9403 in order to determine
which households will perform an intervention and hence, what the costs and benefits of
compliance are.
While these assumptions may seem extreme, the BUD survey provides the only compilation of data on
both lead characteristics of homes and demographic composition of the members of the household. It is,
despite its limitations, the only data available for this analysis. A qualitative discussion of the impact of
these assumption on the results can be found in Section 8.4.2 below.
Given these assumptions, the data on race and income were used to analyze two questions: who bears the
costs of 9403, and who receives the benefits? The null hypothesis is that households that exceed the
9403 standards are equally likely to contain occupants of any race and income group and, as a result, all
groups will face the same relative cost burden and expect the same benefit from 9403. The results of the
statistical analysis are presented in Section 8.4.1.
8.4.1 The Distribution of Costs and Benefits by Race and Income
The cost analysis for 9403 assumes that a home that exceeds any of the four environmental lead levels
(floor dust, window sill dust, paint or soil) established by the 9403 standard will perform the appropriate
interventions. Cost of compliance with 9403 is then simply a function of the condition of the housing
stock--older homes with paint in deteriorating condition or homes with high soil lead levels will have
higher costs of compliance. This analysis goes one step further by connecting the condition of the
housing stock to demographic characteristics of the members of that household. Specifically, this
section tests the hypothesis that all racial and income groups are equally likely to live in housing that
warrants an intervention, and hence, equally likely to bear the cost of compliance with 9403. Similarly
8-8
~403 EA
-------
this section tests the hypothesis that all racial and income groups are equally likely to live in housing with
the same pre-intervention lead levels and, as a result, will experience equal blood-lead reductions from
intervention.
The equity of the cost distribution is determined using two different measures of cost burden. The first
measure is the percentage of any given race or income group that lives in housing that exceeds the ~403
standards. If the distribution of lead hazards is spread evenly across race and income groups then the
cost of mitigating these hazards should also be spread evenly. However, some housing will exceed the
~403 standard for more than one media (i.e., soil, paint or dust) and will require more expensive
interventions. Thus, the distribution of households exceeding the ~403 standard for at least one media
may not provide an accurate estimate of the cost distribution. The second measure compares the per-unit
intervention cost to the median income of each race and income group. The per-unit intervention costs
were calculated as the total intervention costs divided by the number of households with an intervention.
Comparing the per-unit intervention cost to the median income provides an indication of the affordability
of compliance for each race and income group.
The equity of the benefits distribution is more difficult to measure. The monetized benefits described in
Chapter 6 cannot be calculated for each demographic group because these benefits rely on a national
distribution of blood-lead levels that cannot be disaggregated by race or income. Instead of dollar
benefits, a benefits indicator is used to determine the distribution. The indicator is a weighted average of
the reduction in blood-lead concentration resulting from interventions for households in each
demographic group. While not a true representation of the change in blood-lead concentration resulting
from ~403 (this would require an adjustment at the aggregate, nationwide level, to achieve consistency
with NHANES findings), it is a reliable indicator of those changes as long as one examines only the
ordinal and not the cardinal relationship. In other words, if the figures indicate that non-Hispanic whites
have a change in blood-lead of 1.0 Ilg/dl and African-Americans only 0.5 Ilg/dl then we can say with
confidence that non-Hispanic whites will have a higher change in blood-lead than African-Americans but
the exact numbers 1.0 and 0.5 are not reliable.
An analysis of the distribution of benefits by demographic category also needs to recognize that the
expected change in blood-lead concentrations will depend heavily on which blood-lead model is used.
Thus, the figures below are presented for both the IEUBK model and the Empirical model. As explained
in prior chapters of this report, the IEUBK model consistently predicts higher changes in blood-lead
concentrations. However, the ordinal relationship predicted by the IEUBK and the Empirical model are
often consistent and it is only this ordinal relationship that can be used as a reliable indicator of benefits.
The results of both the cost and benefit distributions are presented in Exhibit 8.4.
The results presented in Exhibit 8.4 indicate that risk and the costs of complying with ~403 are not
distributed evenly among racial groups. In comparison to non-Hispanic whites, a much larger share of
the housing stock occupied by African-Americans and a somewhat larger share of the housing stock
occupied by other non-white households warrant an intervention. While the per-unit intervention costs
are higher for white households, the median income for white households is also higher. The per-unit
compliance cost as a percent of the median income are roughly the same for white households and other,
non-white households (5.5%) but slightly higher for African-American households (6.7%).
~403 EA
8-9
-------
However, intervention costs do not vary with income. In general, lower-income households are no more
likely than higher income households to require an intervention. Lower-income and higher-income
households also face approximately the same per-unit compliance costs. However, this means that lower-
income households will have to forgo a larger percentage of their income to comply with ~403.
Exhibit 8.4
Distribution of Intervention Costs and Benefits by Race and Income
Average
Per-Unit
Intervention
Costs
% Exceeding
~403
Standards
Median
Income
Non-Hispanic White 17.7% $37,919 $2,068
African-American 48.8% $21,301 $1,427
Other, Non-White 26.0% $25,837 $1,427
Total 20.7%
< $30,000 20.0% $16,250 $1,837
$30,000 19.1% $48,750 $2,049
Income Not Available 35.8% NA $1,630
Total 20.7%
Blood-Lead Reduction
IEUBK
Model
1.00
2.22
1.02
Empirical
Model
0.52
0.54
0.53
0.73
1.62
0.81
0.12
0.15
0.16
The distribution of benefits across demographic groups varies depending on which model is used. The
IEUBK model predicts that African-Americans will experience greater reductions in blood-lead
concentrations than white or other, non-white households. Similarly, the IEUBK model predicts that
upper-income households will have larger blood-lead reductions. However, the Empirical model predicts
that blood lead reductions do not vary significantly with race or income. Without formal statistical tests
it is difficult to determine if the difference in blood lead reductions for African-Americans and whites
using the IEUBK model is significant.
The data presented in Exhibit 8.4 do not raise significant equity concerns. Because African-American
households are more likely to live in housing units that exceed the proposed 403 standards, they are more
likely to have higher intervention costs. On the other hand, they are also expected to see higher blood-
lead reductions. Similarly, the distribution of costs and benefits do not vary with income although lower-
income households are expected to spend a larger share of their income on lead interventions.
8.4.2 Anticipated changes in results if assumptions are relaxed
As stated previously, there are four main assumptions required in order to use the HUD data for the
equity analysis. The first assumption is that the HUD data present a demographically representative
standard. We know that this is not the case. The HUD survey over-sampled non-Hispanic white
households and under-sampled African-American households. This study has used care to ensure that the
measures of equity are relatively unaffected by this discrepancy. For example, by focusing on
relationships within one racial group the assumption of demographic representation is reduced from
assuming that the sample contains the correct number of African-American households, to an assumption
that the African-Americans that were sampled are representative of the African-American population.
8-10
~403 EA
-------
The second assumption was that the demographic make-up of pre-1978 housing units will not change
from 1997 to 2046. It is not assumed that people will not move for 50 years, but rather that there will be
no major change in the race or income distribution by housing types. In other words, if 40% of pre-1978
housing with certain characteristics is occupied by African-Americans today then 40% of that category of
housing will be occupied by African-Americans in 2046. However, during the preceding 50 years there
has been a massive demographic shift even within a housing type. If one considers the demographic
make-up of the inner-city areas in 1957 with the demographic make-up in 1997 it is apparent that the
percentage of both African-Americans and lower-income households in those housing units has risen,
while the percentage of non-Hispanic whites and upper-income households has declined significantly. To
the extent that newer housing is more expensive than older housing, the expected trend from 1997 to
2046 is that an increasing percent of pre-1978 housing will be occupied by households that earn less than
$30,000 year.
There are numerous reasons to expect that pre-1978 housing will also become more heavily populated by
non-whites over the next 50 years as well. These reasons could be both voluntary and involuntary.
V oluntary reasons might include an increasing desire on the part of middle-class African-Americans to
stay in the city and strengthen traditional black communities. Involuntary reasons include ongoing
housing and fmancing discrimination which prevents some middle-class non-whites from relocating to
newer housing in the suburbs. This is obviously a highly simplified discussion of the complex economic
and sociological relationships between race, income and housing location; however, the point is that the
demographic make-up of pre-1978 housing is not likely to remain constant over the next 50 years.
Moreover the expected shift in the racial and income composition of these homes is likely to place an
ever increasing share of the ~403 compliance burden as well as the compliance benefits on lower-income
and non-white households. To the extent that the ~403 standards result in a more targeted approach to
intervention, these actions will take place where they are most needed.
The third assumption made as part of the equity analysis is that the probability of a child being born into
a home is constant for all home types, and hence for all demographic groups. Exhibit 8.5 presents the
birth rate per 1000 persons in 1990 by race. Data on birth rates by income are not available.
Exhibit 8.5
Birth Rates per 1,000 Households by Race
Birth rate
Non-Hispanic white 14.4
African-American 22.4
Other, non-white 24.5
Source: United States Department of Commerce. 1996. Current
Population Reports, P25-1130, as cited in Statistical Abstract of the
United States (Washington, DC, U.S. Government Printing Office, p.20.
The data in Exhibit 8.5 do not support the assumption that birth rates are equal across demographic
groups. African-American and other, non-white households have higher birth rates than non-Hispanic
whites. Birth rates for all three groups are projected to decrease by the year 2000 but non-white
households are still expected to have higher birth rates than non-Hispanic white households. Thus, if we
believe that households will take actions with respect to 9403 when children are present, then non-white
households will have higher compliance costs. However, it is also true that if more African-American
~403 EA
8-11
-------
and non-white households have children and perfonn the necessary intervention, then more African-
American and non-white children will benefit from 9403.
The final assumption made in the quantitative analysis presented in Section 8.4.1 is that all households
that exceed the 9403 standard perform the appropriate interventions. The Current Population Survey's
(CPS) Lead Awareness Supplement provides infonnation that can be used to calculate the percentage of
total interventions by socio-economic characteristics. This is a HUD-designed supplement to the
monthly CPS conducted in December 1994. Exhibit 8.6 gives the incidence rate by race, and median
income level of households that intervene.
Exhibit 8.6
Characteristics 01 Households Living in Pre-1978 Housing and Pre-1978 Housing with
Interventions
Total
% white
median income
58.6 million
78.6%
$27,500
pre-1978 housing with an
intervention
4.8 million
85.0%
$32,500
pre-1978 housing
Households living in units in which an intervention has occurred are more likely to be white and have
higher incomes than the general population living in pre-1978 housing. These results indicate that the
assumption that all households perfonn the appropriate interventions could underestimate the equity
discrepancy presented in section 8.4.1 above. However, it is unclear whether the fact that intervening
households tend to be more white and have higher income is a reflection of a greater level of knowledge
of lead hazards or a reflection of ability to pay for interventions. The CPS survey also showed that
interventions were strongly and positively related to the level of knowledge about lead-based paint
hazards, which is also related to income. To the extent that the results reflect greater knowledge and
9403 provides infonnation on lead hazards, 9403 may induce more African-American and lower-income
households to perform interventions. However, if these discrepancies are based on ability to pay then the
results presented in Section 8.4.1 will overstate both the cost and the benefits of compliance for non-
white households.
Data from the State of Massachusetts can also be used to assess the assumption that all households are
equally likely to intervene regardless of race or income. Since 1988 all inspections for lead-based paint
and all lead abatements occurring in the state of Massachusetts are required to be performed by licensed
professionals and those professionals are required to report the inspection or abatement (termed
deleading) to the state. Since the inspection and deleading data do not include socio-economic nor
demographic information on the individual housing units, the analysis of the Massachusetts data was
conducted at the census block group level with data from the 1990 U.S. Census of Population and
Housing. Using ordinary least squares (OLS), the inspection and deleading data were regressed on
several socio-economic characteristics for their census block group. Separate equations were estimated
for rental and owner-occupied units. For each, three equations were estimated: number of inspections,
number of inspections that found a violation, and number of deleadings.
The explanatory variables were:
8-12
~403 EA
-------
.
Number of housing units built before 1980;
Number of children;
Number of households that moved in past five years;
Median household income; and
Number of whites.
.
Holding constant for the number of inspections that found a violation, the number of rental unit
deleadings in a census block group increases with an increase in the number of children under age 6 and
with the presence of whites. The number of rental deleadings decrease with increases in median income.
The results are similar for owner-occupied housing, except that the number of owner-occupied deleadings
is inversely related to the number of whites in the block group. Households moving increase deleadings
in owner-occupied units but not in rental units. Since race has no significant effect for rental units and
interventions in owner-occupied units increase as the number of non-whites increases, and the number of
deleadings decreases slightly as income increases, the assumptions made for this analysis that
interventions in a given house type are evenly distributed among units of that type regardless of income
or race, do not appear extreme.
8.5 Executive Order 13045--Protection of Children from Environmental Health Risk and Safety
Risks
Executive Order 13045 requires that regulations undergo review by the Office of Management and
Budget (OMB) if the regulatory action is economically significant and concerns an environmental health
risk or safety risk that an agency has reason to believe may disproportionately affect children. The focus
of the ~403 regulation is on the protection of children's health and the household lead standards were
chosen based on an analysis of the health risks to children only. The benefits from ~403 outlined in
Chapter 5 are a reflection of benefits to children under 6 years old only.
Of the estimated 173 million children born between 1997 and 2046, approximately 131 million children
will be born into housing built prior to 1979. It is estimated that ~403 will result in reductions in
exposure to household lead in soil, dust, and paint for 43.8 million children over the next 50 years. This
reduction in exposure, in turn, will reduce the incidence of elevated blood-lead levels and increase
average IQ. Exhibit 8.7 presents blood-lead and IQ statistics for both the baseline and post-compliance
scenanos.
Notice that the health impacts of ~403 are often substantial. The reduction in the number of children
suffering from elevated blood-lead levels due to pica (direct ingestion of paint chips) is on the order of
1.3 million. The reduction in the number of children with elevated blood-lead levels (greater than 10
Ilg/dl) from all sources is estimated at 2.4 to 6.9 million. The increase in average IQ depends greatly on
which benefits model is used but is estimated to be between 0.8 and 3.1 points for the 43.8 million
children estimated to be affected by interventions. The number of children who will avoid an IQ less than
70 points is between 9,000 and 26,000 depending on the benefits model employed.
~403 EA
8-13
-------
Exhibit 8.7
Beneficial Health Impacts on Children Resulting from ~403
: Number with: Number with i Number with j
: elevated i blood-lead : blood-lead :
. . .
: blood-lead i greater than: greater than
i due to pica i 10 IJg/dl : 20 IJg/dl
i (millions) j (millions) i (millions)
4.12 i 2.4: 10.6 j
3.34 i 1.1 : 3.6:
. .
. .
. .
. .
Mean blood-
lead level
(lJg/dl)
: Number
j Average : avoiding IQ
: IQ point gain i less than 70
1.0 i NA : NA
0.2 j 3.1 : 26,000
Baseline
Post-g403
IEUBK
Post-g403
Empirical
3.91 1.1 8.2 0.7 0.8 9000
8-14
~403 EA
-------
References
Massachusetts Childhood Lead Poisoning Prevention Program. 1996. Inspection and Deleading
databases.
United States Department of Commerce. 1994. Current Population Survey: Lead-based Paint
Awareness Supplement. Public Use file. December.
United States Department of Commerce. 1996. Current Population Reports, P25-1130, as cited in
Statistical Abstract of the United States (Washington, DC, U.S. Government Printing Office,
p.20.
United States Department of Commerce and U.S. Department of Housing and Urban Development.
1996. Property Owners and Managers Survey, Multi-housing Unit Properties.
United States Department of Housing and Urban Development (HUD). 1993. National Survey of Lead-
Based Paint in Housing: Documentation of Analytical Data Files. Prepared for U.S. Department
of Housing and Urban Development by We stat, Inc.. November 30.
United States Environmental Protection Agency. 1996. TSCA Title IV, Sections 402(a) and 404:
Target Housing and Child Occupied Facilities Final Rule, Regulatory Impact Analysis.
United States Small Business Administration. 1993. Small Business Thresholds by 4-digit Standard
Industrial Classification.
~403 EA
8-15
-------
8-16
i403 EA
-------
References
Abt Associates. 1995a. TSCA Title IV, Sections 402(a) and 404: Target Housing and Child-Occupied
Facilities Final Rule Regulatory Impact Analysis. Prepared for U.S. EPA, Office of Pollution
Prevention and Toxics, Regulatory Impacts Branch. September.
Abt Associates. 1995b. Exemption of Lead-Based Paint Architectural Debris from Subtitle C of RCRA,
Analysis of Costs, Benefits, and Risks. Prepared for Office of Solid Waste, U.S. EPA.
September 11.
Alliance to End Childhood Lead Poisoning (AECLP). 1993. Understanding Title X: A Practical Guide
to the Residential Lead-based Paint Hazard Reduction Act of 1992. AECLP, Washington, DC.
January.
American Academy of Pediatrics (AAP). 1995. Treatment Guidelines for Lead Exposure in Children.
Pediatrics. 96(1): 155-160.
Annest, J., 1. Pirke1, D. Makuc, J. Neese, D. Bayes, and M. Kovar. 1983. Chronological Trend in Blood
Lead Levels Between 1976 and 1980. New England Journal of Medicine, 308: 1373-1377.
Ashenfelter, O. and J. Ham. 1979. Education, Unemployment, and Earnings. Journal of Political
Economy 87(5):99-131.
Axelrad, D. 1993. Guidance on the Preparation of Economic Analyses and Regulatory Impact
Analyses in OPPT. Regulatory Impacts Branch, Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency, Washington, D.C., January, pp. 26-27.
Barnett, W.S. 1992. Benefits of Compensatory Preschool Education. The Journal of Human Resources
27(2): 279-312.
Battelle. 1996. Procedures and Results for Input to the Economic Analysis for Section 403 and
Addenda to Report on Inputs to Economic Analysis. Prepared by Battelle, Columbus, OH, for
U.S. EP A, Office of Pollution Prevention and Toxics, Chemical Management Division, June 17
and August 29.
Battelle. 1997. Risk Assessment to Support Standards for Lead in Paint, Dust, and Soil. Prepared by
Battelle, for National Program Chemicals Division, Office of Pollution Prevention and Toxics,
U.S. Environmental Protection Agency. EPA 747-R-97-006, December.
Bishop, J. 1992. The Impact of Academic Competencies on Wages, Unemployment, and Job
Performance. Carnegie-Rochester Series on Public Policy 37: 127-194.
Blackburn, M.L. and D. Neumark. 1992. Unobserved Ability, Efficiency Wages, and Interindustry
Wage Differentials. The Quarterly Journal of Economics, 1421-1436.
~403 EA
R-1
-------
Blackburn, M.L. and D. Neumark. 1993. Omitted-Ability Bias and the Increase in Return to Schooling.
Journal of Labor Economics 11(3): 521-544.
Boadway, RW. 1979. Public Sector Economics. Little, Brown and Company (Inc.), Boston, MA, pp.
29-43.
Centers for Disease Control (CDC). 1991a. Strategic Plan for the Elimination of Childhood Lead
Poisoning, February.
Centers for Disease Control (CDC). 1991b. Preventing Lead Poisoning in Young Children: A
Statement by the Centers for Disease Control. Public Health Service, Centers for Disease
Control and Prevention, U.S. Department of Health and Human Services, October 1991.
Chamberlain, G. and Z. Griliches. 1976. More on Brothers. Harvard Institute of Economic Research
Discussion Paper No. 469. Cambridge, MA.
Clark, S., R Bornschein, P. Succop, S. Roda, and B. Peace. 1991. Urban Lead Exposures of Children in
Cincinnati, Ohio. Chemical Speciation and Bioavailability, 3(3): 163-171.
Council of Economic Advisors (CEA). 1996. Economic Report of the President. Government Printing
Office. Washington, D.C. February.
Cropper and Krupnick. 1989. The Social Costs of Chronic Heart and Lung Disease. Washington, DC:
Resources for the Future.
Elias, R 1993. Personal communication between Rob Elias, U.S. Environmental Protection Agency,
Office of Research and Development and Alice Tome, Abt Associates Inc. March 16.
Environmental Science and Technology. 1996. 33/50 Program Ends, EPA Declares Success.
30(ll):482A-483A.
Eyraud, J. 1993. Economic Incentives Under TSCA: A Regulator's Guide. Volume I. Review Draft.
For the U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxies,
Regulatory Impacts Branch. Washington, D.C., February, pp. 3-12.
Flaherty, P. 1995, 1996. Personal communication between Pat Flaherty, Specialty Construction, Fridley,
MN and Chris Paciorek, Abt Associates Inc. October, 1995 and July, 1996.
Florini, K.L., G.D. Krumbhaar, and E.K. Silbergeld. 1990. Legacy of Lead: America's Continuing
Epidemic of Childhood Lead Poisoning. Environmental Defense Fund, Washington, DC.
March.
Freeman, A. Myrick (1993). The Measurement of Environmental and Resource Values: Theory and
Methods. Resources for the Future, Washington, DC.
R-2
~403 EA
-------
Ganderton, P. and P. Griffin. 1993. Impact of Child Quality on Earnings and the Productivity of
Schooling Hypothesis. Contemporary Policy Issues 1 I (July): 39-47.
Goldman, Lynn. 1994. USEP A, Office of Prevention, Pesticides and Toxic Substances. Memorandum
on Agency Guidance on Residential Lead-Based Paint, Lead-Contaminated Dust, and Lead-
Contaminated Soil. July 14.
Griliches, Z. 1977. Estimating the Returns to Schooling: Some Economic Problems. Econometrics
45:1-22.
Grogger, J. and E. Eide. 1995. Changes in College Skills and the Rise in the College Wage Premium.
The Journal of Human Resources 30(2): 280-310.
Hanushek, E.A, Rivkin, S.G., and D.T. Jamison. 1992. Improving Educational Outcomes While
Controlling Costs. The Carnegie-Rochester Series on Public Policy 37: 205-238.
Hartman, Robert W. (1990). One Thousand Points of Light Seeking a Number: A Case Study of CBO's
Search for a Discount Rate Policy. Journal of Environmental Economics and Management,
Vol 18(2):S3-S7.
Hometech. 1996. Remodeling and Renovation Cost Estimator.
Illinois Department of Public Health (ILDPH). 1995. Get the Lead Out: Illinois Childhood Lead
Poisoning Surveillance Report. ILDPH, Childhood Lead Poisoning Prevention Program.
Springfield, IL.
Kakalik, J. et al. 1981. The Cost Of Special Education. Rand Corporation Report N -1791- ED.
Kolb, Jeffrey A and Scheraga, Joel D. (1990). Discounting the Benefits and Costs of Environmental
Regulations. Journal of Policy Analysis and Management. Vol 9(3):381-390.
Krupnick, AJ. and M.L. Cropper. 1989. Valuing Chronic Morbidity Damages: Medical Costs and
Labor Market Effects. Draft Final Report to the US Environmental Protection Agency, OPPE.
June 26, 1989.
Lead-Based Paint Hazard Reduction and Financing Task Force (Task Force). 1995. Putting the Pieces
Together: Controlling Lead Hazards in the Nation's Housing. HUD-1547-LBP. July.
Lim, B. 1996. Lead paint abatement and repair and maintenance study in Baltimore. Materials from the
Quality Assurance Project Plan sent by Benjamin Lim of U.S. EP A, Office of Pollution
Prevention and Toxics (OPPT), Chemical Management Division to Nishkam Agarwal, U.S.
EPA, OPPT, Regulatory Impacts Branch. February.
Lind, Robert (1990). Reassessing the Government's Discount Rate Policy in Light of New Theory and
Data in a World Economy with a High Degree of Capital Mobility. Journal of Environmental
Economics and Management, Vol 18(2):S8-S28.
~403 EA
R-3
-------
Markowitz, M.E. and J.F. Rosen. 1991. Need for the Lead Mobilization Test in Children with Lead
Poisoning. Journal of Pediatrics. 119: 305-310. [As cited in CDC, 1991.]
Maryland Department of the Environment (MDDE). 1996. Facsimile from Environmental Analysis
Division to Amy Benson of Abt Associates Inc. Childhood Lead Registry (CLR) Occurrence of
Lead Poisoning in Children 0-72 Months; Reportfor 1/94 12/94. Sent November 15.
Massachusetts Childhood Lead Poisoning Prevention Program. 1996. Inspection and Deleading
databases.
McCord, James. 1996. Personal Communication between Dr. James McCord, Children's Health Care of
St. Paul, MN and Amy Benson of Abt Associates Inc. November 25.
McGee and Gordon. 1976. The Results of the Framingham Study Applied to Four Other U.S.-based
Epidemiologic Studies of Coronary Heart Disease. The Framingham Study: An
Epidemiological Investigation of Cardiovascular Disease, Section 31, April.
Means, R.S. 1996. Repair and Remodeling Cost Data.
Miceli, T.S.; Pancak, K.A.; and Sirmans, C.P. (1996). An Economic Analysis of Lead Paint Laws,
Journal of Real Estate Finance and Economics, 12:59-75.
Moore, Michael J. and Viscusi, W. Kip. (1990). Discounting Environmental Health Risks: New
Evidence and Policy Implications. Journal of Environmental Economics and Management, Vol
18(2):S51-S63.
Murnane, R.J., Willet, J.B., and F. Levy. 1995. The Growing Importance of Cognitive Skills in Wage
Determination. The Review of Economics and Statistics, 77(2): 251-266.
Mushak, P., and A. Crocetti. 1990. Methods for reducing lead exposure in young children and other risk
groups: an integrated summary of a report to the U.S. Congress on childhood lead poisoning.
Environmental Health Perspectives, 89: 125-135.
National Center for Lead-Safe Housing (NCLSH). Undated. Creating a Lead-Based Paint Hazard
Control Policy: A Practical, Step-by-Step Approach for Nonprofit Housing Organizations.
Technical Assistance Bulletin 2.
National Center for Lead-Safe Housing (NCLSH). 1995. SpecMaster Database. Provided to Chris
Paciorek, Abt Associates by Jonathon Wilson, NCLSH. December.
National Conference of State Legislatures (NCSL). 1996a. State Training/Certification! Accreditation
Programs for Lead. November 8. Compiled by Doug Farquhar.
National Conference of State Legislatures (NCSL). 1996b. Update on EPA Lead-Based Paint Activities.
Compiled by Doug Farquhar.
R-4
~403 EA
-------
National Conference of State Legislatures (NCSL). 1997. State Lead Poisoning Prevention Statutes.
July 23. Compiled by Doug Farquhar.
Needleman, H.I., Schell, A., Bellinger, D., Leviton, A., and E. Allred. 1990. The Long-Term Effects of
Exposure to Low Doses of Lead in Childhood: An II-Year Follow-up Report. The New
England Journal of Medicine 322(2): 83-88.
Office of the President. 1996. Economic Report of the President.
Organization for Economic Cooperation and Development (OECD). 1993. Risk Reduction Monograph
No.1: Lead. Background and National Experience with Reducing Risk. Paris: Environment
Directorate,OECD. Document No. OCDE/GD(93)67.
Perket, C. 1994. Hazardous Waste Landfill Prices, 1989-1994. EI Digest, November, pp. 33-36.
Piomelli et al. 1984. Management of Childhood Lead Poisoning. Pediatrics 4: 105.
Pirkle, J.L., J. Schwartz, J.R. Landis, and W.R. Harlan. 1985. The Relationship between Blood Lead
Levels and Blood Pressure and its Cardiovascular Risk Implications. American Journal of
Epidemiology, 121: 246-258.
Pockock, S.J., Smith, M., and P. Bagrhurst. 1994. Environmental Lead and Children's Intelligence: A
Systematic Review of the Epidemiological Evidence. BMJ 309: 1189-1197.
Pooling Project Research Group. 1978. Relationship of blood pressure, serum cholesterol, smoking
habit, relative weight and ECG abnormalities to incidence of major coronary events: final report
of the Pooling Project. Journal of Chronic Disease, 31.
Prenney, B. 1987. The Massachusetts Lead Program: Moving Toward Phase 2. Prevention Update.
Developed by the Maternal and Child Health Consortium Project and the National Coalition on
Prevention of Mental Retardation. April.
Rabinowitz, M. and H. Needleman. 1983. Petrol Lead Sales and Umbilical Cord Blood Lead Levels in
Boston, MA. Lancet, 8314/5 (1):63.
Rhode Island Department of Health (RIDH). 1993. R 23-24.6-PB: Rules and Regulations for Lead
Poisoning Prevention.
..
Salkever, D. 1995. Updated Estimates of Earnings Benefits from Reduced Exposure of Children to
Environmental Lead. Environmental Research 70: 1-6.
Santucci, R. 1996. Personal communication between Robert Santucci, cost consultant to the National
Center for Lead-Safe Housing and president of Urban Restoration Corporation, North Carolina,
and Chris Paciorek, Abt Associates Inc. July.
~403 EA
R-5
-------
Schwartz, J. 1993. Beyond LOELs, p Values and Vote Counting: Methods for Looking at the Shapes
and Strengths of Association. Neurotoxicity, 14(2-3),237-246, Summer/Fall.
Schwartz,1. 1994a. Low-Level Lead Exposure and Children's IQ: A Meta-Analysis and Search for a
Threshold. Environmental Research, Vol 65: 42-55.
Schwartz,1. 1994b. Societal Benefits of Reducing Lead Exposure. Environmental Research 66: 105-
24.
Schwartz, J. and H. Pitcher. 1989. The Relationship Between Gasoline Lead and Blood Lead in the
United States. J oumal of Official Statistics, 5 :421-431.
Shannon, M. 1996. Personal Communication between Dr. Michael Shannon, Children's Hospital in
Boston, MA and Amy Benson of Abt Associates Inc. November 25.
Shurtleff, D. 1974. Some Characteristics Related to the Incidence of Cardiovascular Disease and Death.
The Framingham Study: An Epidemiological Investigation of Cardiovascular Disease, Section
30, February.
Silbergeld, E.K., J. Schwartz, and K. Mahaffey. 1988. Lead and osteoporosis: mobilization of lead from
bone in postmenopausal women. Environmental Research, 47, 79-94.
Spittler, T. 1993. Personal communication between Tom Spittler, U.S. EPA Region I, and Alice Tome,
Abt Associates Inc.
United States Department of Commerce and U.S. Department of Housing and Urban Development.
1989. American Housing Survey forthe United States in 1989. DOC, Economic and Statistics
Administration and Bureau of the Census and HUD, Office of Policy Development and
Research.
United States Department of Commerce. 1992. Statistical Abstract of the United States. US
Department of Commerce.
United States Department of Commerce. 1993a. Money Income of Households, Families, and Persons
in the United States: 1992. Current Population Reports, Consumer Income, Series P60-184.
United States Department of Commerce (DOC) and U.S. Department of Housing and Urban
Development (HUD). 1993b. American Housing Survey for the United States in 1993. U.S.
Department of Commerce, Economics and Statistics Administration and Bureau of the Census
and U.S. Department of Housing and Urban Development, Office of Policy Development and
Research. Current Housing Reports H150/93. Issued February 1995.
United States Department of Commerce. 1994. Current Population Survey: Lead-based Paint
Awareness Supplement. Public Use file. December.
R.6
~403 EA
-------
United States Department of Commerce (DOC) and U.S. Department of Housing and Urban
Development (HUD). 1995. American Housing Survey for the United States in 1995. U.S.
Department of Commerce, Economics and Statistics Administration and Bureau of the Census
and U.S. Department of Housing and Urban Development, Office of Policy Development and
Research. Issued 1997.
United States Department of Commerce. 1996. Current Population Reports, P25-1130, as cited in
Statistical Abstract of the United States (Washington, DC, U.S. Government Printing Office,
p.20.
United States Department of Commerce and U.S. Department of Housing and Urban Development.
1996. Property Owners and Managers Survey, Multi-housing Unit Properties.
United States Department of Education. 1993. Digest of Education Statistics. US Department of
Commerce.
United States Department of Health and Human Services, Public Health Service, Agency for Toxic
Substances and Disease Registry (USDHHS). 1988. The Nature and Extent of Lead Poisoning
in the United States: A Report to Congress. July.
United States Department of Health and Human Services, Public Health Service, Food and Drug
Administration (FDA). 1991. Fact Sheet: Lead in China Dishes Lawsuit in California; Lead in
Solder Used in Cans, Crystal, and Food Wrappers. November 18.
United States Department of Health and Human Services, Public Health Service, Food and Drug
Administration (FDA). 1992. Statement by Michael R. Taylor, Deputy Commissioner for
Policy, Food and Drug Administration, Public Health Service, Department of Health and Human
Services Before the Ad Hoc Subcommittee on Consumer and Environmental Affairs, Committee
on Government Affairs, U.S. Senate. March 27.
United States Department of Housing and Urban Development (HUD). 1990. Comprehensive and
Workable Plan for the Abatement of Lead-Based Paint in Privately Owned Housing: Report to
Congress. Washington, DC. December.
United States Department of Housing and Urban Development (HUD). 1993. National Survey of Lead-
Based Paint in Housing: Documentation of Analytical Data Files. Prepared for U.S. Department
of Housing and Urban Development by Westat, Inc.. November 30.
United States Department of Housing and Urban Development (HUD). 1995. Guidelines for the
Evaluation and Control of Lead-Based Paint Hazards in Housing. Office of Lead-Based Paint
Abatement and Poisoning Prevention. June. Available from HUD USER (800-245-2691).
United States Department of Housing and Urban Development (HUD). 1996. Regulatory Impact
Analysis of the Proposed Rule on Lead-Based Paint: Requirements for Notification, Evaluation
and Reduction of Lead-Based Paint Hazards in Federally-Owned Residential Property and
R-7
~403 EA
-------
Housing Receiving Federal Assistance. Prepared by ICF for Office of Lead-Based Paint
Abatement and Poisoning Prevention, U.S. Department of Housing and Urban Development.
United States Department of Housing and Urban Development (HUD). 1997. Moving Toward a Lead-
Safe America: A Report to the Congress of the United States; Office of Lead Hazard Control.
Washington, DC. February.
United States Environmental Protection Agency. 1984. A Survey of Literature Regarding the
Relationship Between Measures of IQ and Income. Prepared by ICF, Inc., Report to US EPA,
Office of Policy Analysis, June.
United States Environmental Protection Agency. 1985. Costs and Benefits of Reducing Lead in
Gasoline: Final Regulatory Impact Analysis. Prepared by the U.S. Environmental Protection
Agency, Office of Policy Analysis, Economic Analysis Division. February.
United States Environmental Protection Agency 1986a. Air Quality Criteria for Lead, June, 1986.
United States Environmental Protection Agency. 1986b. Reducing Lead in Drinking Water: A Benefit
Analysis. Prepared by U.S. Environmental Protection Agency, Office of Policy Planning and
Evaluation, Draft Final Report. December.
United States Environmental Protection Agency. 1987. Methodology for Valuing Health Risks of
Ambient Lead Exposure. Prepared by: Mathtech, Inc. Prepared for: Ambient Standards Branch,
Office of Air Quality Planning and Standards, under EPA Contract No. 68-02-4323. December.
United States Environmental Protection Agency (USEP A). 1989. Review of the National Ambient Air
Quality Standards for lead: Exposure Analysis Methodology and Validation. Office of Air
,Quality Planning and Standards. Research Triangle Park, N.C. June.
United States Environmental Protection Agency. 1990. Review of the National Ambient Air Quality
Standards for Lead: Assessment of Scientific and Technical Information. OAQPS Staff Paper,
Air Quality Management Division, Research Triangle Park, NC., December.
United States Environmental Protection Agency (USEPA). 1991. U.S. Environmental Protection
Agency Strategy for Reducing Lead Exposures. February 21.
United States Environmental Protection Agency (EPA). 1992. Applicability ofRCRA Disposal
Requirernents to Lead-based Paint Abatement Wastes. Office of Pollution Prevention and
Toxics. June.
United States Environmental Protection Agency (EP A). 1994a. Estimating Costs for the Economic
Benefits of RCRA Noncompliance. Office of Regulatory Enforcement. Obtained by Abt
Associates from DPRA, Incorporated, St. Paul, MN.
United States Environmental Protection Agency (USEPA). 1994b. Lead; Requirements for Lead-based
Paint Activities, Proposed Rule, Federal Register, Friday, September 2.
R-a
t403 EA
-------
United States Environmental Protection Agency (USEPA). 1995a. Report on the National Survey of
Lead-Based Paint in Housing. Base Report. USEP A, Office of Pollution Prevention and Toxics.
EPA 747-R95-003. April.
United States Environmental Protection Agency (EPA). 1995b. Lead-Based Paint Risk Assessment
Model Curriculum. Prepared for the Chemical Management Division by The National Center for
Lead-Safe Housing. June. A VI-0006700-1 and A VI-0006700-2.
United States Environmental Protection Agency (EP A). 1995c. Guidance on Identification of Lead-
Based Paint Hazards. Federal Register 60 (175): 47248-47256, September 11.
United States Environmental Protection Agency. 1996. TSCA Title IV, Sections 402(a) and 404:
Target Housing and Child Occupied Facilities Final Rule, Regulatory Impact Analysis.
United States Environmental Protection Agency (USEP A). 1997. Risk Assessment for the Section 403
Rulemaking; USEP A, Chemical Management Division, Office of Pollution Prevention and Toxics.
United States Office of Management and Budget 1988. Regulatory Impact Analysis Guidance
Regulatory Program of the United States Government.
United States Office of Management and Budget. 1992. Guidelines and Discount Rates for Benefit-
Cost Analysis of Federal Programs, Circular No A-94, Revised. October 29, 1992.
United States Office of Management and Budget 1996a. Economic Analysis for Federal Regulations
under Executive Order 12866, January 11, 1996.
United States Office of Management and Budget 1996b. Guidelines and Discount Rates for Benefit-
Cost Analysis of Federal Programs, Circular No A-94, Appendix C, Revised. February 6,
1996.
United States Small Business Administration. 1993. Small Business Thresholds by 4-digit Standard
Industrial Classification.
Wallsten, T.S. and Whitfield, R. G. 1986. Assessing the Risks to Young Children of Three Effects
Associated with Elevated Blood-lead Levels. Report by Argonne National Laboratory. Report
No. ANL/AA-32. Sponsored by the U.S. EPA Office of Air Quality Planning and Standards.
Westat, Inc. 1995. Report on the National Survey of Lead-Based Paint in Housing. Prepared for U.S.
EPA. March 17.
9403 EA
R-9
-------
:.. .
~O-oQ)
-0:"(1»
~, :.. >.-
\II (1).-'"
... (1)(.)
(JC(.)Q)
...ftS(1)'i-
....:..Q)
~.- U) 'C
-oftS(/)
tJ) (1) ~.-
E~ E E
!O(1)Q)
.- :.."""'"
.... c.'-'-
o(1)(1)Q)
c.c=:5
"'-'+-u.-
~_._~
.J .- .-
.. ~ ..: L-
a>..., (1) 0
...C-O
:.. ...
~(1)oC
E E :...2
...(1)~"'"
a> (.) 0'-
c.ftS ~-o
- c
.. Q.tno
o e.5 (.)
C
-------
U.S. DEPARTMENT OF COMMERCE
Technology Administration
National Technical Information Service
Springfield, VA 22161 (703)605-6000
------- |