United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
Air
Final Report
EPA-452/R-98-002
July 1998
oEPA
ECONOMIC IMPACT AND
REGULATORY FLEXIBILITY
ANALYSES OF THE FINAL
ARCHITECTURAL COATINGS
VOC RULE
U.S. Environmental Protection Agency
Region 5. Library (PL-12J)
77 West Jackson Boulevard, 12th floor
Chicago, IL 60604-3590
-------
: TABLE OF CONTENTS
U
— Section
J Executive Summary . ix
V
1 Introduction, Regulatory Background, and
Industry Profile 1-1
1.1 Introduction 1-1
1.2 Regulatory Background 1-2
1.2.1 Regulatory Structure 1-3
1.3 Industry Profile 1-5
1.3.1 Commodities and VOC Content 1-5
1.3.2 Demand for Architectural Coatings . . . 1-8
1.3.2.1 Conceptual View of Coating
Decision 1-8
1.3.2.2 Substitution Effects 1-10
1.3.2.3 Aggregate Demand 1-11
1.3.2.4 Coating Users 1-11
1.3.3 Production of Architectural Coatings . 1-12
1.3.3.1 Raw Material Inputs 1-12
1.3.3.2 Formulations 1-15
1.3.3.3 Manufacturers' Substitution
Options and New
Technologies 1-17
1.3.4 Industry Conditions 1-20
1.3.4.1 Shipments and Manufacturer
Specialization 1-20
1.3.4.2 Company Size and Industry
Structure 1-23
2 Costs of Proposed Regulation for Architectural
Coating Producers 2-1
2.1 Background 2-1
2.2 Overview of Response Options 2-2
2.2.1 Supply 2-2
2.2.2 Demand 2-4
2.3 Cost Analysis 2-5
2.3.1 Costs of Reformulation 2-5
2.3.1.1 Product-Level Reformulation
Cost Estimates 2-8
2.3.1.2 National Reformulation Costs . 2-14
ill
-------
TABLE OF CONTENTS (continued)
Section Page
2.3.2 Exceedance Fee Provision 2-20
2.3.3 Product Withdrawal 2-23
2.3.4 "Best-Response* Analysis 2-24
2.3.5 Tonnage Exemption 2-30
2.4 Cost Analysis Uncertainties 2-30
2.4.1 Upward Bias 2-32
2.4.2 Downward Bias 2-32
2.4.3 Unknown Directional Effects 2-33
3 Architectural Coatings Market Analysis 3-1
3.1 Market Effects of Firm Responses to
Regulation 3-1
3.1.1 Model Execution and Results 3-2
3.1.1.1 Baseline 3-2
3.1.1.2 Quantifying Market Shocks . . 3-3
3.2 Architectural Coatings Industry Employment
Impacts 3-10
4 Traffic Coating User Costs 4-1
4.1 Truck Replacement Cost Method 4-2
4.2 Equipment Retrofit Method 4-5
4.3 National Incremental Cost Calculation .... 4-5
5 Social Cost-Effectiveness Analysis 5-1
5.1 Conversion of Impacts to Current Dollars . . . 5-5
6 Small Business Impact Analysis 6-1
6.1 Background and Affected Entities 6-2
6.1.1 Potentially Affected Entities 6-3
6.1.2 Regulatory Requirements 6-5
6.2 Analysis 6-6
6.2.1 Baseline Market Presence of Small
Architectural Coatings Producers . . . 6-6
6.2.2 VOC Content of Small Business Products:
Technology and Specialization Effects . 6-11
6.2.3 Costs Associated With Regulatory
Compliance 6-17
6.2.4 Reformulation Cost Impact Estimates . . 6-18
6.2.4.1 Small Business Impacts Under
"Reformulation-Only" Option . 6-22
IV
-------
TABLE OF CONTENTS (continued)
Section Page
6.2.5 Cost Impacts Across Market Segments . . 6-23
6.2.6 Average Cost Impacts for Small
Company 6-25
6.2.7 Potential Factors Mitigating Small
Business Impacts: Exceedance Fee,
Withdrawal, and Tonnage Exemption . . . 6-27
6.2.7.1 Fee and Withdrawal Options . , 6-27
6.2.7.2 Tonnage Exemption 6-29
6.3 Regulatory Alternatives to Reduce Impacts . . 6-31
6.3.1 Selection of VOC Content Limits and
Coating Categories 6-33
6.3.2 Low-Volume Exemption Option 6-34
6.3.3 Exceedance Fee Compliance Option . . . 6-38
6.3.3.1 Exceedance Fee Rate 6-41
6.3.4 Extended Compliance Time for Small
Businesses 6-44
6.3.5 Compliance Variances 6-45
6.3.6 Selection of Recordkeeping and Reporting
Requirements 6-47
6.4 Small Business Impact Summary 6-49
7 Epilogue 7-1
7.1 New Product Categories 7-1
7.2 Categories with Higher VOC Content Limits . . 7-6
7.3 Summary 7-7
8 References 8-1
Appendices
A Market Definition, Demand Estimation,
and Data A-l
B Summary of Reformulation Cost Estimates from
Public Comments B-l
C Calculation of Regulation-Induced Costs when
Reformulation Normally Occurs at Fixed Time
Intervals C-l
D Methodology for Computing Market and Welfare
Adjustments D-1
-------
LIST OF FIGURES
Number Page
1-1 Comprehensive classification of emissions
from consumer and commercial products 1-2
1-2 Inputs generally used in the manufacture
of a solventborne coating 1-16
1-3 Inputs generally used in the manufacture
of a waterborne coating 1-16
1-4 Approximate volume relationships of coating
ingredients 1-18
1-5 Location of manufacturing establishments
in the paints and allied products
industry in 1987: SIC 2851 1-27
2-1 Basic stages of architectural coating reformulation
(prototype firm and product) 2-4
2-2 Fee versus reformulation 2-22
4-1 Cost schedules with and without accelerated
replacement 4-4
4-2 Replacement cost schedules with and without
equipment retrofit 4-6
6-1 Chain of ownership 6-4
VI
-------
LIST OF TABLES
Number Page
1-1 Average VOC Content for Architectural Coatings
Expected to Be Covered by Proposed Regulation . . . 1-7
1-2 Consumers of Architectural Coatings 1-12
1-3 Percentage of Solvent in Conventional and
Reformulated Organic Solventborne Coatings 1-19
1-4 Value Shipped of Potentially Regulated Paint
and Allied Products: 1981 Through 1991 1-21
1-5 Number of Companies, Establishments, and
Producer Specialization—Paint and Allied
Products: 1987 1-24
1-6 Commodity Production in 1982: Paint and
Allied Products 1-25
1-7 Large Firm Dominance and Numbers of Companies
and Establishments in the Paint and Allied
Products Industry: 1987 1-27
1-8 Number of Companies and Establishments in
the Coatings Industry, Selected Years,
1972-1991 1-28
1-9 Recent Acquisitions in the Coatings Industry .... 1-29
2-1 Table of Standards 2-6
2-2 National Costs: Reformulation-Only Scenario .... 2-19
2-3 Architectural Coatings Market Segments Baseline
Data 2-27
2-4 Best-Response Option Analysis—Survey Population:
Fee = $2,200/ton 2-28
2-5 Reformulation Cost Analysis Uncertainties 2-31
3-1 Regulatory Effects on Architectural Coatings
Market Output and Prices 3-6
3-2 Architectural Coatings Market Welfare Effects . . . 3-9
3-3 Estimated Employment Effects 3-11
VI1
-------
LIST OF TABLES (continued)
Number Page
4-1 National Incremental Cost of Traffic Coating
Equipment Replacement and Retrofits—Medium
Stripers 4-7
4-2 National Incremental Cost of Traffic Coating
Equipment Replacement and Retrofits—Large
Stripers 4-9
4-3 National Incremental Cost Summary for Traffic
Coating Equipment 4-11
5-1 Social Cost-Effectiveness Estimates 5-4
5-2 Conversion of Summary Impacts to 1996 Dollars . . . 5-5
6-1 Small Business Presence in the Architectural
Coatings Market: Survey Population 6-7
6-2 Baseline VOC Content 6-12
6-3 Specialization-Based Average VOC Content: Small
Business Products 6-15
6-4 Small Business Costs by Market Segment:
Reformulation Option Only 6-19
6-5 Average Regulatory Impact by Firm Size—
"Reformulation-Only* Scenario 6-26
6-6 Average Regulatory Impact for Small Companies—
"Best-Response* Scenario 6-28
7-1 Table of Standards: Interim vs. Final 7-4
Vlll
-------
EXECUTIVE SUMMARY
Under Title I of the Clean Air Act of 1990, the U.S.
Environmental Protection Agency (EPA) is developing
regulations to reduce volatile organic compound (VOC)
emissions from various consumer and commercial products. One
of the first categories of consumer and commercial products to
be regulated is architectural coatings. This report analyzes
the costs and economic impacts of the final architectural
coatings rule.
The general purpose of the regulation is to reduce the
flow of VOCs into the atmosphere from the use and disposal of
architectural coating products. These emissions are
distinguished from the manufacturing-related emissions that
are controlled by other forms of regulation (as are emissions
to land and water).
VOC emissions are regulated because of their contribution
to the formation of ground-level ozone. Elevated levels of
ozone degrade air quality and pose a variety of health risks
to exposed populations.
ES.l COMPLIANCE ACTIONS
The regulation imposes a set of standards for VOC content
for individual architectural coating products. Products that
exceed the limits imposed by these standards must either be
brought into compliance with the limits, have an exceedance
fee assessed on the product's VOC content above the limit, or
ix
-------
be withdrawn from the market. These actions, however, can be
avoided for products subject to the small tonnage exemption.
ES.2 COMPLIANCE COSTS
The number of compliance actions was estimated using
survey data on VOC content and sales volumes for almost
5,000 architectural coating products manufactured by
116 companies. The surveyed products constitute about
three-quarters of industry output. The survey data were used
to estimate the compliance activity for the products and
manufacturers not covered in the survey and is thereby the
basis for the national estimate of costs.
Initially, the regulatory impacts were viewed in a very
restrictive light, assuming that reformulation down to the
standards is the only option available to producers. The
aggregate costs of this restrictive option were then computed
to give a benchmark measure of regulatory costs under a
restrictive set of conditions. The costs in Table ES-1
present both the initial one-time expenditure for the
reformulations and the costs expressed in annualized terms.
TABLE ES-1. NATIONAL COST FOR ARCHITECTURAL COATINGS
PRODUCERS-REFORMULATION-ONLY
Total Initial Expenditure Total Annualized Cost
($1991) ($1991)
204.0 million 34.2 million
The analysis was expanded by progressively shedding the
restrictive assumptions of forced reformulation. First, the
exceedance fee option was incorporated, taking into account
that producers may choose to pay an exceedance fee rather than
reformulate if it is a less costly alternative for them.
-------
Then, the least-cost compliance option (fee or reformulation)
was compared with benefit streams (net revenues) to determine
if the least-cost option is also profitable. If the value of
the benefit stream is less than the cost of compliance, firms
are assumed to remove the products from the market as a
best-response strategy. Alternative response options reduce
the cost of the regulation by approximately 20 percent for the
architectural coating producers included in the survey. Cost
reductions are likely to be greater for the nonsurvey
population and are further reduced when market-level responses
are factored in (see below). Most of the cost savings is
attributable to adopting the exceedance fee, which is
projected to be the compliance option for a number of products
that are either very small in volume or have a VOC content
relatively close to the limit. Because the fee is generally
adopted for relatively small sources of VOC "exceedance," the
effect on VOC emissions reductions is projected to be
relatively small.
ES.3 MARKET EFFECTS
The compliance actions lead to a reallocation of
society's resources toward VOC controls, which imposes
opportunity costs directly on the producers and indirectly on
other members of society as producers act, markets respond,
and prices and output change. The purpose of the
architectural coatings market analysis is to characterize the
reallocation of resources and quantify them in
dollar-denominated terms to assess the distribution of costs
and economic impacts of the regulation.
The collective effect of some producers removing
unprofitable products and some producers bearing a per-unit
fee on output will contract the aggregate supply of
XI
-------
architectural coatings and lead to changes in market prices
and quantities. The optimal best-response actions and
resulting market outcomes will determine how the welfare costs
of the policy are distributed across producer groups,
consumers, and the government sector.
Several scenarios were modeled for the standards. In
general, market model results indicate a very small change in
baseline market conditions as a result of the regulation.
This derives from the expectation that aggregate costs of the
regulation are a small share of aggregate industry costs.
However, because there is a high degree of producer
heterogeneity within the architectural coatings sector, the
costs for some producers may be large. The distribution of
impacts across affected parties is presented in Table ES-2.
TABLE ES-2. MARKET IMPACTS SUMMARY
Aggregate Welfare Effects on.... (MM $1991)
Architectural coating producers -22.0
Architectural coating consumers -4.3
Government (fee receipts) +4.0
Net social welfare effect ("social cost*) -22.3
A portion of the cost for architectural coating producers
is passed on to consumers in the form of higher prices, which
lowers their welfare. An important impact to consider is the
effect of the fee payments. While these payments constitute
losses for the producers paying the fee (less the amount they
are able to pass on to the consumer via higher prices), these
fee payments are simply transfer payments to the government
and therefore do not constitute a net increase in social
costs. In other words, while the fee serves as a private cost
for firms that do not reduce VOCs to the statutory limit and a
xii
-------
continuing incentive for producers to reduce VOCs to the
limit, it does not constitute an allocation of society's
resources to a particular use as, in contrast, the allocation
of resources for reformulation does.
The net social cost estimate is substantially lower than
the annualized cost estimate under the reformulation-only
scenario described above. The market analysis demonstrates
the potential for substantial cost savings due to adopting the
fee alternative and how this cost savings is likely to accrue
especially to producers of small volume products. Moreover,
this cost savings is not expected to have a significant impact
on undercutting aggregate emissions reduction targets.
ES.4 TRAFFIC COATING USER COSTS
The economic analysis up to this point has focused
entirely on the primary impacts of the regulation, those borne
directly by producers in the architectural coatings industry
in the form of higher costs and indirectly by the consumers of
architectural products in the form of higher prices. The
driving force of those impacts is the requirement that, except
for products subject to the tonnage exemption, noncompliant
products must either be reformulated to a compliant VOC level,
be subject to a fee on the excess VOCs over the allowable
level, or be withdrawn from the market. However, this
analysis considered a type of secondary impact, one that is
caused by the costs that users of a newly compliant product
must incur to purchase the special equipment necessary to
apply the compliant coating. The secondary impact analysis
focuses exclusively on users of traffic marking paints,
primarily government entities such as state transportation
departments, for whom the costs of switching application
XI11
-------
equipment ("striper" trucks) are thought to be potentially
significant.
Traffic coating user costs are summarized in Table ES-3.
Costs are estimated as the incremental cost associated with
the accelerated replacement of striper trucks and are
expressed both in terms of the present value of the one-time
acceleration ($53.2 million total) and on an annualized basis
($3.7 million).
TABLE ES-3. NATIONAL INCREMENTAL COST SUMMARY FOR TRAFFIC
COATING EQUIPMENT ($1996)
Striper Type Present Value of Annualized
Cost Cost
Medium (see Table 4-1) $42,844,912 $2,999,144
Large (see Table 4-2) $10,393,011 $727,511
Total $53,237,923 $3,726,655
ES.5 SOCIAL COST-EFFECTIVENESS
The social cost estimates from the market analysis and
the estimate of traffic coating user costs can be used to
compute measures of the social cost-effectiveness of the
regulation. The distinction of "social" cost-effectiveness is
made to illuminate the fact that the costs evaluated are the
net costs imposed on society (i.e., the net welfare costs
estimated in the architectural coatings market plus the
resource costs incurred by traffic coating users to switch
application equipment).
Cost-effectiveness results are summarized in Table ES-4.
Emission reduction effects of the regulation are estimated by
taking the national target for VOC emission reductions from
xiv
-------
TABLE ES-4. SOCIAL COST-EFFECTIVENESS SUMMARY
Social Cost Estimated Emissions Social Cost per Mg
($1991) Reduction (Mg) ($1991)
25.6 million 103,471 247
architectural coatings and modifying that total to account for
market responses (fee adoption and changes in output levels).
This estimate allows for an evaluation of
cost-effectiveness implications of the fee option. Allowing
the fee reduces social costs by about $12 million but foregoes
about 1,802 Mg of emissions reduction, about 1.7 percent of
the targeted reductions. Dividing the cost savings by
foregone reductions approximates the marginal social cost of
the foregone reductions. This figure is $6,580/Mg, which is
substantially higher than the $247/Mg average social
cost-effectiveness measure reported above. This difference
indicates that the fee's main effect is to reduce the very
most expensive emission reductions without substantially
undercutting the achievement of emissions reduction.
For external reporting purposes, the economic impacts are
reported in 1996 dollars. Costs are converted from the base
year used in the analysis (1991) to 1996 using the Gross
Domestic Product (GDP) price deflator. The conversion results
are presented in Table ES-5.
TABLE ES-5. CONVERSION OF SUMMARY IMPACTS TO 1996 DOLLARS
Impact Estimate $1991 $1996
Net social cost $25.6 million $29.2 million
Net social cost per Mg $247/Mg $282/Mg
of emissions reduction
xv
-------
ES.6 SMALL BUSINESS IMPACTS AND REGULATORY FLEXIBILITY
ANALYSIS
The potential for significant impacts on small businesses
of the regulation arises from two primary sources:
• Products made by small producers, on average, have a
higher VOC content than the industry average.
• The costs of reformulating products to comply with the
regulation are fixed and thereby impose higher average
costs on small volume coatings.
The first problem is related to small producers' tendency
to specialize in coatings categories that are naturally higher
in VOC content and to their tendency to concentrate in the
"high-VOC" end of the distribution of products within a given
category. Thus, the potential for disproportionate impacts of
VOC reduction regulation on small businesses follows partly
from the fact that small businesses contribute a
disproportionate amount of the aggr«gate VOC emissions that
are targeted for reduction.
The second problem follows from the nature of
reformulation costs. A coating's formula is the product of an
intellectual capital investment, much like the development of
a drug or a computer software product. The cost of the
investment follows directly from the level of effort necessary
to revise the formula to meet both the VOC standards imposed
by the regulation and performance standards imposed by the
marketplace. This level of effort is essentially independent
of the quantity of the product that is eventually sold.
Therefore, the relative impacts on smaller volume products is,
by definition, greater.
The data used in this analysis suggest that these two
primary factors are relevant in the case of small
architectural coating producers. The average VOC content of
xvi
-------
the products made by the small business producers in the
survey is 75 percent higher than the average VOC content of
all products combined (see Table ES-6). A little over half of
the difference in the averages is attributed to the
specialization of small producers in high-VOC content product
categories, with the remainder attributed to the tendency for
small businesses to produce higher VOC products within each
product group. Moreover, the average product volume of
products made by small businesses is less than 20 percent of
the average product volume for the entire survey population,
implying much larger average reformulation costs (see
Table ES-7). Thus, without mitigating factors, the impacts on
some small businesses are potentially significant.
TABLE ES-6. BASELINE VOC CONTENT
Size
Category*
All products
Small business
products
VOC
Emissions
(Mg)
344,059
21,431
Sales
(kL)
1,853,623
65,914
Average VOC
Content
(g/L)
186
325
a The survey had 116 respondents and 36 of those identified themselves as
having under $10 million in annual sales. Twelve survey respondents did
not report company size.
Source: Industry Insights. Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey. Prepared for
National Paint and Coatings Association in cooperation with the
AIM Regulatory Negotiation Industry Caucus. Final Draft Report.
1993.
At proposal, the Agency included specialized coating
categories and limits designed to preserve niche product
markets. To evaluate whether further steps were still needed
to accommodate niche market coatings, the Agency requested
that commenters identify any additional specialty coatings
xvn
-------
that would not comply with applicable VOC content
requirements. The Agency also requested comment on whether to
include an "exceedance fee" in the final rule, which would
allow companies the option of paying a fee, based on the
amount that VOC content limits are exceeded, instead of
achieving the limit. In addition, the Agency requested
comment on the concept of a low volume cut-off, under which a
coating might be exempt from regulation. In the final rule,
the Agency has included the exceedance fee compliance option
and a provision that enables each manufacturer to claim as
exempt a specified amount of VOC per year (known as the
tonnage exemption). Also, in response to public comments, the
Agency created seven new niche product categories and
increased the VOC content limits for four product categories
in the final rule. The Agency also added an extended period
for compliance after promulgation to allow additional time for
reformulations. These provisions are designed to mitigate
rule impacts on small businesses' low production volumes and
to allow for the preservation of several niche markets.
However, based on the limited data available to the Agency,
only the mitigating impact of exceedance fees can be
quantified.
The analysis shows that, when reformulation is the only
option for compliance, the cost/revenue ratio is estimated to
be 2.5 percent on average (see Table ES-7). When the
alternative compliance options of the exceedance fee or
product withdrawal are considered, the ratio decreases to
2.0 percent (see Table ES-8). This ratio would decrease
further if the cost effects of the additional niche product
categories, use of the tonnage exemption, and reduction in
cost to reformulate due to resin supplier assistance could be
specifically quantified.
XVlll
-------
TABLE ES-7. AVERAGE REGULATORY IMPACT BY FIRM SIZE-
"REFORMULATION-ONLY* SCENARIO*
Industry Small Firm
Average Average
Revenue" ($1991) 38,990,000 4,614,000
Number of products" 42.4 27.5
Number of products facing major 9.9 7.8
reformulation0
Annualized reformulation cost5 ($1991) 144,272 113,669
Ratio of annualized reformulation 0.4 2.5
cost to revenues (percent)
* The survey has 116 respondents and 36 of those identified themselves as
having under $10 million in annual sales. Twelve survey respondents did
not report company size.
b Data for revenues and products per firm were based on data reported in
Table 6-1. The number of products per firm is based on the total number
of products for which quantity data are available.
c This number represents two-thirds of the products over the 1998 TOS.
Industry experts estimate that approximately two-thirds of the products
with VOC contents exceeding the TOS limits face a 'major' reformulation.
d Annualized cost of reformulation is the number of major reformulations
multiplied by the annualized reformulation cost estimate per product of
$14,573 ($1991).
Source: Industry Insights. Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey. Prepared for
National Paint and Coatings Association in cooperation with the
AIM Regulatory Negotiation Industry Caucus. Final Draft Report.
1993.
The Agency prepared analyses to support both the proposed
and final rules that are equivalent to those required by the
Regulatory Flexibility Act (RFA) as modified by the Small
Business Regulatory Enforcement Fairness Act of 1996 (SBREFA).
The Agency undertook these analyses because of the large
presence of small entities in the architectural coatings
industry and because the initial impact analysis indicated
that there could be a significant economic impact on a
substantial number of small entities if mitigating regulatory
options were not adopted for the rule. After evaluating
public comment on the proposed mitigating options, EPA made a
xix
-------
TABLE ES-8. AVERAGE REGULATORY IMPACT FOR SMALL
COMPANIES-'BEST-RESPONSE* SCENARIO
Compliance
Strategy
Reformulate
Fee
Withdrawal
Total
Average percent
Percent of All
Constrained
Survey Products
Selecting
Option
60.5%
35.5%
4.0%
100.0%
of sales
"Expected*
Number of
Products
Selecting
Strategy"
4.7
2.8
0.3
7.8
Average
Compliance
Cost per
Product
($1991)
14,573
7,197b
12,705C
11,879
Compliance
Cost
($1991)
68,767
19,936
3,955
92,658
2.0%
a Equals average number of constrained products for small companies (7.8)
multiplied by percentage of all constrained products in the survey
selecting each strategy.
b Average fee cost computed by taking the average fee rate ($0.084/L),
multiplying by the average size per small company product (65,914 L),
and adding the recordkeeping cost per product of $590.
c Equals the average value of foregone profits for the 46 surveyed
products that select the fee as the best-response strategy.
number of changes to the proposed rule to further mitigate the
rule's small business impacts. As a result, the Agency
believes that it is highly unlikely that the rule will have a
significant economic impact on a substantial number of small
entities. However, in light of the Agency's inability to
quantify the effect of the mitigating options, the EPA has
elected to conduct a regulatory flexibility analysis and to
prepare a SBREFA compliance guide to eliminate any potential
dispute on whether EPA has fulfilled SBREFA requirements.
ES.7 EPILOGUE
Because regulatory development is an evolving process,
the final Table of Standards for VOC content limits differs
slightly from the interim Table of Standards used in the
analysis reported here. The main difference between the two
xx
-------
sets of standards is the addition of seven new categories in
the final standards and an increase in the VOC content limits
for three categories.
By and large, new categories were added to accommodate
specialty products that were previously included in other
categories with lower VOC limits. As a result, some products
that would be over the limit in the previous category, thereby
necessitating a compliance action (reformulate, fee payment,
withdrawal), are no longer constrained by the regulation.
Therefore, in most cases the addition of the new categories
reduces the number of required compliance actions and, as a
result, also cuts compliance costs. In addition, raising the
VOC limits in the other categories reduces compliance actions
and costs as well.
However, one of the new product categories, concrete
curing and sealing (CCS) compounds, applies to products that
were considered outside of the regulated universe in the
economic analysis presented in this report. Therefore, the
costs associated with the compliance actions required for
those products are not estimated in the analysis. If they
were, the cost estimate would be larger.
Data were available to approximate cost effects for only
two of the seven new product categories. One of these was the
CCS category, which allowed for an estimate of the
corresponding increase in costs just described. The other new
product category for which data were obtained is zone
markings. The original 1991 emissions inventory provided data
to analyze the cost reductions due to the increase in content
limits for three product categories. Taken together, the
available data allowed for quantification of a $580,000
(1991 dollars) net increase in the estimate of annual social
costs. However, this increase in cost must be considered
against the unquantified decrease in costs from the expected
fall in compliance activity in the five other new categories
xxi
-------
for which data were unavailable. Without additional data, it
is difficult to conclude whether the cost reductions from
those categories will together outweigh the net cost increases
quantified. Given that the social cost effects quantified
here are less than 3 percent of the total estimated social
costs of the regulation, factors that reduce (or reverse the
sign) of these costs lead to the conclusion that the total
social cost estimate is not greatly affected by the
differences between the interim standards used in the analysis
and the final standards issued in the rule.
xxii
-------
SECTION 1
INTRODUCTION, REGULATORY BACKGROUND, AND INDUSTRY PROFILE
1.1 INTRODUCTION
Under Title I of the Clean Air Act of 1990, the U.S.
Environmental Protection Agency (EPA) is developing
regulations to reduce volatile organic compound (VOC)
emissions from various consumer and commercial products. One
of the first categories of consumer and commercial products to
be regulated is architectural coatings.
This report analyzes the economic impacts of the final
architectural coating regulation. Section 183(e)(l)(B) of the
Clean Air Act Amendments of 1990 defines a consumer or
commercial product as
any substance, product (including paints, consumer
and commercial products, and solvents), or article
(including any container or packaging) held by any
person, the use, consumption, storage, disposal,
destruction, or decomposition of which may result in
the release of volatile organic compounds.
Thus, the general purpose of the regulation is to reduce the
flow of VOCs into the atmosphere from consumption and disposal
of products that contain VOCs. Figure 1-1 shows the
dissipative emissions and the disposal emissions into the air
that are the target of this regulation.1 These emissions are
1-1
-------
agro-chtmleal*
InduslriaVantrgy
Mt I (fit Is us* t consumption (pollution
takodfetf in coniMrcial and eontumtr goods)
application to soils of agro-chi nlcils
Araa-sourc*
agricultural tmissions
Point- and ar«a-sourct
production (missions
Oisslpalivt
tmissions
Disposal emission*
AIR, LAND. AND WATER
Figure 1-1. Comprehensive classification of emissions from
consumer and commercial products.
Source: Adapted from Stigliani, William M. Chemical Emissions from the
Processing and Use of Materials: The Need for an Integrated
Emissions Accounting System. Ecological Economics £(4):325-341.
1990. (Figure 2).
distinguished from the manufacturing-related emissions that
are controlled by other forms of regulation. The regulatory
structure is presented here followed by an overview of the
architectural coatings industry.
1.2 REGULATORY BACKGROUND
Section 183(e)(3)(A) directs the EPA to list categories
of consumer or commercial products that account for at least
1-2
-------
80 percent of VOC emissions on a reactivity-adjusted basis in
ozone nonattainment areas. The EPA divided this category list
into four groups and established priorities for regulation.
Architectural coatings is in the first group of categories to
be regulated.
The design of regulatory strategies to reduce VOCs
emitted by architectural coatings is shaped in specific ways
by the Clean Air Act as amended. Two components of the
legislation are of particular importance:
• determining regulated entities and
• establishing best available controls.
Regulations developed under Section 183(e) may be imposed
only with respect to "manufacturers, processors, wholesale
distributors, or importers of consumer or commercial products
for sale or distribution in interstate commerce in the United
States" or certain entities that supply such products to the
former Sections 183(e)(1) (C) and 183(e) (3) (B) . The definition
of regulated entities excludes retailers and users.
The regulations affecting architectural coatings will
require best available controls. The EPA Administrator, on
the basis of "technological and economic feasibility, health,
environmental, and energy impacts," will determine the desired
degree of emissions reduction that
is achievable through the application of the most
effective equipment, measures, processes, methods,
systems or techniques, including chemical
reformulation, product or feedstock substitution,
repackaging, and directions for use, consumption,
storage, or disposal. (Section 183[e][l]).
1.2.1 Regulatory Structure
One hundred sixteen architectural coatings manufacturers
responded to a survey conducted by the National Paint and
1-3
-------
Coatings Association for products manufactured and their VOC
contents.2 The Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey (the survey)
provides VOC content information and 1990 sales quantities by
product. Based in part on these data, EPA is promulgating VOC
content limit standards, which manufacturers and importers
will be required to meet in 1999. Once the regulation becomes
law, manufacturers and importers of architectural coatings
subject to the regulation must limit the VOC content per liter
of coating to the standards specified for each coating product
they manufacture. The EPA has included an option of allowing
manufacturers and importers to choose to pay an exceedance fee
instead of meeting the limit for a particular product
category. Another option manufacturers and importers have is
to use a tonnage exemption to claim a set amount of product as
exempt from VOC limits. The VOC content limits are presented
in the Table of Standards (TOS) for 1999 in Section 2
(Table 2-1). The limits specified in this table were used in
this economic impact analysis. They cover all the major
architectural coatings categories as well as certain special
purpose coating products for which a less stringent limit is
granted in order for the coating to adequately perform its
designed purpose (e.g., high-temperature coatings).
Architectural coatings manufacturers who choose to pay a
fee on their products that do not meet the standards will pay
the fee on the VOC content of the product that is in excess of
the limit." The fee rate is $2,500 (1996 dollars, adjusted to
$2,200 in 1991 dollars) per metric ton (Mg) of excess VOCs.
Fees will be paid semi-annually and will be placed in a
"special fund* specified under Section 183(e). If EPA is able
•Excess VOCs are defined as the maximum VOC content of the coating,
as applied, in grains per liter of coating, less water and exempt compounds,
minus the applicable VOC standard.
1-4
-------
to obtain these funds through a subsequent Congressional
appropriation they may be used by the Administrator to support
the administration of the regulation or to promote additional
VOC emission reductions from architectural coatings through
technological development grants, award programs, or other
means.
This report includes an overview of the architectural
coatings industry, products, and technologies and an analysis
of the economic impacts on the affected entities and the
industry as a result of the TOS VOC content limits, exceedance
fees, and tonnage exemption. An economic model of the
architectural coatings industry is developed to obtain
estimates of the potential price and quantity changes
associated with the regulation. In addition, a Regulatory
Flexibility Analysis is conducted, which estimates the impacts
of the regulation on small businesses and presents
alternatives that may be implemented to mitigate those
impacts.
1.3 INDUSTRY PROFILE
This profile of the architectural coatings industry
describes commodities and VOC content, demand for
architectural coatings, production of architectural coatings,
and industry conditions.
1.3.1 Commodities and VOC Content
The "architectural coatings* regulation applies primarily
to products that the U.S. Census Bureau also categorizes as
architectural coatings, but some products in the Census
categories of special purpose coatings and miscellaneous
allied paint products are affected as well.3 Unless otherwise
indicated, the term "architectural coatings* is used
1-5
-------
throughout this report to indicate the entire group of
regulated products. Product categories covered under the
regulation are listed in Table 1-1.* The products are grouped
into the three Census categories in which they are found.b As
indicated, the largest quantity of regulated coatings is
included in the architectural coatings category, but some
coatings are classified with the special purpose and allied
paint products categories, which also include other products
not covered by this regulation such as marine paints and
putty.
Examples of Census-defined architectural coatings, all of
which are represented in Table 1-1, include exterior and
interior organic solventborne and waterborne tints, enamels,
undercoats, clear finishes, stains, and architectural
lacquers. These coatings are used for general purpose on-site
application to residential, commercial, institutional, and
industrial structures. They are intended for ordinary use and
exposure and provide protection and decoration.
Special purpose coatings are similar to architectural
coatings in that they can be classified as stock or shelf
goods, rather than formulated to customer specifications, as
are OEM coatings. The difference is that they are formulated
for special applications or environmental conditions such as
extreme temperatures, chemicals, fumes, fungi, or corrosive
conditions.
VOC content varies substantially between specific types
of coating products. Most of this variety is due to the type
of solvent used in the coating and the ratio of the solvent to
other ingredients in the formulation. Based on the 1990
bSee Appendix A for a detailed explanation of products for regulation
and their corresponding Census classification.
1-6
-------
TABLE 1-1.
AVERAGE VOC CONTENT FOR ARCHITECTURAL COATINGS TO
BE COVERED BY REGULATION
Sales-Weighted Average VOC Content
(g/L)
Product Category
Organic Solvent
Waterborne
Architectural coatings
Exterior flat architectural coatings
Exterior nonflat architectural coatings
Interior flat architectural coatings
Interior nonflat architectural coatings
Semitransparent stains
Opaque stains
Undercoaters
Primers
Sealers
Waterproofing sealers, clear
Waterproofing sealers, opaque
Quick dry undercoaters, primers, and
sealers
Bituminous coatings
High performance architectural coatings
Roof coatings
Lacquer
Varnish
Special purpose/industrial maintenance
Coatings
Swimming pool coatings
Dry fog coatings
Mastic texture coatings
Metallic pigmented coatings
Fire retardant coatings
Antigraffiti
Concrete curing compounds
Form release compounds
Graphic arts coatings
High-temperature coatings
Industrial maintenance coatings
Multicolored coatings
Pretreatment wash primers
Sanding sealers
Shellacs
Traffic marking paints
Allied paint products
Below ground wood preservatives
Semitransparent wood preservatives
Clear wood preservatives
Opaque wood preservatives
336
404
315
413
527
429
379
374
607
659
242
441
290
431
269
667
481
554
365
278
461
a
577
717
601
386
560
392
321
718
531
539
398
541
591
493
446
68
76
.48
74
85
56
41
48
41
200
a
31
4
113
28
300
143
a
149
107
a
23
131
71
a
42
a
112
a
a
192
a
85
a
67
419
a
* Sales-weighted average VOC content not available.
Source: Industry Insights. Architectural and Industrial Maintenance Surface
Coatings VOC Emissions Inventory Survey. Prepared for the National Paint
and Coatings Association in Cooperation with the AIM Regulatory Negotiation
Industry Caucus. Final Draft Report. 1993.
1-7
-------
survey data collected, the sales-weighted average VOC contents
for surveyed coating products are listed in Table 1-1.c<5
1.3.2 Demand for Architectural Coatings
1.3.2.1 Conceptual View of Coating Decision. The demand
for architectural coatings derives from the demand for the
treatment of architectural surfaces. Surface treatment
services include not only coating treatment, but also
noncoating treatment alternatives such as wallpaper or
exterior siding. While the choice among coating alternatives
is emphasized below, it is implicitly recognized that the
substitution between coating and noncoating surface treatments
is possible as well.
The coatings themselves are an input into the production
of surface treatment services, the final product of interest.
Each surface possesses certain attributes that affect the
demand for surface treatment. These include surface material
(substrate), age, exposure (e.g., weather, chemicals), and
other physical factors that intrinsically affect the relative
performance of treatment alternatives.
In an economic decisionmaking context, we think of the
owner of the surface as seeking to maximize the utility
derived from the services provided by the surface (i.e.,
shelter, decoration, etc.). Let process i indicate the
activity of treating a surface defined by the attributes
above. Through this process, labor, capital, and materials
are employed to treat the surface. Thus we can characterize
'Sales-weighted average VOC content is
(VOC Content^ • (Sales)i
t
(Sales)i
i-l
where VOC content is equal to the percentage by weight, sales are measured
in pounds per year, and n equals the number of product categories.
1-8
-------
the production of a unit* of surface treatment through process
i as follows:
Qi = Q(L,K,XJ
where Qi is the surface area unit (e.g., 1,000 ft2) treated
using process i and L, K, and Xt are the quantities of labor,
capital, and material (e.g., coatings) used to produce Q^.
For the processes that include coatings application,
assume there is a fixed proportions relationship between each
input and output, determined by the type of coating being
used. For example, process A requires 1 gallon of coating A,
40 hours of labor, and 10 units of capital to cover a unit
area of a given surface type. Therefore, for a given set of
input prices, there is a (constant) per-unit cost of
treatment. Costs of noncoating alternatives can be similarly
computed. Considering all n possible treatment alternative
for a given surface generates an array of costs (C^Cj, . . . ,Cn) .
Each owner/consumer places a subjective value on the
outcome of each treatment alternative. This value derives
from such factors as innate preferences for the visual appeal
of treatment alternatives and perceptions of the structural
quality and durability. For example, consumer A may prefer
the look of glossy solvent-based coatings to flat water-based
coatings and/or may perceive other differences in product
quality. The consumer explicitly or implicitly monetizes
these preferences, and the associated monetary values for each
of the n alternatives comprise the array of perceived benefits
for (B1,B2, . . . ,Bn) .
In evaluating the choice among treatment alternatives,
the consumer weighs each alternative's monetized benefit, Bi,
against the cost of treatment, Ci- The subjective payoff from
each alternative can be expressed as
1-9
-------
nt =
The consumer maximizes utility with respect to the
surface treatment choice by selecting the alternative with the
highest payoff. This of course presumes that at least one of
the payoffs is not negative. If all potential payoffs are
negative, the consumer is better off by choosing no surface
treatment at all.
1.3.2.2 Substitution Effects .
The purpose of this discussion is to describe how
consumption choices may change in response to any price
effects of the regulations. If the regulations induce a
change in the price schedule of various architectural
coatings, the unit costs of treatment alternatives will be
directly affected. Furthermore, the regulations may induce a
change in the structural characteristics of the coating that
alters the application technology. For example, a different
VOC content may change the volume of the coating that must be
applied and the amount of labor and capital necessary to
achieve the same surface area treatment; consequently, the
technological parameters may change with the new VOC
requirements. Therefore, treatment costs will be affected
jointly by what we call the factor price effect and the
technology effect. If, for example, the VOC-content
regulations would raise the price of the affected coatings and
reduce the technological efficiency of the treatment process
(e.g., more coats necessary), then both the factor price
effect and technology effect would combine to increase the
cost of the affected treatment alternatives, generating a new
set of treatment costs {CV / • • . »Cn' ) . Alternatively the new
formulas could improve technical efficiency, but at a higher
cost and the net effect on price would be unknown.
1-10
-------
VOC-content regulations may also affect consumer
valuation of the treatment alternatives through a change in
visual characteristics and altered perceptions of quality or
durability. These changes generate a new set of subjective
values for the treatment alternatives of (B^ , . . . ,Bn') . As a
result, evaluating the new arrays of benefits and costs
produces a new array of treatment payoffs, (n^1...,nn' ) . The
consumer can again be expected to select the treatment
alternative with the highest payoff. This situation may
produce a different optimal selection than the no-regulation
case. The consumer may in fact choose a noncoating
alternative or no-treatment alternative, where coating
treatment would be selected without the regulation.
1.3.2.3 Aggregate Demand.
If all consumers' preferences were identical and all
surfaces to be treated possessed the same characteristics, the
consumer choice model above would predict only one optimal
type of surface treatment throughout the economy. A wide
array of treatments and coatings are actually applied,
however, indicating a variety of surfaces with different
characteristics as well as individual preferences that vary
across consumers.
Aggregating over all consumers and all surfaces, we can
see how the regulatory changes can induce substitution among
treatment alternatives and changes in aggregate demand for the
affected coatings. These aggregate changes in demand and the
associated effect on consumer welfare are the focus of this
study.
1.3.2.4 Coating Users. Users of coatings can be divided
into two groups: professionals and nonprofessionals. The
nonprofessional is typically a "do-it-yourselfer* who
purchases only a small amount of coatings each year. The
application of coatings by nonprofessionals is limited
1-11
-------
primarily to residential architectural coatings. Professional
users of coatings may be professional painters or contractor/
builders. These professionals apply coatings to a broad array
of surfaces in residential, commercial, institutional, and
industrial settings. Table 1-2 shows that in 1991
do-it-yourselfers purchased two-thirds of all residential
architectural coatings.6 It seems reasonable to assume that
contractors purchased all of the nonresidential architectural
coatings and thus accounted for 60 percent of the use of all
architectural coatings.
TABLE 1-2. CONSUMERS OF ARCHITECTURAL COATINGS
Percentage of Total
Gallons in 1991 (%)
Residential
Do-it-yourselfers 41
Contractors 19
Nonresidential" 40
Total 100
• Commercial, institutional, light industrial.
Source: National Paint and Coatings Association. U.S. Paint Industry Data
Base. Menlo Park, CA, SRI International. 1992.
1.3.3 Production of Architectural Coatings
1.3.3.1 Raw Material Inputs. Coatings comprise four
basic types of materials: pigment, resin (binder), solvent,
and additives. Pigment is the solid component consisting of
uniform particles of a controlled size that are insoluble in
the vehicle (the liquid portion of the coating). Pigments are
used in coatings to decorate and protect and as fillers.7
Pigmentation, although it varies depending on desired
1-12
-------
properties, is similar in both waterborne and solventborne
formulations.
Film-forming binders surround and hold together the
elements of the coating film and make up the nonvolatile
portion of the vehicle. Resins aid in adhesion; determine the
cohesiveness of the dried film; affect gloss; and provide
resistance to chemicals, water, and acids. Natural and .
synthetic resins and oils, along with certain additives such
as driers and plasticizers, serve as binders in coatings and
are one of three types: multiuse resins (acrylics, vinyls,
urethanes, polyesters); thermoset resins (alkyds, epoxies);
and oils (drying oils, bodied oils).
The vehicles in organic solventborne and waterborne
paints differ not only by the type of resin used, but also in
the way they form a film and dry (or cure). Alkyd paints are
oxidizing film formers in which the drying oils react with the
oxygen in the air when the paint dries. The chemical reaction
binds the molecules of the vehicle into a hard, dry film.
Alkyd coatings continue to oxidize long after they dry and
eventually provide a rock hard surface. Latexes consist of
tiny, heat-sensitive plastic particles (latex) that are
dispersed but not dissolved in water along with the pigment.
As the water evaporates, a layer of closely packed plastic
particles and pigment is left behind. The softened plastic
particles then lose their shape and molecules diffuse and
reattach to form a binding film.8 The chemical
characteristics of latex and alkyd paint influence some of
their characteristics, such as gloss and resistance to
blocking and water. Heat-sensitive plasticizers in latex
paint cause the residual tackiness called blocking, which is
more of a problem in glossy latex paints where the ratio of
resin to pigment is higher. Precise control of particle shape
and size in the film former is necessary to increase gloss.
1-13
-------
The plastic mesh also breathes better, allowing water and air
to pass through it. The oxidizing process of alkyds forms a
smooth (thus glossier), watertight skin of hardened resin that
provides durability and water resistance.
Petroleum distillates in alkyd paints and water in latex
paint function as the carrier, or volatile vehicle, that
disperses the pigment and resin and provides the necessary
fluidity for applying the coating. Basically there are two
types of solvents: water and organic. In alkyd paints
organic solvents dissolve the components of the film former,
keeping them in solution. In latex paints, water separates
and suspends the droplets of film former. Following
application, the evaporation rate of the particular solvent
controls the rate at which the film forms, leaving the pigment
and resin bonded to the surface. Latent solvents, which
dissolve the film former when combined with true solvents, and
diluents may be added to the true solvent.9 Diluents can be
blended with the dissolved solution to extend the true and
latent solvents. Water is the true solvent used in latex
paints but may function as a diluent in alkyd formulations.
Three types of organic solvents are used in coatings:
hydrocarbons (aliphatic, aromatic); oxygenated solvents
(alcohols, esters, ketones, glycol ethers); and chlorinated
solvents (1,1,1-trichloroethane, methyl chloroform).10
Architectural solventborne paints are mainly formulated with
aliphatic hydrocarbons.
Additives are used in relatively small amounts in both
organic solventborne and waterborne formulations to provide
additional necessary properties or augment the properties of
other inputs. They may be added to the film former, solvent,
or pigment. Waterborne paints in particular may use additives
such as agents to reduce foaming or bubbling of paint when it
1-14
-------
is shaken and applied; wetting agents, which can improve
pigment dispersion or adhesion; freeze-thaw agents, which
reduce the temperature at which the paint will freeze to
prevent coagulation; and coalescing agents, which aid the flow
of the latex particles to form a more continuous film.11 VOC
contents in latex paints (4 to 10 percent, or 50 to 200 g/L)
are due to the additives used.12 Solvents such as alcohols and
ethylene glycols are added as co-solvents to waterborne
formulations. They are often necessary to allow the plastic
particles to soften and be mobile enough to bind into a
continuous film.13
The additives used in the largest volume are thickeners,
fungicides and preservatives, plasticizers, and defearners.14
Figures 1-2 and 1-3 show the principal raw material
ingredients discussed above as they are used in organic
solventborne and waterborne coating formulations.
1.3.3.2 Formulations. One of the distinguishing
characteristics of each coating is the relative amount of the
three main material inputs contained in the coating: pigment,
binder, and solvent. Different formulations, particularly
different ratios of pigmentation in the dried film to total
volume of the dried film (pigment-volume concentration), will
lead to correspondingly different protective and decorative
functions.15 For example, a coating designed to hide surface
irregularities (like a mastic texture coating) has a higher
pigment-volume concentration than a gloss varnish whose
decorative function is to impart a shiny transparent or semi-
transparent coating. Low pigment-volume concentrations have
an increased resin content and in general have high
durability, gloss, and washability. The ratio of solvent to
nonvolatile components ("solids") also characterizes types of
1-15
-------
Alkyd
Resin
Aliphatic Titanium
Hydrocarbons Dioxide
Talc
Nonvolatile Volatile
Component Component Hiding
(Film Former) (Carrier) Pigment Extender
Additives / \ Additives
All Other Inputs
Figure 1-2.
Packaged Coating
Inputs generally used in the manufacture of a
solventborne coating.
Aoyfc
Retm
Water
Titanium
Dioxide
>lat*> I
Nonvolatile Volatile
Component Component Hidtfig
(Film Former) (Carrier) Figment
Additives
Talc
Extender
AlOher Input*
Figure 1-3.
Inputs generally used in the manufacture of a
waterborne coating.
1-16
-------
coatings.d Penetrating stains have a low solids-to-solvent
ratio, and, when the solvent evaporates, virtually no film is
left behind.
Figure 1-4 shows typical formulations and average VOC
contents for a few architectural coatings.16 The coatings in
Figure 1-4 with higher solvent content also have higher VOC
content. A low solids-to-solvent ratio, as with
semi-transparent stain, is associated with high VOC content in
coatings with organic solvents because VOCs are contained
almost exclusively in the solvent portion of the coating. Two
ways to reduce the amount of VOCs released from coatings are
to increase the solids-to-solvent ratio and to substitute
water for an organic solvent.
1.3.3.3 Manufacturers' Substitution Options and New
Technologies. Manufacturers face two substitution
possibilities to reduce VOC emissions from coatings. They may
reformulate the coating to increase the solids-to-solvent
ratio. Alternatively, manufacturers may reformulate the
coating so that it contains the same amount of solvent but
emits fewer VOCs during application (i.e., substitute water
for an organic solvent). Certain coatings such as interior
flat wall paint, interior semigloss, and exterior house and
trim paint have been formulated using water for several years.
Between 1950 and 1980, waterborne coatings replaced
approximately 70 percent of solventborne coatings.17 The
performance of latex paints often meets and even exceeds alkyd
counterparts; therefore, manufacturers may choose to
discontinue organic solventborne paints in these product
classes.
^Nonvolatile components are often referred to as the "solids* portion
of the coating, which includes pigments, resins, and other additives,
although resins are not really solid until the film forms and are
considered part of the nonvolatile vehicle, or liquid portion of the
formulation.
1-17
-------
Semitransparent
Stain
Exterior Nonflat
Paint
General Primer or
Undercoater
Solvent
Pigment
Binder
Solvent
Pigment
Binder
1
Solvent
Pigment
Binder
1
Average VOC Content
Organic solvent 527
Waterborne 85
Figure 1-4.
Average VOC Content Average VOC Content
Organic solvent
Waterborne
76
Organic solvent
Waterborne
48
Approximate volume relationships of
coating ingredients.
Note: VOC content in grains per liter from Table 1-1.
Source: Whittington, Trevellyan V. Paint Fundamentals. In Paint
Handbook. Guy E. Weismantel (ed.). New York, McGraw-Hill.
Pp. 1-1 to 1-23. 1981. (Adapted from Figure 1.4)
Other products, including stains, clears, high-gloss
enamels, outdoor varnishes, and some special purpose coatings,
are more difficult to reformulate. According to a 1990
article, clear coatings have two problems associated with
them: waterbornes are transparent to UV radiation, whereas
organic solventbornes absorb UV rays thus protecting the
substrate; and waterborne acrylic polymers are not strong
enough.18 Quality performance in reformulated products is
currently possible, but the cost may be very high.19 As new
technologies become more refined, new resin systems, such as
1-18
-------
alkyd systems once used only in solventbornes, will be used in
more coatings, so prices will become more competitive.
High solids content formulation is an alternative
technology to waterborne formulations that manufacturers have
employed to reduce VOC emissions from coatings. A high solids
coating is formulated with a high solids-to-solvent ratio."
Since a smaller percentage of solvent is contained in the
coating, fewer VOCs are released during application.
Table 1-3 shows example reduced solvent contents of three
different types of reformulated organic solventborne
products.20
TABLE 1-3. PERCENTAGE OF SOLVENT IN CONVENTIONAL AND
REFORMULATED ORGANIC SOLVENTBORNE COATINGS
Conventional Solvent Reformulated Solvent
Product Content (%) Content (%)
Interior semigloss
Clear coatings
Stains
60
55 - 62
72 - 85
47
35 - 37
30 - 35
Source: Bakke, Timothy 0. Clean Air Paints. Popular Science. 237:85.
August 1990.
Disadvantages noted in the past of higher solids organic
solventborne paints include increased viscosity, longer drying
time, reduced durability, and generally higher prices.21
Reformulated organic solventbornes may be thicker, which would
make them harder to apply and extend drying time, but they may
offer greater protection. Durability may be compromised
because of the reduced strength of shorter chain alkyd
molecules substituted for longer chain molecules to improve
flow.22 Reformulated alkyd products can offer some advantages
•Note that the definition of high solids varies by coating type.
1-19
-------
however. Durability may be traded for flexibility, which
provides increased resistance to cracking and peeling.
Presealers may not be necessary for wood substrates because
the thicker coatings penetrate more evenly. Reduced VOC
emissions, lower odor, and reduced toxicity and flammability
are other benefits.
Raw material suppliers are expanding and improving upon
existing technologies to meet demand for performance in new
waterborne and high solids formulations. Solvents for use in
waterborne formulations (i.e., glycol ethers) and high solids
(keytones, esters) are replacing many of the hydrocarbon
solvents used in solventborne formulations. Resins are being
developed with a goal toward improved performance in new
low-VOC formulations; similarly additives are being developed
to improve flow and leveling characteristics of the new
resins. Additional low-VOC technologies are reactive diluent
technology, radiation curing technologies, and powder
coatings, which currently are mainly used in manufacturing
applications.
1.3.. 4 Industry Conditions
1.3.4.1 Shipments and Manufacturer Specialization. In
1991, the architectural coatings segment of the paint and
allied products industry shipped $4,881.9 million in
potentially regulated products (Table 1-4) .23.24.25,26.27,28,29,30,31,32
The value of shipments steadily increased by approximately
59 percent between 1981 and 1991, with a slight decrease
between 1990 and 1991. The strong construction market
throughout the 1980s helped contribute to this growth, but the
industry as a whole was generally considered to be maturing in
the early 1990s.33 In 1991, the size of the architectural
coatings component relative to the total coatings industry was
37.8 percent. New products are important to the paint and
allied products industry, because growth for individual
1-20
-------
TABLE 1-4. VALUE SHIPPED OF POTENTIALLY REGULATED PAINT AND ALLIED PRODUCTS:
1981 THROUGH 1991 ($106)
1987
Product
Code
28511
28511 93
28511 00
28513
28513 0
28513 11
Product Class
Architectural coatings
Exterior solventborne
Exterior waterborne
Interior solventborne
Interior waterborne
Architectural lacquers
Architectural
coatings, n.s.k.*
Special purpose
coatings
Industrial new
construction and
maintenance paints
Interior
Exterior
Traffic marking paints
1991
4,881.
856.
1,166.
647.
2,054.
83.
73.
293.
503.
132.
y
8
0
2
8
3
7
8
2
4
1990
4,913.
890.
1,201.
644.
2,018.
85.
72.
265.
484.
138.
b
3
9
1
9
5
9
9
7
0
1989
4,525
819
1,062
655
1,850
73
63
240
497
138
.J
.7
.3
.1
.8
.5
.9
.0
.4
.0
1988
4,426
817
1,014
621
1,779
84
109
208
395
136
.8
.7
.3
.4
.2
.7
.6
.6
.2
,-i
1987
4,245
816
952
603.
1,620.
81
170.
179,
343
104
.4
.9
.7
.0
.1
.7
.9
.9
.1
6
Year
1986
4,010.0
765.3
899.8
553.5
1,491.9
81.2
218.2
128.7
321.8
78 5
1985
3,830
757
851
544
1,405
79
192
149
330
95
.8
.8
.1
.7
.5
.2
.4
.7
.2
9
1984
3,559,
731,
875,
512,
1,343,
64.
32,
126
252
.3
.0
.4
.6
.0
.9
.9
.9
.9
N/A
1983
3,320
672
833
490
1,210
57
35
103
191
.8
.4
.8
.2
.7
.6
.9
.5
.9
N/A
1982
3,092.2
621.9
789.1
473.1
1,119.7
45.3
42.9
120.0
209.7
N/A
1981
3,065.6
609.6
719.7
445.5
1,090.6
37.1
163.2
88.4
203.9
N/A
to N/A . Not available.
M • n.s.k. » Not specified by kind.
Sources: U.S. Department of Commerce.
Printing Office. 1983
U.S. Department of Commerce.
Printing Office. 1984.
U.S. Department of Commerce.
Printing Office. 1985.
U.S. Department of Commerce.
Printing Office. 1986.
U.S. Department of Commerce.
Printing Office. 1987.
U.S. Department of Commerce. Current Industrial Reports:
Printing Office. 1988.
U.S. Department of Commerce.
Printing Office. 1989.
U.S. Department of Commerce. Current Industrial Reports:
Printing Office. 1990.
U.S. Department of Commerce.
Printing Office. 1991.
U.S. Department of Commerce.
Printing Office. 1992.
Note: Inflation factors used by the Census to produce estimates for
disaggregated figures here. For architectural coatings: 1987-
purpose coatings: 1987-1991, 1.06; 1982-1986, 1.00; and 1981,
relationships and for 1982-1986 on 1982 Census relationships.
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Current Industrial Reports:
Paints and
Paints and
Paints and
Paints and
Paints and
Paints and
Paints and
Paints and
Paints and
Paints and
the entire
1991, 1.00
1.08. The
Allied Products, 1982.
Allied Products, 1983.
Allied Products, 1984.
Allied Products, 1985.
Allied Products, 1986.
Allied Products, 1987.
Washington, DC, Government
Washington, DC, Government
Washington, DC, Government
Washington, DC, Government
Washington, DC, Government
Washington, DC, Government
Allied Products, 1988. Washington, DC, Government
Allied Products, 1989. Washington, DC, Government
Allied Products, 1990. Washington, DC, Government
Allied Products, 1991. Washington, DC, Government
industry for aggregated data were applied to the
; 1982-1986, 1.004; and 1981) 1.04. For special
inflation factors for 1981 are based on 1977 Census
-------
producers is predicted to come from market share expansion,
new product introductions, and improvements in established
products.34
Sales in the architectural sector generally reflect
activity in house redecoration, maintenance and repair, as
well as sales of existing homes, new home building, and, to a
lesser extent, commercial and industrial construction. Among
interior and exterior architectural coatings, the waterborne
coatings market dominates the sector and experienced a larger
percentage increase in growth than did organic solventborne
coatings. Interior waterbornes grew the most, 88.4 percent
from 1981 through 1991. In 1991, 76 percent of interior
coatings and 57.6 percent of exterior coatings were
waterborne. Partly in response to environmental regulations
aimed at the reduction of VOC emissions, the industry has
shifted from manufacturing conventional organic solventborne
paints in favor of paints with high solids-to-solvent ratios
and waterborne and solventless paints.35 However, much of this
trend has also been driven by consumer demand.
Although the historical Census data do not identify value
of shipments for paint products within the four product
classes, other sources indicated that the majority of interior
wall and exterior siding paint jobs use waterborne
products.36'37'38 Therefore, the exterior and interior solvent-
borne shares probably account for mainly coatings used on
exterior and interior trim, floors, decks, and high-gloss
enamels.
Industrial new construction and maintenance paints and
traffic marking paints are classified by the Census as special
purpose coatings, which comprised 22 percent of the total
coatings market in 1991. Market shares for industrial
maintenance and traffic marking paints within the special
purpose segment were 28 percent and 4.6 percent, respectively.
1-22
-------
Growth prospects for this segment are expected to be above
average, especially for industrial and machinery maintenance
coatings.
For all companies classified in the paints and allied
products industry in 1987, 98 percent of their value of
shipments was generated from the manufacture of paints and
allied products (Table 1-5).fi39 Only 3 percent of the value of
paints and allied products shipped were manufactured by
companies outside the industry. The top three secondary
producers of paint and allied products account for about half
the value produced as secondary products in other industries
and are shown in Table 1-6: adhesives and sealants, plastics
materials and resins, and printing ink.40 Because coating
products often function as sealants, the adhesives and
sealants industry is a logical secondary producing industry.
1.3.4.2 Company Size and Industry Structure.
Information on industry structure is highly dependent on one's
definition of the industry in question. The data used in this
discussion apply to the entire Paint and Allied Products
Industry (SIC 2851). As indicated above, architectural
coatings account for just under 40 percent of industry
shipments. Unfortunately, the industry structure data are not
available for the architectural coatings component of the
industry. Therefore, the information presented here may not
always accurately reflect the structure of the architectural
coatings sector.
In 1987, the paint and allied products industry comprised
1,121 companies owning a total of 1,428 establishments
(Table 1-7) ,41-42 Single establishments were held by
approximately 77 percent of the companies, and they had an
£Industry statistics, unless otherwise noted, include figures for all
segments of the paint and allied products industry, not just those to be
regulated.
1-23
-------
TABLE 1-5. NUMBER OF COMPANIES, ESTABLISHMENTS, AND PRODUCER
SPECIALIZATION-PAINT AND ALLIED PRODUCTS: 1987
Industry
Primary
Industry/ Number of Product Coverage
SIC Primary Number of Establish- Specialization Ratio
Code Product Class Companies ments Ratio (%)* (%)b
Product
Class
Total Made
in All
Industries
2851 Paints and
allied
products
28511 Architectural
coatings
28513 Special
purpose
coatings
1.123
1,426
282
131
98
97
12,078.8
• Value of primary products for the industry divided by the sum of the value of
primary products produced by the industry and the value of secondary products
produced by the industry.
b Value of primary products for the industry divided by the total value of
products for that industry produced in any industry.
Source: U.S. Department of Commerce. 1987 Census of Manufactures, Industry Series:
Paint and Allied Products. Washington, DC, Government Printing Office.
1990.
average value added of $1.1 million. The multiestablishment
companies had an average value added of $20.4 million and
produced almost 85 percent of the total value added for the
industry. Also shown in Table 1-7, the 50 largest companies
in 1987 produced 66 percent of the total value of shipments
for the industry. Data from the Small Business Administration
(SBA) indicate that in 1991 there were 1,152 companies and
approximately 98 percent of those were classified as small
businesses as defined by having fewer than 500 employees.43
Figure 1-5 displays the location of manufacturing
establishments in the paint and allied products industry by
state.44 California has the greatest number, 201, followed by
Illinois with 118. Paint manufacturing is fairly well
represented in most states east of the Mississippi River.
1-24
-------
TABLE 1-6. COMMODITY PRODUCTION IN 1982: PAINT AND ALLIED
PRODUCTS
(SIC 2851) Value* Percentage
Product Examples Producing Industries ($106) Produced
Interior and exterior Primary 8,243.3 96.5
paint, lacquers, and
varnishes; OEM All secondary
coatings; industrial producers 303.2 3.5
new construction and
maintenance paints, All producers 8,546.5 100.0
traffic paints,
automotive refinish
paints, marine paints, Top three secondary: 142.6 1.7
aerosol coatings, Adhesives and
paint and varnish sealants 68.8 0.8
removers, thinners, Plastics materials
putty and glazing and resins 45.8 0.5
compounds, brush Printing ink 28.0 0.3
cleaners
All other secondary 160.6 1.9
producers
• Measured at producers' prices.
Source: U.S. Department of Commerce. The 1982 Benchmark Input-Output
Accounts of the United States. Washington, DC, Government Printing
Office. 1991.
In the 1980s, consolidation was a major trend in the
paint and allied products industry. The maturity of the
industry and increased technology requirements are factors
contributing to the restructuring. A large number of mergers
and acquisitions took place in response to pressure from the
higher cost of paint ingredients, intense industry
competition, compliance with government regulations, and low
profit margins.45 Other companies divested their paint and
coating operations to focus on other businesses or as an
alternative to making the capital and research and development
(R&D) commitments required to remain competitive. The number
of coating manufacturers and the number of establishments
1-25
-------
TABLE 1-7.
LARGE FIRM DOMINANCE AND NUMBERS OF COMPANIES AND ESTABLISHMENTS
IN THE PAINT AND ALLIED PRODUCTS INDUSTRY: 1987
Industry
(SIC 2851)
Paint and allied
product producers
Multiestablishment
companies
Single-establishment
companies
Number of
Companies
1,121
259
862
j_, ' Percentages consist of the
• companies) , divided by the
to
Number of
Establish-
ments
1,428
566
862
Value
Added
6,220.7
5,271.3
949.3
Average
Value
Added"
(S106)
5.5
20.4
1.1
sum of value of shipments of
total value of shipments of
Value of Shipments
Percentage Accounted for
Total 4 Largest 8 Largest 20 Largest
($10') Companies Companies Companies
12,702.4 27 40 53
10,651.9
2,050.5
the four largest companies (or 8, 20,
the industry.
by •
50 Largest
Companies
66
or 50
b The total value added divided by the number of companies.
Sources: U.S. Department of Commerce. Census of Manufactures, Subject Series: Type of Organization.
Washington, DC, Government Printing Office, February 1991.
Source for Percentage Accounted for Data: U.S. Department of Commerce. 1987 Census of
Manufactures, Concentration Ratios of Manufacturers. Washington, DC, Government Printing
Office. 1992.
-------
I
to
Figure 1-5. Location of manufacturing establishments in the paints and
allied products industry in 1987: SIC 2851.
Source: U.S. Department of Commerce. 1987 Economic Censuses. Volume 1, Report Series, Release ID.
Census of Manufactures: Location of Manufacturing Plants. file MC87LMCO. 1991.
-------
operated by these manufacturers has decreased. As indicated in
Table 1-8, from 1972 to 1991, the number of companies
decreased by 12 percent, and the number of manufacturing
establishments decreased by over 20 percent.46
TABLE 1-8. NUMBER OF COMPANIES AND ESTABLISHMENTS IN THE
COATINGS INDUSTRY, SELECTED YEARS, 1972-1991
Year
1972
1977
1982
1987
1991
% change 1972-1991
Number of
Establishments
1,599
1,579
1,441
1,426
1,400.
-12.4%
Number of Companies
1,317
1,288
1,170
1,123
1,030'
-21.8%
• 1991 figures are from Finishers' Management. The U.S. Paint and
Coatings Industry, pp. 23-25. April 1991.
Source: U.S. Department of Commerce. 1987 Census of Manufactures,
Industry Series: Paints and Allied Products. Washington, DC,
Government Printing Office. 1990.
On average, 35 to 40 mergers or acquisitions took place
each year in the coatings industry in the late 1980s and early
1990s.47 A transaction involves the transfer of production
capacity from one company to another but does not necessarily
indicate the dissolution of the company making the transfer.
The selling company could sell only a division or product line
and remain in business. Some of the larger acquisitions
reported in trade journals, by the press, and in companies'
annual reports are listed in Table 1-9.*8
1-28
-------
TABLE 1-9. ACQUISITIONS IN THE COATINGS INDUSTRY: CIRCA 1990
Selling Company
Acquiring Company
Division Sold
DeSoto
Whittaker Corp.
Azko Coatings
Inc.
DeSoto
Clorox Co.
Sherwin Williams
Morton International
Reliance Universal
Inc.
Valspar
PPG Industries, Inc.
Consumer Paint Operation
Specialty Chemicals
Operation
Buyout
Coil Coatings Operation
Olympic and Lucite
finishes
Source: Loesel, Andrew. Coatings Industry Faces New Mix. In Chemical
Marketing Reporter. 238(18):SR3-SR8. New York, Schnell Publishing
Co. 1990.
Most of the larger companies produce architectural,
original equipment manufacturer (OEM), and special purpose
coatings. Several of the largest coatings producers are
chemical corporations; however, paint manufacturing represents
only a small part of their overall business.49 In 1991, merger
activity slowed down and left the industry basically divided
into two groups: a few, well-financed and highly diversified
multinationals and a large number of regional paint
companies.
50
1-29
-------
SECTION 2
COSTS OF REGULATION FOR ARCHITECTURAL COATING PRODUCERS
This section estimates the costs to comply with the
architectural coatings regulation and examines the economic
impacts of these costs as they are distributed across
producers and consumers of the regulated products through
market processes. The analysis in this section focuses on
the(primary)impacts defined within the architectural coatings
product markets. An assessment of impacts on users of traffic
coatings addresses selected secondary impacts in other sectors
of the economy. That analysis is presented in Section 4.
2.1 BACKGROUND
The EPA plans to control VOC emissions from architectural
coatings using a combined regulatory approach: (1) product-
specific VOC content limits, (2) an option for producers of
products that exceed the content limits to pay a fee on the
VOC content in excess of the limit, and (3) a phased tonnage
exemption that allows each manufacturer the option to claim as
exempt a limited number of products that result in a specified
amount of emissions annually. Using reformulation cost
estimates and an exceedance fee rate, the Agency analyzed the
potential impacts of the regulation, first using static
analyses of regulatory response options and second using a
2-1
-------
dynamic market analysis that estimates changes in prices,
quantities, and social welfare.
2.2 OVERVIEW OF RESPONSE OPTIONS
The regulation to reduce the VOC content of architectural
coatings will affect both production decisions for the
suppliers of the coatings (through its impact on costs and
revenues) and consumption decisions for the demanders (through
its impact on product prices). Before developing a formal
economic model to analyze these regulations, the Agency needed
to characterize the scope of responses available to producers
and consumers.
2.2.1 Supply
The EPA is proposing a set of limits for the VOC content
in specific product categories to be met in 1999. Firms that
produce products exceeding the VOC limits essentially have
three compliance options:
• reformulate the products so that they comply with the
standard,
• pay a fee on the excess VOC content over the standard,
or
• remove the product from the market.
Each producer also may exempt a small quantity of product from
compliance.
This analysis assumes that firms will choose the option
that maximizes their net benefits, as measured by the expected
(discounted) value of the profits generated under each option.
Although decisions in the short-run may differ from decisions
made to maximize net benefits in the long run, this analysis
primarily considers the long-run decisions and their impact on
2-2
-------
the architectural coatings markets. Uncertainties pertaining
to short-run decisions are discussed in Section 2.7.
The first option for producers to comply with the rule is
to reformulate products that exceed the specified VOC content.
Product reformulation often involves an investment in research
and development (R&D) to develop a compliant product. The
extent of the reformulation necessary to bring a product into
compliance can vary from product to product. In some cases,
compliance can be achieved for a particular product without
large R&D investments because the product is similar enough to
an existing formula or another product undergoing
reformulation. A major reformulation, as is discussed
throughout this analysis, typically requires a significant
resource and time commitment. The process can take several
years and is divided into a number of different stages.
Figure 2-1 identifies the basic reformulation stages for a
prototype architectural paint (other coatings such as
varnishes may have fewer stages).51 The firm may subsequently
need to alter its capital equipment to produce the
reformulated product, but these physical capital adjustments
are usually small compared to developing the intellectual
capital to devise the new formula.
The analysis that follows assumes that manufacturers bear
the full cost of each reformulation. Since the VOC content
limits in the rule reflect available resin technologies, it is
likely that the costs associated with reformulation will at
least partially be shared by resin manufacturers/suppliers.
In that regard, the direct impacts on manufacturers will be
overstated in the analysis. This and other potential upward
and downward biases in the cost estimation methodology are
addressed later in this section.
2-3
-------
Formulate a Whoa Paint
I
Conduct FteW T*tHno *nd Cotof Formutotfon
1
I
Formulate the Bases 1
(that wM be tinted) |
I
Determine Prescription for Raul
Stores
1
Develop Information
(technical and safety sheets, marketing
Information, labels)
Scale Up Production, Introduce, and Distribute
Estimated Effort 2 to 3 person-years
Estimated Elapsed Time: 1.5 to5.5 year*
Figure 2-1. Basic stages of architectural coating reformulation
(prototype firm and product).
Source: AIM Coatings Regulatory Negotiation Committee meeting.
July 28-30, 1993, Washington, DC. Meeting Summary.
2.2.2 Demand
The regulation can be expected to induce changes in the
prices of the affected products. Product consumers may alter
their selection of coatings based on the relative prices of
coating products and on the relative prices of coating versus
noncoating alternatives. For example, consumers might opt for
a waterborne coating rather than its solventborne alternative
if the regulation-induced change in prices increases the
relative price of the solventborne product. Moreover, a
potential user of a high-VOC coating product facing
reformulation may even opt for a noncoating alternative if the
price rises too much.
2-4
-------
The reformulated products can also possess different
characteristics that affect their demand. For instance, VOC
content reduction in a typically high-VOC product may change
consumers' perceptions of the product's performance,
durability, and ease of application. The lower VOC content
may also work as a signaling device for the "green* consumer
in pursuit of products deemed more friendly to the
environment.* These factors collectively affect the benefit
consumers derive from using the product and thus their
willingness to pay for the reformulated product versus other
product alternatives.
2.3 COST ANALYSIS
This section evaluates the costs imposed on manufacturers
to reformulate noncompliant products, describes and quantifies
the exceedance fee provision, and incorporates the option of
withdrawing products from the market into the decision
process.
2.3.1 Costs of Reformulation
Of the compliance options referenced above, reformulation
of products that have a VOC content exceeding the category
limit in the TOS (see Table 2-1) is the most significant both
in terms of potential cost and emission reductions. The
economic analysis begins by estimating the national cost of
the regulation in the absence of other compliance options
(fee, withdrawal) and ignoring market responses. This will
provide an upper-bound estimate for the true national costs of
the regulation. The national estimate will be modified
(reduced) as the other compliance options and market behavior
are explicitly considered below.
•Some manufacturers currently produce zero-VOC-content coatings that
are marketed as "clean air* coatings.
2-5
-------
TABLE 2-1. TABLE OF STANDARDS'
VOC Content Limit
Architectural Coating (g/L)
Antenna coatings 500
Antifouling coatings 450
Antigraffiti coatings 600
Bituminous coatings and mastics 500
Bond breakers 600
Chalkboard resurfacers 450
Concrete curing compounds 350
Concrete protective coatings 400
Dry fog coatings 400
Extreme high-durability coatings 800
Fire-retardant/resistive coatings
Clear 850
Opaque 450
Flat coatings, N.O.S.
Exterior 250
Interior 250
Floor coatings 400
Flow coatings 650
Form release compounds 450
Graphic arts coatings (sign paints) 500
Heat reactive coatings 420
High-temperature coatings 650
Impacted immersion coatings 780
Industrial maintenance coatings 450
Lacquers (including lacquer sanding sealers) 680
Magnesite cement coatings 600
Mastic texture coatings 300
Metallic pigmented coatings 500
Multicolor coatings 580
Nonferrous ornamental metal lacquers 870
(continued)
2-6
-------
TABLE 2-1. TABLE OF STANDARDS' (CONTINUED)
VOC Content Limit
Architectural Coating (g/L)
Nonflat coatings, N.O.S.
Exterior 380
Interior 380
Nuclear power plant coatings 450
Pretreatment wash primers 780
Primers and undercoaters, N.O.S. 350
Quick dry coatings
Enamels 450
Primers, sealers, and undercoaters 450
Repair and maintenance thermoplastic coatings
650
Roof coatings 250
Rust preventive coatings 400
Sanding sealers 550
Sealers 400
Shellacs
Clear 650
Opaque 550
Stains
Opaque 350
Clear and semitransparent 550
Waterborne low solids 120
Swimming pool coatings 600
Thermoplastic rubber coatings and mastics 550
Traffic marking paints 150
Varnishes 450
Waterproofing sealers and treatments
Clear 600
Opaque 400
Wood preservatives
Below ground 550
Clear and semitransparent 550
Opaque 350
N.O.S. = Not otherwise specified.
* The final Table of Standards included in the regulation differs
slightly from this list. See Section 7 for a discussion.
2-7
-------
The method for estimating the national costs of the
regulation under this scenario is to:
1. Estimate reformulation cost per product
2. Estimate the total number of products nationwide
facing reformulation
3. Multiply the cost per product times the number of
reformulations
These steps are now presented in sequence.
2.3.1.1 Product-Level Reformulation Cost Estimates.
Developing a new formula for an architectural coating
involves altering the mix of the four coating components:
resins, solvents, pigments, and additives. For solventborne
products, a new formula might increase the ratio of solids
(resins) to solvents to reduce the solvent's contribution to
VOC emissions.
Reformulation is a one-time investment to develop a
formula that complies with the VOC requirement. This
generally involves applying R&D effort to develop and test the
new formula. Various other expenses (e.g., administrative and
marketing) are incurred to get the reformulated product to
market; however, for the purposes of this report, all relevant
costs are collectively referred to as "reformulation" costs.
The level of effort for reformulation varies across
products, depending on the product's characteristics and the
difference between a product's VOC content and the standard.
For the analysis at proposal, EPA used information provided at
a regulatory negotiation meeting on July 28, 1993 on the cost
of developing a new product formula to meet a standard that
was more stringent than that which was proposed.52 Because
other data were not available to gauge the reasonableness of
this estimate, the EPA solicited input during the public
comment period for this rule to determine the appropriateness
2-8
-------
of the value used at proposal. Appendix B provides a summary
of the information received. These data show that the value
used at proposal was considerably above estimates provided by
commenters. Thus, the value used for this analysis is revised
to reflect both the initial estimate from the regulatory
negotiation and the subsequent estimates provided during the
public comment period. Not enough information was provided in
these comments, however, to estimate separate costs for each
specific product category; therefore, the average of the
estimates provided is used as the cost of reformulation for
all products subject to the regulation. That average cost is
$87,000 per product and will be used throughout this analysis
to estimate the economic impacts, unless otherwise indicated.15
Cost annualization. Several of the comments received
during the public comment period indicate a concern that the
cost estimate used at proposal was too low. However, the
lump-sum cost estimate used at proposal ($250,000) was
considerably higher than the estimates provided in the public
comments. Therefore, the concern appears to be centered
around the annualized cost estimate used at proposal ($17,772
per year). In many cases, commenters appeared to be comparing
the annualized cost used in the proposal to their estimate of
lump-sum costs to reformulate. The purpose of annualizing
costs and the methods for doing so in this analysis are
presented below.
Reformulation is a one-time effort to develop a new
formula. But the useful life of the formula goes beyond the
year in which reformulation occurs. In this regard, it is
"Please note that because the base year for all information to
develop the regulation (i.e., product inventories, VOC content limits,
estimated emission reductions, etc.) is 1991, all costs and economic
impacts presented in the analysis are expressed in 1991 dollars unless
otherwise indicated. All cost and economic impact measures are
transformed to present dollars in Section 7 for external reporting
purposes.
2-9
-------
much like any other capital investment (in this case,
"knowledge* capital), so the cost must be amortized over the
useful life of the investment.
The standard formula for annualizing a lump sum
investment cost is
a = I • [i(l+i)n / ((l+i)n - 1)]
where a equals the annualized amount, I is the initial lump
sum investment cost, i is the interest (discount) rate, and n
is the useful life of the investment. As indicated above, the
revised value for the lump-sum investment used throughout this
analysis is $87,000 per product. The discount rate is
7 percent, which is the rate recommended by the Office of
Management and Budget (OMB) for cost-benefit analysis of
federal regulations.53 Determining the number of years to use
in the annualization formula, n, requires considering the
^useful life" of the knowledge developed in reformulation.
More specifically, how long do the benefits of the current
investment accrue? Reformulation allows the firm to continue
to sell the current product (at a lower VOC content), rather
than remove the product from the market. Therefore, the time
stream of the benefits to the firm is at least as long as the
reformulated product will remain on the market (i.e., the
product life). This is a complicated issue. A particular
version (formula) of a product may remain on the market for
many years, then be reformulated to add different product
attributes and kept in the market as a new and improved
version of the old product. This product reformulation
rotation may recur continuously into the future. If so, what
is the best way to estimate the useful life of the VOC
reduction technology induced by the regulation?
2-10
-------
Two assumptions were considered to capture the range of
possibilities for the useful life of the reformulation
investment.
1. The low-VOC technology developed for the new formula
is applicable only to that formula and cannot be
transferred to future adjustments of the product.
2. The low-VOC technology developed for the new formula
is applicable to that formula and is transferrable
to all future versions of the product forever.
Case 1: In the first case, if the reformulated product
is expected to remain on the market for a certain number of
years (T), then the useful life of the VOC reduction
investment is T years and the initial cost should be
annualized accordingly (n=T). Moreover, if the current
product is simply replaced T years hence by a reformulated
version of the product, it is assumed that the VOC reduction
technology developed for the current product is
nontransferrable to the next product. Thus, an entirely new
investment in VOC reduction technology T years in the future
(the time of the next reformulation) is assumed necessary.
This defines the most pessimistic (i.e., shortest) estimate
for the useful life of the current VOC reduction investment.
Because shortening the useful life of an investment reduces
the amortization period, it also raises the annualized cost of
compliance, therefore providing the upper-bound estimate for
this analysis.
Estimating the cost under the first assumption requires
determining an appropriate product life for a typical
architectural product. Attempts to obtain this information
from secondary data and industry sources proved unsuccessful
since a "typical* product was too difficult to define. A life
of T=8 years was assumed to be a reasonable, if conservative,
base case estimate of a single product life cycle. Thus a^ is
2-11
-------
the annualized reformulation cost per product for case 1 (high
estimate), with an $87,000 reformulation investment, a useful
life of 8 years, and discounted at 7 percent, which is
computed as follows:
at = $87,000 [0.07(1.07)" / ( (1.07)8-1) ] = $14,573.
Case 2: In the second case, the low-VOC technology
developed for the regulation applies to all current and future
versions of the reformulated product. In other words, once
the VOC technology is developed for the new formula, it does
not need to be re-developed in the future, even if the product
is modified in the future to add new attributes. As a result,
the useful life is the length of time the firm expects to
remain in the product market. In the extreme case, the firm
has no plans to remove the product from the market and the
useful life is essentially infinite. Under this assumption,
the cost is amortized in perpetuity to make it comparable with
the benefits of the VOC technology. Thus, the cost
annualization formula yields a2, the estimate of reformulation
cost per product:
a2 = $87,000 • 0.07 = $6,090.
Because a firm may not expect to remain in the market
forever and/or the current VOC reduction technology may not
transfer perfectly to all future versions of the current
product, the assumption for case 2 can be viewed as a
lower-bound estimate of annualized costs.
However, under an alternative interpretation, the costs
may be lower still. Suppose a company, in the absence of the
VOC standards, would routinely reformulate its product every
few years. Then, the VOC regulation can be viewed not as
2-12
-------
forcing firms to reformulate the product; rather, it forces
them to reformulate their products sooner than they otherwise
would. Thus, the one-time cost to the firm is the present
value of accelerating the series of costs that would occur
(later) without the regulation. This cost will, in general,
be less than the lump-sum cost of reformulation referenced
above; therefore, the annualized measures would be lower as
well. This is demonstrated by numerical example in
Appendix C.
To summarize, data from the regulatory negotiation and
public comment periods were used to provide EPA's best
estimate of the cost of reformulation. The average
reformulation cost estimate is $87,000 per product. This is a
one-time cost that must be annualized for policy analysis.
The annualized cost estimate depends on the assumption about
the new formula's useful life. Under a useful life estimate
of 8 years, the annualized cost per product is $14,573. As
indicated, a number of assumptions can be justified on
theoretical and empirical grounds that would reduce this
estimate. For example, the useful life of the reformulation
investment may well exceed 8 years. Also, reformulations
occur as a normal business practice and the cost of
reformulation for VOC content may not be entirely incremental.
However, the $14,573 estimate is the maintained value
throughout the analysis, except where otherwise indicated,
thereby providing a conservatively high cost estimate.
South Coast Air Quality Management District (SCAQMD)
study. As a point of comparison, estimates of the cost of
architectural coatings reformulation are provided in a study
conducted for the SCAQMD to address economic impacts of VOC
content regulations in California.54 This study identified
costs associated with product reformulation and temporary and
permanent product sales losses. Reformulation costs varied
2-13
-------
depending on the extent of the reformulation necessary. Most
of the small firms surveyed indicated that they did not have
full-time R&D employees. Costs for additional research and
development due to the regulation ranged from $1,000 to $5,000
annually for firms with few products affected by the
California rule and more than $50,000 for firms with many
affected products and little or no research staff.
The SCAQMD study also identified other compliance costs
not related to R&D. Rough estimates of the cost of equipment
adjustments necessary to accommodate reformulation ranged from
$5,000 to $35,000 per firm. Costs attributed to temporarily
or permanently discontinued products ranged from zero to
$3,000 for firms with few affected products to more than
$75,000 for firms with many affected products. Per-product
estimates were not presented. Employment changes for the
surveyed firms in the SCAQMD study were expected to be
minimal, affecting only the possible addition of R&D chemists.
Because the timing, number of reformulated products, cost
components, and regulatory structure associated with each
SCAQMD cost estimate are not apparent from the report, they
cannot be combined with the estimates presented above in any
meaningful fashion to improve the estimate of regulatory
costs.
2.3.1.2 National Reformulation Costs. The analysis of
national reformulation costs begins with the recognition that
the population of regulated products can be broken into two
groups: those included in the emissions survey and those
omitted from the survey. The methods used to estimate costs
for each group are presented in turn.
Survey population. In this section, aggregate
reformulation costs are for the products reported in the
Architectural and Industrial Maintenance Surface Coatings VOC
Emissions Inventory Survey (the survey).55 The survey
2-14
-------
population represents roughly three-fourths of total industry
output. The analysis is then extended to the industry level to
calculate a national estimate.
To estimate reformulation costs for the entire survey
population, the number of architectural products that will
need reformulation to comply with the standards is determined.
This number depends on the number of architectural products
with a VOC content exceeding the standards for the respective
product categories.
The survey reports the number of products, sales volume,
and average VOC content for specific VOC content ranges (e.g.,
0 to 50 g/L, 51 to 100 g/L, 101 to 150 g/L) within specific
product groups (e.g., exterior flat waterborne, exterior flat
solventborne, interior flat waterborne). Knowing the limits
imposed by the TOS, the number, volume, and average VOC
content of products over the limit can be derived using the
survey data. These data can be used to generate estimates of
the expected cost of reformulating products subject to the
TOS, as well as the associated reduction in emissions
accomplished by the reformulations.
Nonsurvey population. By definition, characterizing the
population of nonsurveyed products introduces further
uncertainty into the analysis. To estimate the number of
nonsurveyed products facing reformulation, one must use
product information from the survey population and apply it to
the nonsurvey population subject to some assumption about the
correspondence between the two populations. The economic
analysis presented at proposal performed this task subject to
the assumption that the overall survey population was
representative of the nonsurvey population. Further scrutiny
suggested a more appropriate assumption would be that the
nonsurvey population was more accurately represented by the
small company component of the survey population. A
2-15
-------
supplemental analysis in the appendix of the proposal analysis
addressed this issue and indicated that national cost of the
regulation is higher when the assumption that all nonsurveyed
products are produced by small companies is applied. That
assumption is maintained and further refined to generate cost
estimates for the nonsurvey population in this analysis, as
described below.
For each of the 13 defined market segments in the
architectural coatings industry, data were available on total
market volume (in liters) derived from the Census of
Manufactures data for the baseline year (1991) and the total
volume of surveyed products for that category. From that data
the total volume omitted from the survey (i.e., volume
produced by the nonsurvey population) can be computed:
Nonsurveyed volume = Market volume - Surveyed volume (2.1)
If the average size of nonsurveyed products is known, the
number of nonsurveyed products can be estimated as follows:
Nonsurveyed products = Nonsurveyed Volume / Average
volume of an nonsurveyed product (2.2)
If the proportion of nonsurveyed products needing
reformulation is known, then the number of nonsurveyed product
reformulations can be computed:
Nonsurveyed product reformulations =
Nonsurveyed products • Proportion of
nonsurveyed products needing reformulation (2.3)
and the corresponding reformulation costs are then
2-16
-------
Cost of nonsurveyed product reformulations =
Nonsurveyed product reformulations •
Reformulation cost per product (2.4)
Because no specific data on nonsurveyed products were
available for this analysis, the average product volume needed
in Eq. (2.2) and the reformulated product proportions needed
in Eq. (2.3) are not known. However, the information from the
surveyed products can be used to impute values for the
nonsurveyed products. One option is to assume that
nonsurveyed products are the same average size and have the
same rate of product reformulation as surveyed products.
However, as indicated above, the survey population is not
necessarily representative of the nonsurvey population,
because the former includes mostly large companies and the
latter mostly small companies. To more appropriately capture
the differences between the nonsurvey population and the
survey population, the following assumptions are proposed:
(1) Let the average size of nonsurveyed products in each
market segment equal the average size of small
company products reported for that market segment in
the survey data.
(2) Let the nonsurveyed product reformulation rate in
each market segment equal the reformulation rate for
small company products reported for that market
segment in the survey data.
The effect of assumption (1) is to increase the number of
nonsurveyed products and thereby increase the number of
nonsurveyed product reformulations and associated costs,
relative to the alternative assumption that nonsurveyed
products are produced by both large and small companies.
Assumption (2) adjusts the estimates based on market segment-
specific reformulation rates, which is greater on average for
small companies. The combined effect of these two assumptions
2-17
-------
is to raise the cost of the regulation relative to the
alternative assumption.
National estimate. Typically during the development of
an air pollution regulation, an engineering analysis
identifies the pollution control equipment required to comply
with the rule and estimates the total installed capital cost
in a memorandum to the public docket or as a section of the
rule's Background Information Document (BID). The economic
analysis typically uses this information to amortize costs on
an annual basis and perform a market analysis. For the
architectural rule, the control cost estimates are highly
dependent on decisions made by the regulated producers in a
market setting to either reformulate, pay an exceedance fee,
or remove the over-limit product from the market. With the
market emphasis, all costs were expressed in annual terms in
the economic analysis presented at proposal. EPA received
public comments suggesting that an estimate of total initial
reformulation cost (the analog to total installed capital
cost) would also be informative. This cost is computed and
presented below, along with the standard annual cost
estimates.
The national reformulation costs can then be estimated as
follows:
National reformulation cost =
Cost of surveyed product reformulations +
Cost of nonsurveyed product reformulations (2.5)
Table 2-2 presents the results of the analysis for the
TOS.56 The first row of Table 2-2 reports reformulation costs
and emissions reduction summed across all surveyed products.
A total of 1,730 products from the survey exceed the limits
that manufacturers and importers will be subject to, which is
36 percent of the total number of products in the survey
2-18
-------
TABLE 2-2. NATIONAL COSTS: REFORMULATION-ONLY SCENARIO
to
i
M
VO
Annualized Cost of
Population
Surveyed
products
Nonsurveyed
products
Total
Number of Products
Over Limit
1,730
1,759
3,519
Estimated
Reformulations*
1,153
1,193
2,345
Initial Lump-Sum Cost -
($1991)
$100,282,029
$103,760,202
$204,042,231
Reformulation ($1991)
Low*
$7,019,742
$7,263,214
$14,282,956
High*
$16,797,240
$17,379,834
$34,177,074
Totals subject to rounding error.
NA = not applicable
• Based on the assumption that one-third of products over the limit do not need a major reformulation (see
text).
6 Based on an initial investment of $87,000 cost per product.
c Based on an annual!zed value of $6,090 cost per product.
d Based on an annual!zed value of $14,573 cost per product.
Source: Industry Insights. Architectural and Industrial Maintenance Surface Coatings VOC Emissions Inventory
Survey. Prepared for the National Paint and Coatings Association in cooperation with the AIM
Regulatory Negotiation Industry Caucus. Final Draft Report. 1993.
-------
(4,846).c A presentation to the Regulatory Negotiation
Committee indicated that roughly one in three products that
exceeds the limits would not need a reformulation, primarily
because the product lines are similar to others that will be
reformulated. Thus, the costs are assessed for the remaining
two-thirds of products over the limit to compute the
aggregate cost estimate. After reducing the number of
products, the estimated number of reformulations for the
survey population is 1,153, yielding a range for an aggregate
cost of reformulation of $7.0 to $16.8 million dollars (1991
dollars), depending on which useful life assumption is used to
annualize the lump-sum value.
Nationally, about 2,345 products are subject to
reformulation. The initial lump-sum cost to reformulate these
products (at $87,000 per product) is just over $200 million.
Depending on the annualized cost per product estimate used,
annualized costs range from about $14 to $34 million per year.
Again, these estimates overstate the expected cost of the
regulation because they do not account for producers' best
response (i.e., their lowest cost option) to the regulation.
The next section discusses the part of the analysis that
accounts for these actions.
2.3.2 Exceedance Fee Provision
Architectural coatings producers have the alternative of
paying a fee per unit of output for products that exceed the
limit. The fee will be computed as follows:
fee = (actual VOC content - VOC limit) • fee rate. (2.6)
°The actual survey total number of products is 4,920. However,
throughout Section 2 of this report 4,846 is used as the total number (and
the corresponding quantity and emissions) because product-level data were
unavailable for 74 products in the survey.
2-20
-------
VOC content is measured in grams per liter (less water and
exempt compounds) , and the fee rate is paid on the grams per
liter in excess of the limit. The fee rate is $0.0024 per
excess gram per liter with annual adjustments based on the
gross domestic product (GDP) price deflator. Total fee
payment per product simply equals the per-liter fee times
total liters of production.
In this step of the analysis, the premise is that
architectural coatings producers will choose the less costly
of the reformulation and exceedance fee options as a
compliance strategy. The choice is based largely on two
product-specific factors: quantity of output produced and the
"excess* VOC per unit.
The diagram in Figure 2-2 helps explain the effect that
output quantity has on the choice between reformulating the
product and paying an exceedance fee. The vertical axis
represents the cost per liter of compliance and the horizontal
axis measures product volume in liters annually. Since the
cost of reformulation is a fixed cost (i.e., it is independent
of the level of output), the average reformulation cost per
liter of output falls as output levels increase. This
situation is represented by the downward-sloping line in
Figure 2-2. However, the exceedance fee per unit of output is
constant with respect to the output levels. Let F be the
exceedance fee per liter of output; the flat line extending
from F on the vertical axis indicates that the fee rate is
constant. For the purposes of this discussion, we ignore the
role of fixed recordkeeping costs under the fee option. These
costs are included in the empirical analysis that follows. In
Figure 2-2, for all output levels less than QT the average
cost of reformulation is higher than the per-unit fee, and for
all output levels greater than QT, the average cost is below
2-21
-------
S/liter
Av«ng« Co»t of
Reformulation
QT
Output L*v«l
Favors Fee Favors Reformulation
Figure 2-2. Fee versus reformulation.
the fee. This relationship indicates that the fee is the less
costly alternative when output is less than QT and
reformulation is the less costly alternative when output is
greater than QT. Thus small volume producers are more likely
to choose the fee, all else equal. As Figure 2-2 illustrates,
the existence of a fee places an upper limit on the per-liter
costs of complying with the regulation: F • Q.
Figure 2-2 also illustrates the effect of different fee
rates on the "threshold point* of quantity, below which the
fee is the preferred option. If the fee were F' instead of F,
reflecting either a higher assessment rate per Mg of emissions
or a higher amount of excess VOC per unit, the threshold point
would be lower. Thus, for higher excess VOC categories and
for higher fee rates, fewer producers would probably select
the fee option, all else equal. Because the fee will be more
cost-effective only for lower-volume products and lower
2-22
-------
excess-VOC categories, allowing the fee option should have a
relatively small impact on variation from the aggregate
emission reduction targets as long as the fee assessment rate
is not set at an extremely low level.
2.3.3 Product Withdrawal
Up to this point, the analysis has focused on firms
responding to the regulation by choosing the less costly
alternative between reformulation and the fee regulatory
response. However, this view of a producer's likely response
is incomplete because the cost of the regulatory response must
be weighed against the benefits of the action to the firm.
Here the analysis equates regulatory compliance with the
decision to pay the costs and remain in the market. Thus, the
benefits of the compliance action are the net returns
(revenues minus variable costs) obtained from continuing to
produce the product. The net payoff of compliance for a
particular architectural coating exceeding the limit can be
expressed as follows:
nR = P • q - c(q) - r*. (2.1)
To ease the notational burden, all terms are expressed in
their annualized form: P is product price, q is annual
output, c(q) is the product cost function (without regulation)
with respect to annual output, and r* is the annualized cost
of the least-cost option among regulatory responses (i.e.,
reformulation or fee). In other words, r* gives the cost of
the solution to the least-cost decision discussed in the
previous section.
The firm is assumed to select an output level (q*) that
maximizes profits (nR*) . In a competitive market, this is the
point at which the marginal cost of production equals the
market price. However, the firm will only operate in this
2-23
-------
market if it can cover its production costs and compliance
costs; that is, if the following condition is met:
nR* (q** r*) * 0. (2.8)
If the condition in Eq. (2.8) is not met, then the firm's best
response is to withdraw the product, produce no output (q?'=Q),
and generate zero profits for the product (nR*=0) . In this
regard, product withdrawal would be the firm's least cost
option, because the alternative implies they lose money by
remaining in the market.
2.3.4 "Best-Response* Analysis
The analysis presented here determines which option (fee,
reformulation, or withdrawal) is the best response for
specific products within a certain VOC content range from the
survey.
For the purpose of this analysis, a product stratum is
defined as all products existing in a specific VOC content
range for a specific product category. An example of a
stratum would be all exterior flat waterborne products in the
101 to 150 g/L VOC content range. For the TOS, all strata in
the survey were examined to determine those that exceed limits
for their respective product categories. As indicated above,
the survey includes data on the number of products, sales
volume, and baseline VOC emissions for each stratum. These
data were used to compute average sales volume per product for
all strata exceeding the TOS limits. These average volume
estimates formed the basis for computing exceedance fee costs
and product-level profits.
An example of a best response determination is as
follows:
2-24
-------
(Best-Response Example)
Suppose the average sales volume per product for one
stratum is 100,000 L/yr. To determine the exceedance fee
for each stratum, the midpoint of the VOC content range
was used as an estimate of average VOC for the stratum.
This measure was used to compute excess VOC content
because it is consistent with the regulatory definition
of VOC content (grams per liter less water and exempt
compounds) and is available for each stratum.
First the fee rate was adjusted to 1991 dollars by
multiplying the fee rate (in 1996 dollars) of 0.0028/g by
the ratio 1991/1996 of GDP price deflators. The
resulting fee rate is 0.0024/g. Suppose the midpoint of
the stratum is 150 g/L above the limit. The associated
fee per unit would be 150 • $0.0024 = $0.36/L. The total
exceedance fee payment for the product is
($0.36/liter) • 100,000 liters = $36,000 per year. Fixed
recordkeeping costs must also be incurred for products
subject to the fee. Fee-related recordkeeping costs
were estimated to be $590 per product per year.57 Adding
these numbers together, the compliance cost under the fee
option is $36,590 per year. This exceeds the annualized
cost of reformulation ($14,570 per year). Under these
conditions, it is assumed that products in this stratum
would reformulate rather than pay the exceedance fee.d
This decision would be reversed if, for instance, the
stratum exceedance were 50 g/L, in which case the fee
"By conducting the fee-versus-reformulation decision at the stratum
level, and basing the decision on average cost and fee for each stratum, it
is implied that all products within the stratum are identical to the mean
values. In reality, there will be some variation around the mean so that
some producers may find one alternative less costly while others find the
other alternative less costly. This analysis is unable to capture this
heterogeneity with the available data, but presumably these effects are
smoothed out as the analysis compares means across the hundreds of strata
in the survey.
2-25
-------
payments would be $12,000, which, adding in the fixed
cost of $590, is below the reformulation cost per
product.
To simulate the reformulation/fee/withdrawal decision,
per-unit profits were estimated to compare with unit costs for
each stratum and computed as follows:
(2.9)
where P is the output price and m is the profit margin. For
each product category analyzed, the average market price for
the market in which the product category belongs was used (see
Table 2-3) .58-59 The model derives the returns-to-f ixed-factors
(RFF) profit margin as follows:
m = 1 - (variable cost/revenues). (2.10)
The ratio of variable cost to revenue can be computed using
values provided by the NPCA. The variable cost component in
the numerator includes the cost of goods sold plus variable
selling and storage costs. These variable costs comprise
81.7 percent of revenues for the mean producer surveyed by
NPCA, so the estimate of the RFF profit margin is 0.183.
These average reformulation cost per liter and profit
calculations were performed for each stratum above the TOS
limits to determine the relative frequency of reformulation/
fee/withdrawal selections and their impact on compliance
costs. These analyses were performed directly for the survey
population, with the results used to impute values for the
nonsurvey population. Results are presented for the survey
population in Table 2-4.
Under the chosen fee rate of $0.0024 (1991 dollars), the
fee is the preferred alternative for 409 (35.5 percent) of the
2-26
-------
TABLE 2-3
ARCHITECTURAL COATINGS MARKET SEGMENTS BASELINE
DATA FOR 1991
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
Market Segment4
Exterior & high performance
solventborne coatings
Exterior & high performance
waterborne coatings
Interior solventborne
coatings
Interior waterborne coatings
Solventborne primers &
undercoaters
Waterborne primers &
undercoaters
Solventborne clear coatings,
sealers, & stains
Waterborne clear coatings &
stains
Architectural lacquers
Wood preservativesc
Traffic marking paints
Special purpose coatings
Industrial maintenance
coatings
Totals /averages
Quantity
Produced (kL)b
162,937
468,345
94,935
833,434
61,298
75,212
134,678
120,738
40,011
27,449
91,067
34,568
231,261
2,375,933
Value ($103)
540,511
1,046,383
302,264
1,747,341
171,583
160,960
412,743
266,174
83,320
493,965
132,358
141,633
797,006
6,296,241
Average
Price
($/L)
3.32
2.23
3.18
2.10
2.80
2.14
3.06
2.20
2.08
1.45
1.45
4.10
3.45
2.65
• See Appendix A for an explanation of products included in each market
segment.
b The quantities and values are taken from Census data except the quantity
for wood preservatives, which is taken from the survey.
c For wood preservatives the quantity is taken from the survey, but the
price is taken from the Census data.
Sources: U.S. Department of Commerce. Current Industrial Reports: Paints
and Allied Products, 1991. Washington, DC, Government Printing
Office. 1992.
Industry Insights. Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey. Prepared for
the National Paint and Coatings Association in cooperation with
the AIM Regulatory Negotiation Industry Caucus. Final Draft
Report. 1993.
2-27
-------
TABLE 2-4.
BEST-RESPONSE OPTION ANALYSIS-SURVEY POPULATION:
($1991)
FEE = $2,200/TON
Number'
of Products
Above the
Limit
Reformulation
selected
Fee selected
Withdrawal
selected
Total
697
409
46
1,153
Product
Quantity (L)
239,
38,
278,
183
235
772
191
,643
,442
,807
,893
j^j • Total products over-limit times (2/3)
I b If fee not selected, reformulation costs
£j? withdrawal is not selected, reformulation
Reformulation
Cost
$10,160,142
$0
$0
$10,160,142
Fee
Payments
$0
$3,225,366
$0
$3,225,366
for the 409 products that
costs for the 46 products
Foregone
Profits
$0
$0
$415,178
$415,178
Total
Costs
$10,160,
$3,225,
$415,
$13,800,
Cost Savings
from Not
Reformulating6
142
366 $2,738,
178 $255,
686 $2,993,
$0
616
042
658
selected the fee are $5,963,982. If
that selected withdrawal are $670,220.
Source: Industry Insights. Architectural and Industrial Maintenance Surface Coatings VOC Emissions Inventory
Survey. Prepared for the National Paint and Coatings Association in cooperation with the AIM
Regulatory Negotiation Industry Caucus. Final Draft Report. 1993.
-------
1,153 products facing the reformulation versus fee decision."
However, these products only account for 38 million liters of
output, about 14 percent of the volume subject to the
decision, reinforcing the notion that the fee is selected for
lower-volume products. The total fee payment for those
products is about $3.7 million (average is $0.08/L), but the
estimated avoided reformulation cost for the 409 products
choosing the fee is over $5.9 million for a net aggregate
savings to producers of about $2.7 million. Moreover, because
the fee payment is simply a transfer from one sector of
society (architectural coatings producers) to another (the
government), the social cost savings due to incorporating the
fee are the full $5.96 million reformulation cost savings,
less any costs of administering the fee.
Table 2-4 indicates that 46 products elect withdrawal as
the best response strategy to the regulation, which is less
than 0.1 percent of the 4,846 products surveyed. The
estimated foregone profits for those products total
approximately $415,000, which should be considered a component
of "compliance cost" of the regulation. However, this
produces a $255,000 savings to society over the reformulation-
only option.
All told, allowing for options other than reformulation
substantially reduces compliance costs for the survey
population. The option to pay the fee or to withdraw reduces
the compliance cost estimate by about $3.0 million, or about
18 percent of the costs that would be incurred by the survey
population if reformulation were the only compliance option.
•Note that 1,153 products represent two-thirds of the total number
exceeding the limits because the other one-third were assumed to
reformulate without incurring the "major* reformulation cost.
2-29
-------
2.3.5 Tonnage Exemption
All producers will be allowed to exempt the following
quantity of VOC emissions from control that is phased in over
three years:
Period 1: 23 Mg (25 tons)
Period 2: 18 Mg (20 tons)
Period 3: 9 Mg (10 tons)
Because these represent relatively small volumes, especially
after the 3-year phase-in, the tonnage exemption will likely
serve in lieu of the exceedance fee for small volume products
and thereby reduce fee payments by producers employing the
tonnage exemption.
To the extent that the tonnage exemption replaces the fee
as a compliance option for some products, the foregone fee
payments represent the reduced impact on producers. Consider
the post-year 3 case where 9 Mg of VOC emissions are exempted
from control. Suppose that 3.6 Mg of these emissions are
"exceedance" emissions (i.e., emissions above the amount
allowed in the VOC content standards). If a fee were assessed
to these emissions, the cost to the firm would be 3.6 • $2,200
= $7,920 ($1991). Therefore, the exemption allows the firm to
avoid this impact. Note that while this reduces the private
impact on firms subject to the exemption/fee, there is no
corresponding effect on the social cost of the regulation as
the reduced fee payments are just reduced transfers from one
party (producers) to another party (government).
2.4 COST ANALYSIS UNCERTAINTIES
Table 2-5 lists the key assumptions and main areas of
uncertainty surrounding the cost estimates. Items of
2-30
-------
TABLE 2-5. REFORMULATION COST ANALYSIS UNCERTAINTIES
Assumptions
• Initial reformulation cost is $87,000.
• Useful life of reformulation is (1) 8 years, (2) forever.
• Discount rate is 7 percent.
Potential upward bias factors
• Effects of tonnage exemption not considered.
• Costs assumed constant in the future; but may fall over time
as new technology is developed and disseminated.
• Industry trends since 1991 have moved toward lower VOC
formulations.
• Costs may be borne partly by material suppliers.
• Regulatory baseline is changing. State regulations have
been implemented (e.g., Massachusetts), and some producers
have already developed formulations and incurred
reformulation costs to comply with new as well as existing
regulations. These formulas can be applied to a federal
rule at a minimal cost.
Potential downward bias factors
• Costs are confined to the reformulated product itself; users
may incur additional costs to adapt application systems.
• Multiple products may be lumped together as one in the
survey. Therefore, multiple reformulations may be necessary
in some cases where a single reformulation is projected.
Potential factors with unknown directional effects
• Estimate is for a "typical* product; individual products may
differ.
• Lower-bound estimate of 8 years for useful life of
reformulation is speculative.
• Reformulation may positively or negatively affect variable
production costs (e.g., materials).
• Effects on product quality and performance are unknown;
anecdotal evidence shows both positive and negative effects
depending on the product.
• Costs may rise/fall based on amount of "excess VOC" to
reduce.
• The number of reformulations for nonsurveyed products may be
mis-estimated due to lack of data.
2-31
-------
uncertainties are grouped by the likely direction of bias on
the cost estimate: upward, downward, or unknown.
2.4.1 Upward Bias
As indicated in the previous section, one source of
upward bias in the cost estimates is that the analysis does
not directly account for the effect that the tonnage exemption
would have on cost mitigation.
The analysis may overstate reformulation costs incurred
by architectural producers by not explicitly accounting for
cost-saving technological innovation. Spillover effects from
early reformulation efforts could substantially reduce the
costs for other formulas. This may be facilitated by the role
that material suppliers play in developing formulas,
particularly in the case of smaller architectural coatings
manufacturers. Economies of scale may occur because material
suppliers solve the problem for multiple clients and formulas.
Since this rule was initially proposed, for example,
Massachusetts has implemented its own regulation for
architectural coatings. In compliance with that regulation,
104 companies have registered compliant architectural coatings
with the Massachusetts Department of Environmental
Protection.60 Many of those companies operate on a national
scale. Therefore, products those companies make that
currently meet the Massachusetts regulation do not need to be
further reformulated to comply with the national rule. Those
costs are not "backed-out* in this analysis, which imparts an
upward bias of unknown magnitude on the costs presented.
2.4.2 Downward Bias
A couple of factors may lead to an understatement of the
reformulation costs presented here. First, by focusing on
costs to the coatings manufacturer, the current analysis does
not account for any fixed costs that coating users may bear as
they switch to compliant formulas. Based on public comments,
2-32
-------
the item of greatest concern in this category is application
equipment for traffic marking coatings. These costs are now
explicitly addressed in a separate section of this report and
included in the final cost-effectiveness analysis below.
The second item that may cause downward bias in the cost
estimates relates to the definition of products in the survey
data. The analysis treats each survey entry as a separate
product and assigns each noncompliant entry a single
reformulation. If, instead, survey respondents combined
several products requiring several reformulations into one
survey entry, total reformulation costs for the survey
population would be underestimated. It is impossible to
determine whether this is a systematic problem with the survey
data and, if so, the extent to which it biases the current
estimate.
While the reformulation cost estimate is the main source
of uncertainty in the analysis, another item that bears
mentioning relates to the selection of nonreformulation
response options (fee or withdrawal). The analysis assumes
that producers will select the lower-cost option
(reformulation or the fee) and exit if the lower-cost option
exceeds the value of the profit stream. However, some
rigidities (e.g., shortage of scientist hours for new formula
development) might make reformulation difficult in the very
short run. However, the phased tonnage exemption period
mentioned above should provide some relief in overcoming the
short-run rigidity particularly for smaller producers.
2.4.3 Unknown Directional Effects
Several items that have unknown directional effects on
the cost estimates are listed in Table 2-9. Of particular
relevance is the absence of variable production cost effects,
notably the difference in material costs. The EPA was unable
to obtain verifiable information on material cost effects of
2-33
-------
reformulation. Anecdotally, it was suggested that
solventborne material costs might rise in some situations
(e.g., those described in the comment) but might fall in
others (e.g., substitution of water carriers for solvent).
The net effect across all products is unknown. Without any
hard data on the size or direction of material cost effects,
the EPA assumed no net material cost effects in the analysis.
The compliance strategy decision is likely to be
complicated by issues other than cost that relate to the
profitability of reformulation. If a product serves a narrow
market niche, reformulation may fundamentally alter the
product's attributes and erode the niche position. In such a
case, the producer may find that choosing reformulation is not
profitable. Although concerns regarding the regulation's
constraints on product differentiability are undoubtedly real
in some cases, this complexity is not explicitly addressed in
the quantitative analysis, primarily because of the difficulty
in observing both levels of and changes in product quality.
2-34
-------
SECTION 3
ARCHITECTURAL COATINGS MARKET ANALYSIS
In this section, market effects of the regulatory action
are analyzed by presenting a model of how the outcome of the
reformulate/fee/withdrawal decision collectively affects
aggregate supply conditions and market outcomes in the
architectural coatings industry. Then, operationalizing the
model using baseline market data and regulatory costs is
discussed to analyze the social cost effects of these market
outcomes in the architectural coatings industry. The section
ends with an analysis of employment impacts.
3.1 MARKET EFFECTS OF FIRM RESPONSES TO REGULATION
Firms' decisions to either reformulate or pay the
exceedance fee and remain in the market or to do neither and
exit the market collectively affect market outcomes (price,
quantity, and welfare). The change in market price depends on
the aggregate effects of the supply responses of the
individual producers. Product exits will shift the aggregate
supply function inward, and marginal cost effects, such as the
per-unit fee, will shift the function upward. This change can
be expected to raise the post-regulatory market price as the
new equilibrium is attained. This process is described in
more detail in Appendix D.
3-1
-------
Appendix D describes the methodology for incorporating
the reformulation/fee/withdrawal effects into a linked
multiple-market model framework. This appendix also presents
the methodology for measuring the social welfare effects
(e.g., producer and consumer surplus) of the changes in market
equilibrium, which is affected by the regulation.
3.1.1 Model Execution and Results
To estimate the effect of VOC content limits on
architectural coatings markets, a baseline characterization of
affected markets was constructed, empirically estimated shifts
in market supply and demand as a result of the regulations
were computed, and the market equilibrium model was applied to
the data to generate changes in prices and quantities in each
market.
3.1.1.1 Baseline. The coatings categories are grouped
into market segments, as defined in Table 2-3. The price and
quantity data necessary to analyze market effects are not
provided in the survey conducted for this study but are
available from the U.S. Census Bureau Current Industrial
Reports publications.61 Because the Census Bureau categorizes
architectural coatings products differently than they are
classified in the survey for this study, the market segments
were constructed so that data can be used from both sources
and provide the necessary level of resolution for market
analysis. This process resulted in the 13 market segments
presented in Table 2-3. Appendix A provides the details of
this product/market cross-referencing scheme.
Table 2-3 lists quantities and value of shipments for
each market segment. From these data, the average price for
each market was imputed. Because the market segment price is
an average value, it may obscure heterogeneity of products
within each group. Although the model aggregates different
products together to construct individual market segments, the
3-2
-------
objective in aggregating to the market segments in Table 2-3
is to provide a level of resolution that both highlights
differences in the end use of the product (e.g., exterior
coatings versus interior coatings) and distinguishes between
groups that will be affected differently by the VOC content
regulation (e.g., solventborne versus waterborne). Eight of
the 13 segments consist of four pairs of related product
groups; one in each pair represents solventborne products and
the other represents waterborne products (e.g., interior
coatings). Although the products in each of the paired market
segments possess different attributes, they perform similar
functions, thereby suggesting a high degree of product
substitutability in demand. Demand elasticities were
estimated using procedures outlined in Appendix A. Supply
elasticities could not be econometrically estimated because of
data limitations; therefore, the aggregate supply elasticity
for each market segment was assumed to be unitary (1.0) .
3.1.1.2 Quantifying Market Shocks. The best-response
regulatory strategy for each stratum in the survey exceeding
the TOS limits is computed in the previous section. For the
market analysis, the least-cost solution obtained previously
was compared to an estimate of per-unit profits. If the cost
term exceeded the profit term, that stratum was identified as
a "withdrawal" stratum. Throughout this section, the market
results using upper bound of product reformulation cost
($14,573 per year) are presented unless otherwise indicated.
If the profit term exceeded the cost term and the least-cost
option was reformulation, the stratum was identified as a
"reformulation* stratum. If the profit term exceeded the cost
term and the least-cost option was the fee, the stratum was
identified as a "fee" stratum. The model computes the total
quantity share of the withdrawal strata by summing the total
quantity from these strata (Qsx) and dividing by the total
3-3
-------
baseline quantity from all strata for that market segment in
the survey (QST) . This share was then multiplied by two-thirds
(the previously referenced share of all noncompliant formulas
needing reformulation) to compute the market quantity subject
to the withdrawal option, which is denoted as the term RX.B
Rx = (QSX/QST) • (2/3) . (3.1)
Similarly, the model computes the total quantity shares for
the reformulation R superscript) and the fee strata
(F superscript), respectively:
RR = (QSR/QST) • (2/3) (3.2)
RF = (QSF/QST) • (2/3) . (3.3)
Finally, all quantities not allocated to the exit,
reformulation, or fee actions can be viewed as the
unconstrained share:
R° = 1 - Rx - RR - RF. (3.4)
To perform the market and welfare effects calculations,
the initial baseline market-level values for the exiting,
reformulating, fee-paying, and unconstrained sectors are
obtained for reasons explained in the methodology description
in Appendix D. The model derives baseline quantities by
multiplying the quantity shares derived from the survey data
by the initial baseline market quantity, Q0:
•Multiplication by two-thirds incorporates the previously discussed
assumption that one-third of all products exceeding the limit can be
costlessly reformulated (and thus would not be withdrawn) .
3-4
-------
Qx = Rx • Qo (3.5)
QR = RR • Q0 (3.6)
QF = RF • Qo (3.7)
Qu = Ru • Q0. (3.8)
To quantify the supply effects of the per-unit fee on the
fee-paying sector, as indicated in the equilibrium model
discussion in Appendix D, the model computes a value for the
unit fee as follows.
F = E F. • (QsiF/QsF) (3-9)
where FA is the fee for fee-paying stratum i, QsiF is stratum
i's quantity, and N is the number of fee strata in the market.
Finally, note that the measure of producer surplus losses
requires an estimate of marketwide reformulation costs. The
model estimates this cost by taking the estimated number of
(surveyed and nonsurveyed) products in each market opting to
reformulate and multiplying this number by the annualized cost
of reformulation.
Changes in Output and Price. Table 3-1 reports the
estimated output and price effects of the final regulation.
In general, the annual output and price effects are quite
small relative to baseline values. Price increases are
typically well below 1 percent of baseline price, with the
exception of the solventborne primers and undercoaters market
segment, where the projected price increase is $0.012/L
(0.4 percent). In fact, to show any price effect, the change
in price is displayed to the fourth significant digit. In
3-5
-------
TABLE 3-1. REGULATORY EFFECTS ON ARCHITECTURAL COATINGS MARKET OUTPUT AND PRICES
U)
No.
1
2
3
4
5
6_
7
8
9
10
11
12
13
Change in
Output
Produced
Market Segment (103 L)
Exterior & high performance—
solventborne
Exterior & high performance—
waterborne
Interior solventborne
Interior waterborne
Solventborne primers and
undercoaters
Waterborne primers and undercoaters
Solventborne clear coating, sealers,
stains
Waterborne clear coatings and stains
Architectural lacquers
Wood preservatives
Traffic marking paints
Special purpose
Industrial maintenance
Total
-195
69
-36
22
-349
52
-172
19
0
-2
-42
-15
-277
-926
% Change
from
Baseline
-0
0
-0
0
-0
0
-0
0
0
-0
-0
-0
-0
-0
.12%
.01%
.04%
.00%
.57%
.07%
.13%
.02%
.00%
.01%
.05%
.04%
.12%
.04%
Change in % Change
Price from
($1991) Baseline
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
.0029
.0003
.0010
.0001
.0120
.0015
.0029
.0004
.0000
.0003
.0013
.0035
.0083
0
0
0
0
0
0
0
0
0
0
0
0
0
.09%
.02%
.03%
.01%
.43%
.07%
.10%
.02%
.00%
.02%
.09%
.08%
.24%
New
Quantity
(103 L)
162,
468,
94,
833,
60,
75,
134,
120,
40,
27,
91,
34,
230,
2,375,
741
414
900
456
950
264
506
757
Oil
446
025
554
984
006
New
Price
($/L)
$3.32
$2.24
$3.19
$2.10
$2.81
$2.14
$3.07
$2.21
$2.08
$1.45
$1.46
$4.10
$3.46
-------
other words, the average market price for nearly all 13 market
segments changes by less than 1 cent per unit. Estimated
quantity reductions, across all architectural coatings markets
are approximately 926,000 L/yr. This figure is less than
one-tenth of a percent of the industry baseline quantity.
The results indicate differential impacts across market
segments. For example, solventborne primers and Industrial
Maintenance show the largest reduction in output. However,
four of the waterborne market segments show a net increase in
output produced. These projected increases result as
consumers substitute away from the solventborne counterparts
because of the regulation-induced supply contraction and price
increases in those segments. While noteworthy, these
increases are quite small in absolute terms.
Total Social Costs. The method for estimating changes in
consumer and producer welfare effects is demonstrated in
Appendix D. In general, the net welfare effect (social cost)
of the regulation equals the sum of consumer surplus, producer
surplus, and government surplus measures. Costs are
distributed across parties in such a way that reformulating,
fee-paying, and exiting producers experience welfare losses by
incurring the regulatory costs (or withdrawing products) and
consumers bear welfare costs through higher prices. Changes
in consumer surplus measure losses to consumers from higher
prices and foregone consumption. The total change in producer
surplus for each scenario equals the sum of the change in
producer surplus for the exiting products, fee-paying
products, reformulating products, and unconstrained products.
Losses to exiting products reflect the foregone profits the
producers would have received had the products stayed in the
market. Losses for fee-paying products measure the net effect
of fee payments and recordkeeping costs plus the partial
3-7
-------
offset of these losses by the rise in price caused by the
regulation.
In Table 3-2, the producer losses for reformulating
producers total -$20.4 million. The model actually projects
total reformulation costs of $19.0 million, but $0.8 million
of total reformulation costs are recovered from offsetting
price gains accruing to the reformulating producers.
Note that the producer surplus effect for unconstrained
products is positive, reflecting the fact that producers of
these products gain the benefits of the regulation-induced
rise in price, without any change in their cost structure
caused by the regulation. However, the welfare gains accruing
to the unconstrained products are transfers from coating
consumers and, as such, should not be viewed as a net welfare
gain to society due to the regulation.
The net annual welfare cost estimate is $22.3 million.
This is approximately $12 million (41 percent) less than the
initial cost estimate for the regulation under the
reformulation-only scenario (Table 2-2) . Therefore,
accounting for economic responses substantially reduces the
estimate of regulatory costs. Welfare gains accrue to
unconstrained producers through higher prices ($3.2 million)
and the recipient of exceedance fee revenues ($4.0 million),
identified here as the government sector.b However, the
government may redistribute these revenues back to any of the
parties affected directly by the regulations or back to the
citizenry via the Federal Treasury. From society's perspec-
tive, the net welfare effects of the current transfer method
(architectural producers to the government) or alternative
'Note that the difference in losses to fee-paying producers
($4.9 million) and government receipts ($4.7 million) is due to two
factors: the payment of fee-related recordkeeping costs (+$0.6 million)
and gains from offsetting price increases (-$0.4 million).
3-8
-------
TABLE 3-2. ARCHITECTURAL COATINGS MARKET WELFARE EFFECTS
Change in Producer Surplus ($1991)
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
Market Segment
Exterior & high
performance— solventborne
Exterior & high
performance-waterborne
Interior solventborne
Interior waterborne
Solventborne primers and
undercoaters
Waterborne primers and
undercoaters
Solventborne clear
coating, sealers, stains
Waterborne clear
coatings and stains
Architectural lacquers
Wood preservatives
Traffic marking paints
Special purpose
Industrial maintenance
Total
Exiting
Pro-
ducts
($103)
-31.73
-0.92
-13.89
-11.15
-138.79
0
-79.38
0
0
0
-33.16
-6.44
-268.98
-584.44
Fee- Paying
Products
($103)
-960.
-13.
-165.
-4.
-923.
-6.
-500.
-16.
0
-10.
0
-166.
-1,386.
-4,154.
37
60
22
35
23
96
31
76
67
63
52
61
Reformu-
lating
Producers'
($103)
-1,865
-276
-6,584
-64
-1,582
-17
-2,142
0
-416
-288
-1,475
-159
-5,571
-20,444
.42
.45
.09
.01
.38
.74
.77
.23
.86
.26
.66
.14
.01'
Uncon-
strained
(Net)
Total PS
Sector
($103)
252
148
48
107
454
116
230
43
0
4
47
107
1,600
3,163
.74
.42
.77
.77
.72
.42
.17
.36
.28
.91
.94
.84
.33
Total
($103)
-2,604
-142
-6,714
28
-2,189
91
-2,492
26
-416
-295
-1,460
-224
-5,625
-22,019
.79
.54
.43
.26
.67
.72
.30
.60
.23
.25
.51
.79
.80
.73
Change in
Consumer
Surplus
($1991
103)
-463
-161
-93
-107
-726
-116
-396
-43
0
-6
-120
-119
-1,907
-4,263
.76
.17
.62
.81
.69
.66
.04
.38
.89
.82
.53
.30
.66
Change in
Govern-
ment
Surplus
($1991
103)
910.87
1.82
128.44
1.74
913.45
4.51
481.06
1.85
0
9.47
0
140.84
1,370.29
3,964.32
Net
Welfare
Effects
($1991 103)
-2,157
-301
-6,679
-77
-2,002
-20
-2,407
-14
-416
-292
-1,581
-203
-6,162
-22,319
.68
.90
.62
.81
.91
.43
.28
.93
.23
.67
.33
.47
.80
.06
Actual reformulation cost is $19.0 million, but $0.8 million is recovered by producers through price increases.
-------
distributions (e.g., back to architectural producers) are
zero.
As a point of comparison, market results were estimated
subject to the lower-bound cost assumption for reformulation
($6, 090/product/year) . The total welfare cost under that
scenario is $13.2 million per year. Because of the low
reformulation cost, few products would opt for the fee under
that cost scenario.
3 . 2 ARCHITECTURAL COATINGS INDUSTRY EMPLOYMENT IMPACTS
Regulation-induced reductions in industry output may lead
to corresponding reductions in architectural coatings
employment. Employment impacts are estimated by multiplying
the baseline industry employment level (L0) by the
proportional change in industry output from its baseline
level:
AL = (AQ/Qo) • L0. (3.10)
This assumes a fixed relationship between output and
employment, at least for the marginal changes considered here.
Table 3-3 presents the employment impacts results. Total
employment for SIC 2581 is 51,100 employees.62-63 The
architectural coatings sector is a subset of SIC 2581, so the
architectural coatings employment was computed by taking the
ratio of architectural coatings output to SIC 2581 output and
multiplying it by SIC 2581 employment. This produced an
estimate of approximately 26,100 employed in the architectural
coatings sector.
The proportional change in architectural coatings output
was computed by taking the ratio of the change in output from
the market model (summed across all market segments) over
3-10
-------
TABLE 3-3. ESTIMATED EMPLOYMENT EFFECTS
Architectural Coatings
Share of Baseline
Output Change Output* Imputed Employment Change
(103 L) (%) (no. of employees)
-926 -0.039% -10.2
Baseline quantity and employment computations are as follows:
Output
Sector
SIC 2581
Architectural
model
(103 gal)
1,229,800
627,723
(103 L)
4,654,793
2,375,933
Industry Employment
51,100 from Census
26,083 imputed from
output share
Sources: U.S. Department of Commerce. Current Industrial Reports: Paints
and Allied Products, 1991. Washington, DC, Government Printing
Office. 1992.
U.S. Department of Commerce. 1991 Annual Survey of Manufactures:
Statistics for Industry Groups and Industries. Washington, DC,
Government Printing Office. 1992.
baseline architectural coatings output. This computation was
performed for all four scenarios of the market model.
Given that the output change estimates in the market
model are relatively small, it follows that the estimated
employment impacts are also small. Under the standard
scenario, approximately 10 jobs are lost nationwide, a
0.04 percent reduction.
3-11
-------
SECTION 4
TRAFFIC COATING USER COSTS
The economic analysis up to this point has focused
entirely on the primary impacts of the regulation, those borne
directly by producers in the architectural coatings industry
in the form of higher costs and indirectly by the consumers of
architectural products in the form of higher prices. The
driving force of those impacts is the requirement that
noncompliant products must either be reformulated to a
compliant VOC level, be subject to a fee on the excess VOCs
over the allowable level, or be withdrawn from the market.
However, in this section a type of secondary impact is
considered, one that is caused by the costs that users of a
newly compliant product must incur to purchase the special
equipment necessary to apply the compliant coating. The
analysis focuses exclusively on users of traffic marking
paints, primarily consisting of government entities such as
state transportation departments, for whom the costs of
equipment switching are thought to be potentially significant.
While it is possible that other significant secondary impacts
exist, the extent and size of those is unknown and therefore
not quantified in this report.
One complicating factor in estimating the cost of the
regulation for traffic coating users is the fact that
equipment replacement is a normal activity that would occur in
the absence of the regulation. Therefore, rather than viewing
the regulation as creating equipment replacement
4-1
-------
responsibilities, it is more correct to say that a different
(accelerated) time pattern of equipment replacement is
required. This section presents the issue analytically and
then computes the incremental costs imposed on the population
of traffic coating users.
According to the data collected for this study, the
service life of traffic marking coating trucks (stripers) is
typically 20 years.64 If the average truck is midway through
its replacement cycle, it will be replaced 10 years in the
future in the absence of the regulation. However, to apply
waterborne coatings that are likely to result from the
regulation, users will be required to change the application
equipment. The application equipment can be changed by either
purchasing new trucks with the proper equipment or
retrofitting the current trucks with special equipment to
handle the new coatings. The incremental costs of each are
discussed in turn below.
4.1 TRUCK REPLACEMENT COST METHOD
In an example of truck replacement, new trucks will be
purchased now rather than 10 years in the future, and this
acceleration imposes costs on the government entity. To
estimate the costs of this replacement acceleration process,
the cost of a large replacement truck ($250,000) is used to
compute the net present value (NPV) today (at a 7 percent real
interest rate) of replacing the truck 10 years in the future:
NPV(-IO) = $250,000/1.07" = $127,087. (4.1)
Instead, the government entity is now required to replace the
truck today at a cost of
4-2
-------
NPV(O) = $250.000. (4.2)
Assuming no salvage value for the current truck, the NPV cost
of accelerating the next replacement is then the difference in
these values.
Initial net effect = NPV(O) - NPV(-IO) = $122,913. (4.3)
Thus, if the regulation just accelerates the next replacement,
the one-time cost of that acceleration is approximately
$123,000.
However, accelerating the replacement of the current
equipment by 10 years also accelerates the next round of
equipment replacements (from 30 years hence to 20 years hence)
and so on. Thus, the effects reverberate into all future
replacement decisions. This point is demonstrated graphically
by the alternative time lines of expenditures in Figure 4-1.
The regulation effectively moves up the entire replacement
schedule by 10 years. The computation must therefore be
expanded to measure the present value of the current and all
future adjustments. To start, the present value of an initial
$250,000 cash expenditure repeated every 20 years thereafter
is computed:
V(0) = $250,000 + $250,000*(1/( (1.07)20 - 1))
= $337,118. (4.4)
Without the regulation, this stream of costs would be deferred
10 years into the future. Evaluating this in present value
terms gives
V(-10)= V(0)/1.07" = $171,373. (4.5)
4-3
-------
$250,000
$250,000
$250.000
Cost Schedule Without
Accelerated
Replacement
YMT
10 20 30 40
50
$250.000
$250,000
$250.000
Cost Schedule With
Accelerated
Replacement
T*ar
10
20
30
40
50 . .
Figure 4-1. Cost schedules with and without accelerated
replacement.
Thus, the difference in present value between the two
replacement cost streams is the total cost of accelerating
this and all future purchases:
Total net effect = V(0) - V(-10) = $165,744.
(4.6)
This can be viewed as a one-time cost of the regulation
for the component of a government entity's traffic coating
striper fleet that is 10 years old. This explicitly accounts
for the present value of the regulation's effect on all future
replacement costs.
4-4
-------
4.2 EQUIPMENT RETROFIT METHOD
An alternative to early replacement of a traffic coating
truck is to retrofit the current truck with equipment that can
use the compliant coating. This allows the government entity
to continue to use the current truck until the end of its
service life, at which time it will be replaced with a new
truck that is able to apply compliant coatings. Assuming that
the replacement schedule for the truck is unaffected by the
retrofit, then none of the costs of accelerated replacement
just discussed will apply. This is demonstrated in
Figure 4-2. As with the example in Figure 4-1, replacement
costs without the regulation would occur 10, 30, 50, etc.
years hence. Under the retrofit example, the government
entity incurs the retrofit costs now (Year 0) but still
maintains the same future replacement cost schedule.
Therefore, assuming no salvage value for the retrofit
equipment, the one-time cost of the regulation is simply the
cost of purchasing the retrofit equipment in Year 0. The
present value of all future costs is identical with and
without the regulation.
4.3 NATIONAL INCREMENTAL COST CALCULATION
The cost of the regulation for traffic coating users is
computed separately for the estimated current fleet of medium
stripers (Table 4-1) and large stripers (Table 4-2). Costs
are aggregated across both types and summarized in Table 4-3.
4-5
-------
Replacement Cost
Schedule Without
Retrofit
$250.000
$250,000
$250.000
Y«ar
10 20
30
40 50 .
Replacement Cost
Schedule With
Retrofit
T
$250,000
$250,000
$250.000
T**r
10 20
30
40 50
Figure 4-2. Replacement cost schedules with and without
equipment retrofit.
Data on the vintage of the national fleets of medium and
large stripers are provided in the traffic coating analysis
report by ERG.65 The government entities facing the decision
to replace trucks now or to retrofit each vintage striper in
the fleet are assumed to select the option that minimizes the
present value of costs. When the PV of a new truck vs.
retrofit is calculated, it appears that it would cost less for
government entities to retrofit medium trucks that are under
15 years old than to purchase new trucks. As a result, all
medium stripers currently older than 15 years (i.e., will be
replaced within 5 years) are projected to be scrapped (at no
4-6
-------
TABLE 4-1. NATIONAL INCREMENTAL COST OF TRAFFIC COATING
EQUIPMENT REPLACEMENT AND RETROFITS-MEDIUM STRIPERS ($1996)
Assumptions
Baseline year equipment vintage
Replacement cost
T = service life
PV of replacement cost every 20
years
Retrofit cost
Retrofit cutoff age
i = discount rate
Salvage value
.Replacement scheme
1999
$100,000
20
$134,847 computed
$35,000
15
0.07
0
This is the present value (PV) (Year 0) of accelerating the
replacement schedule.
PV
Replacement PV Number
Scheduled Cost Replacement PV of PV Total
Replace- Without Cost With Incremental Replace- Replace- Annualized
Age ment Year Regulation Regulation Cost ments ment Cost
20
19
18
17
16
0
1
2
3
4
134,847
126,025
117,781
110,075
102,874
134,847
134,847
134,847
134,847
134,847
0
8,822
17,066
24,772
31,973
150
150
150
150
150
0
1,323,265
2,559,962
3,715,753
4,795,932
0
92,629
179,197
260,103
335,715
12,394,912 867,644
(continued)
4-7
-------
TABLE 4-1. NATIONAL INCREMENTAL COST OF TRAFFIC COATING
EQUIPMENT REPLACEMENT AND RETROFITS-MEDIUM STRIPERS ($1996)
(CONTINUED)
Retrofit scheme
Assume that replacement schedule is unaffected by retrofit.
Therefore service life of retrofit is equal to the remaining life of the
current equipment.
Scheduled
Replacement =
Useful Life of PV per
Retrofit Retrofit Retrofit
15 5
14 6
13 7
12 8
11 9
10 10
9 11
8 12
7 13
6 14
5 15
4 16
3 17
2 16
1 19
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
35,000
Number of
Retrofits
150
90
90
90
90
90
90
90
90
0
0
0
0
0
0
Sum
PV of
Retrofits
5,250,000
3,150,000
3,150,000
3,150,000
3,150,000
3,150,000
3,150.000
3,150,000
3,150,000
0
0
0
0
0
0
30,450,000
42,844,912
Annual i zed
Cost
367,500
220,500
220,500
220,500
220,500
220,500
220,500
220,500
220,500
0
0
0
0
0
0
2,131,500
2,999,144
* The PV of the replacement scheme is the PV cost of an accelerated replacement
schedule. This is a one-time event; thus, we annualize this value by multiplying
it by the discount rate. All service life issues are implicitly captured in the
PV calculation.
b The PV of each retrofit is $35,000. This is also a one-time cost (i.e., it does
not need to be repeated). Therefore, it is also annualized by multiplying by the
discount rate.
Note: The replacement of retrofitted vehicles will follow the same schedule as
without regulation, so there is no replacement acceleration taking place.
4-8
-------
TABLE 4-2. NATIONAL INCREMENTAL COST OF TRAFFIC COATING
EQUIPMENT REPLACEMENT AND RETROFITS-LARGE STRIPERS ($1996)
Assumptions
Baseline year equipment vintage 1999
Replacement cost $250,000
T = service life 20
PV of replacement cost every 20 $337,118 computed
years
Retrofit cost $45,000
Retrofit cutoff age 17
i = discount rate 0.07
Salvage value 0
Replacement scheme
**** This is the PV (Year 0) of accelerating the replacement schedule.
PV
Replacement PV Number
Scheduled Cost Replacement PV of PV Total
Replace- Without Cost With Incremental Replace- Replace- Annualized
Age ment Year Regulation Regulation Cost ments ment Cost
20
19
18
0
1
2
337,
315,
294,
118
063
452
337,
337,
337,
118
118
118
0
22,054
42,666
25
25
25
551,
1,066,
1,618,
0
361
651
Oil
38,
74,
113,
0
595
666
261
(continued)
4-9
-------
TABLE 4-2. NATIONAL INCREMENTAL COST OF TRAFFIC COATING
EQUIPMENT REPLACEMENT AND RETROFITS-LARGE STRIPERS
($1996)(CONTINUED)
Retrofit scheme
Assume that the replacement schedule is unaffected by retrofit.
Scheduled
Replacement =
Useful Life of PV per
Retrofit Retrofit Retrofit
17 3
16 4
15 5
14 6
13 7
12 8
11 9
10 10
9 11
8 12
7 13
6 14
5 15
4 16
3 17
2 18
1 19
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
45,000
Number of
Retrofits
25
25
25
15
15
15
15
15
15
15
15
0
0
0
0
0
0
Sum
PV of
Retrofits
1,125,000
1,125,000
1,125,000
675,000
675,000
675,000
675,000
675,000
675,000
675,000
675,000
0
0
0
0
0
0
8,775,000
10,393,011
Annual! zed
Cost
78,750
78,750
78,750
47,250
47,250
47,250
47,250
47,250
47,250
47,250
47,250
0
0
0
0
0
0
614,250
727,511
* The PV of the replacement scheme is the PV cost of an accelerated replacement
schedule. This is a one-time event; thus, we annualize this value by multiplying
it by the discount rate. All service life issues are implicitly captured in the
PV calculation.
b The PV of each retrofit is $45,000. This is also a one-time cost (i.e., it does
not need to be repeated). Therefore, it is also annual!zed by multiplying by the
discount rate.
Note: The replacement of retrofitted vehicles will follow the same schedule as
without regulation, so there is no replacement acceleration taking place.
4-10
-------
TABLE 4-3. NATIONAL INCREMENTAL COST SUMMARY FOR TRAFFIC
COATING EQUIPMENT ($1996)
Striper Type PV of Cost Annualized Cost
Medium (see Table 4-1) $42,844,912 $2,999,144
Large (see Table 4-2) $10,393,011 $727,511
Total $53,237,923 $3,726,655
salvage value) and replaced with new trucks, while all medium
stripers under 15 years old are projected to retrofit the
current vehicles. The corresponding age threshold for this
decision is 17 years for large stripers.
Present value costs are computed for each vintage year,
dependent on the replacement/retrofit decision, and then are
multiplied by the number of stripers of that vintage in the
fleet. This calculation is then summed across all vintage
years to estimate the present value of national costs. As
Table 4-3 indicates, the present value of total national costs
is estimated at $53.2 million - $42.8 million for medium
stripers and $10.4 million for large stripers.
This present value figure is the one-time cost of the
regulation for the government entities faced with equipment
replacement. For comparability with the other estimates in
this analysis, this figure must be expressed in annualized
terms. Because the acceleration (and its costs) are a one-
time event not to be repeated in the future, the appropriate
form of annualization is to compute the corresponding
perpetual annuity value-the amount, if paid out in annual
installments into perpetuity, that would have a present value
equal to the one-time cost estimate. This number is computed
simply by multiplying the one-time cost estimate by the
discount rate of 7 percent
4-11
-------
Annualized cost = ($53.2 million) • .07 = $3.7 million
This is the conceptually correct figure for the annualized
costs incurred by government entities to switch equipment for
traffic marking coating application. This annual estimate is
used to compute cost-effectiveness measures in the next
section.
4-12
-------
SECTION 5
SOCIAL COST-EFFECTIVENESS ANALYSIS
The social cost estimates from the market analysis and
the estimate of traffic coating user costs can be used to
compute measures of the social cost-effectiveness of the
regulation. The distinction of "social* cost-effectiveness is
made to illuminate the fact that the costs evaluated are the
net costs imposed on society, i.e., the net welfare costs
estimated in the architectural coatings market plus the
resource costs incurred by traffic coating users to switch
application equipment.
The measure of social cost-effectiveness is computed as
follows:
SCE = (|AWF| + TMEC)/|AE|. (5.1)
|AWF| is the absolute value of the aggregate annual net change
in welfare (i.e., total social costs), summed across all
markets in the market analysis. TMEC is the annualized
traffic marking equipment costs, and |AE| is the absolute
value of emission reductions. The |AWF| of 20.2 million is
produced by the market model. The TMEC value is estimated at
$3.7 million in the previous section and is adjusted to
$3.3 million {1991 dollars) for comparison with the market
results, leading to a total social cost estimate of
$23.5 million. For external reporting purposes, all numbers
will be converted to 1996 dollars later in this section.
5-1
-------
The emissions reduction estimate needs some elaboration.
To correspond with the cost estimates, a national estimate of
emissions reduction must be used. The baseline estimate of
national VOC emissions from regulated architectural coatings
products is 509,900 Mg.*'66 Given the reduction of 20.6 percent
in 1998, the aggregate emissions reduction in 1998 is
105,075 Mg, which is AET. However, the emissions target must
be adjusted by two market-related factors: foregone emissions
reduction due to selecting the fee option and changes (net
reduction) in emissions due to regulation- induced changes in
industry output.
The first adjustment, AE™, was computed by taking the
total quantity of "exceedance" emissions for products electing
the fee option. These targeted emissions reductions will not
be accomplished because of the fee option:
AEra = (AE/VAEs1) • AET. (5.2)
The second adjustment was computed by taking the ratio of
the change in industry output to baseline industry output and
multiplying by baseline industry emissions:
AE° = (AQ/Q0) • E0. (5.3)
AQ is the change in industry output, which is the sum of
market-level changes, Q0 is baseline industry output
(2.375 billion liters), and E0 is baseline emissions
(509,000 Mg indicated above).
Thus, the net emissions reduction is computed as follows:
This estimate is based on a national baseline emissions estimate
provided by Eastern Research Group of 560,900 tons, which is converted to
Mg by multiplying by the ratio of tons/Mg = 0.9072. The result is a
national estimate of 509,900 Mg.
5-2
-------
AE = AET + AEFR + AEQ. (5.4)
Absolute reductions are reported in Table 5-1. The net
reduction equals the targeted reduction, less foregone
emissions reductions (due to fee), plus emission changes due
to changes in industry output via regulation-induced market
interactions.
The analysis focuses on computing social cost per Mg of
emissions reduction based on the market welfare costs and
traffic marking coating user costs estimated in the previous
sections. Table 5-1 presents the results. The social cost-
effectiveness estimate is $247/Mg.
This estimate allows for an evaluation of cost-
effectiveness implications of the fee option. Allowing the
fee reduces social costs (compared to the static national
reformulation cost estimate of $34 million) by about
$12 million but foregoes about 1,802 Mg of emissions
reduction, about 1.7 percent of the targeted reductions.
Dividing the cost savings by foregone reductions approximates
the marginal social cost of the foregone reductions. This
figure is $6,580/Mg, which is substantially higher than the
$247/Mg average social cost-effectiveness measure reported in
Table 5-1. This indicates that the fee's main effect is to
reduce the most expensive emission reductions.
An important implication of these estimates is that the
fee option, while leading to a substantial reduction in the
social costs of the regulation, does not significantly
undercut the emissions reduction target. Moreover, by
charging the VOC exceedance fee, firms that opt for the fee
have a continued incentive to achieve marginal reductions in
VOC content.
5-3
-------
TABLE 5-1. SOCIAL COST-EFFECTIVENESS ESTIMATES
Social Cost:
Architectural
Coatings
Market
($1991)
22,319,063
Traffic
Coating
User Costs
($1991)
3,269,994
Total
Social Cost
($1991)
25,589,057
National
Emissions
Reduction*
(Mg)
105,075
Estimated
Foregone
Emissions
Change1"
(Mg)
1,802
Estimated
Market
Output
Adjustment0
(Mg)
198
Net
Emission
Reduction*
(Mg)
103.471
Social
Cost per
Mg
($1991)
$247
• National estimate of baseline emissions (509,900 Mg) times 0.2061 (estimated proportional emissions reduction in
1998).
b National emissions foregone due to adoption of exceedance fee, as estimated in architectural coatings market model.
c Baseline emissions times ratio of industry market quantity reduction to baseline industry output.
a Net emissions reduction = targeted reduction - foregone reduction + quantity adjustment.
-------
5.1 CONVERSION OF IMPACTS TO CURRENT DOLLARS
As indicated previously, all impacts presented in the
analysis are in constant 1991 dollars. Some commenters
indicated a preference for values to be expressed in more
recent years. Therefore, this section provides a
demonstration of how 1991 dollars can be converted into a
value closer to the current year. This conversion is
performed using the 3GDP price deflator. At the time of this
analysis, the most recent year of data was for 1996; thus a
conversion is provided for 1996. Given that the GDP index in
1991 is 97.4 and in 1996 the index is 111.0, a conversion
factor of 1.1397 can be applied to any value in the report.
Table 5-2 demonstrates the conversion to 1996 dollars. The
estimated annual net social welfare cost of the regulation of
$25.6 million in 1991 dollars converts to $29.2 million in
1996 dollars. Thus, social cost-effectiveness estimate
converts from $247 Mg ($1991) to $282 Mg ($1996).
TABLE 5-2. CONVERSION OF SUMMARY IMPACTS TO 1996 DOLLARS
Impact Estimate
$1991
$1996
Net social cost
Net social cost per
Mg of emissions
reduction
$25.6 million
$247/Mg
$29.2 million
$282/Mg
5-5
-------
-------
SECTION 6
SMALL BUSINESS IMPACT ANALYSIS
The Regulatory Flexibility Act (RFA) of 1980
(5 U.S.C. 601, et seq.), as amended by the Small Business
Regulatory Enforcement Fairness Act of 1996 (SBREFA), requires
the EPA to give special consideration to the effect of federal
regulations on small entities and to consider regulatory
options that might mitigate any such impacts. The EPA is
required to prepare a regulatory flexibility analysis,
including consideration of regulatory options for reducing any
significant impact, unless the Agency determines that a rule
will not have a significant economic impact on a substantial
number of small entities.
The Agency prepared analyses to support both the proposed
and final rules to meet the requirements of the RFA as
modified by SBREFA. The Agency undertook these analyses
because of the large presence of small entities in the
architectural coatings industry and because the initial impact
analysis indicated that there could be a significant economic
impact on a substantial number of small entities if mitigating
regulatory options were not adopted for the rule. The
analysis supporting the proposed rule was published in the
report titled, "Economic Impact and Regulatory Flexibility
Analysis of Air Pollution Regulations: Architectural and
Industrial Maintenance Coatings* (June 1996). The proposed
rule contained a number of provisions to mitigate the rule's
6-1
-------
impact on small businesses, and the Agency requested comment
on additional measures to reduce the impacts.
This section presents the small business impacts and the
final regulatory flexibility analysis, including responses to
significant issues raised by public comments on proposed
compliance options to mitigate the rule's impact on small
entities. After evaluating public comment on the proposed
mitigating options, EPA made a number of changes to the
proposed rule to further mitigate the rule's small business
impacts. As a result, the Agency believes that it is highly
unlikely that the rule will have a significant economic impact
on a substantial number of small entities. However, in light
of the Agency's inability to quantify the effect of the
mitigating options, the EPA has elected to conduct a
regulatory flexibility analysis and to prepare a SBREFA
compliance guide to eliminate any potential dispute on whether
EPA has fulfilled SBREFA requirements.
6.1 BACKGROUND AND AFFECTED ENTITIES
Small businesses can be defined using the criteria
prescribed in the RFA or some other criteria identified by
EPA. The SBA's general size standard definitions for Standard
Industrial Classification (SIC) codes is one way to define
small businesses. These size standards are presented either
by number of employees or by annual receipt levels, depending
on the SIC code. For SIC 2851, Paint and Allied Products (of
which architectural manufacturers represent approximately
40 percent), the SBA defines small business as fewer than 500
employees. The coatings manufacturing industry, however, is
not labor-intensive. For example, given the average value of
shipments per employee (based on data presented in Sections 1
and 3), a firm with 400 employees might have close to
6-2
-------
$100 million in sales (1991 $) . Therefore, use of this SBA
definition would result in almost all firms in the
architectural coatings industry being classified as small,
which does not appear appropriate given the sales level of
many firms. Alternatively, based on input from the regulatory
negotiation process, the EPA has defined small businesses as
having less than $10 million in annual architectural coatings
sales and less than $50 million in total annual sales of all
products. Using this definition, the section assesses the
baseline presence of small producers in specific architectural
coatings markets. The distribution of small producers by
market segment is important because impacts vary substantially
by market segment. After the baseline assessment, an analysis
is performed to estimate the extent to which specialization in
higher VOC products causes small companies to incur
disproportionate impacts. This is followed by an estimate of
the average impacts of regulatory compliance on small
architectural coatings companies, as measured by the ratio of
compliance costs to sales. The role of special provisions
such as the fee and small tonnage exemption allowance are also
examined in terms of their mitigating impacts on small
producers.
6.1.1 Potentially Affected Entities
A regulatory action to reduce VOC emissions from
architectural coatings products will potentially affect the
business entities that produce the products. Firms, or
companies, that produce architectural coatings are legal
business entities that have the capacity to conduct business
transactions and make business decisions. Figure 6-1 shows
the chain of ownership may be as simple as one facility owned
by one company (firm) or as complex as multiple facilities
owned by subsidiary companies.
6-3
-------
Parent Company
Parent Company
Parent Company
(Direct Owner)
Other Companies
or Legal Entities
Subsidiary
Company
(Direct Owner)
I
Subsidiary
Company
(Direct Owner)
Facility
I
Facility
Facility
B
Figure 6-1. Chain of ownership.
Determining the total number of firms that will be
affected by the regulation is difficult because most of the
available Census data are reported at the four-digit SIC code,
and architectural coatings manufacturers, for whom this
regulation applies, are a subset of the entire coatings
industry represented by SIC 2851. The 1987 Census of
Manufactures, Industry Series: Paint and Allied Products
identified 530 companies with shipments of $100,000 or more
6-4
-------
that manufacture architectural and special purpose coatings.a>67
For the purpose of this analysis, 500 architectural coating
manufacturers were assumed to exist. Data from the
Architectural and Industrial Surface Coatings VOC Emissions
Inventory Survey (the survey) conducted by the National Paint
and Coatings Association provided data for 116 firms, 36 of
which identified themselves as having under $10 million in
annual net sales.b-68 While small businesses represent about
31 percent of the firms in the survey, a larger share of
nonsurveyed firms appear to fall in the small business
category.0
6.1.2 Regulatory Requirements
As discussed in Section 2, the regulation constrains
firms that produce architectural coatings products over the
VOC content limits in one of three ways:
• requires they produce products with VOC content under
the established set of limits,
• imposes a fee on each unit of product that exceeds the
limits established in the regulation, or
• requires they withdraw the product from the market.
Thus, absent the small tonnage exemption, firms with a heavy
(baseline) concentration of products above the limit for their
respective product categories are more tightly constrained by
•These are the two Census categories within SIC 2851 where most of
the architectural coatings products are represented, and this figure
includes companies that produce architectural products, whether or not it
is their primary product.
Twelve survey respondents did not indicate company size.
•The 116 survey respondents comprise about one-fifth of the firms
making architectural coatings products but account for about three-fourths
of industry output. Thus the nonsurveyed firms are relatively numerous but
produce relatively little volume. '
6-5
-------
the regulation than those with a lighter concentration of
above-limit products, all else equal.
6 .2 ANALYSIS
The quantitative analysis of small business impacts draws
from the NPCA survey data for the 36 companies classified as
small (less than $10 million in architectural sales and
$50 million in total sales). While this is a relatively small
sample of all potentially impacted small companies (less than
10 percent), it is assumed that the surveyed small companies
are fairly representative of the nonsurveyed small companies.
As described below, efforts were made to expand the sample
beyond the 36 surveyed small companies, but the inability to
estimate firm-specific costs made such an extension
problematic. Therefore the results of this analysis should be
interpreted with the usual caution surrounding small samples.
6.2.1 Baseline Market Presence of Small Architectural
Coatings Producers
Small business presence in specific coatings markets
indicates one dimension of how small firms may be affected by
the regulation. For certain product markets, small businesses
predominate and thus may be disproportionately affected if
limits are particularly restrictive on those categories.
Table 6-1 lists the coatings product categories provided in
the survey.69 The survey data represent producers that account
for approximately three-quarters of the total industry product
volume.d
Small companies produce more than 20 percent of the
products in the survey, but these products account for just
3.6 percent of total coatings volume and 3.7 percent of total
"This is based on the ratio of Census product volume (part of the
total SIC 2851 volume) to the survey product volume.
6-6
-------
TABLE 6-1.
SMALL BUSINESS PRESENCE IN THE ARCHITECTURAL COATINGS MARKET":
SURVEY POPULATION
Total Survey Population
(116 respondents)11
Market
Segment Regulation
12
13
3,4
3.4
1,2
1,2
5,6
13
1,2
1,2
7,8
7,8
11
7,8
7,8
Bond breakers
Magnesite
cement
coatings
Flat, interior
Nonflat,
interior
Flat, exterior
Nonflat,
exterior
Primers
Industrial
maintenance
coatings
Roof coatings
Bituminous
coatings and
mastics
Stains,
semitransp.
Stains, opaque
Traffic
marking paints
Waterproofing
sealers, clear
Average
Imputed Volume
Revenues' Number per
Sales (10' of Product
(10' L) $1991) Products (103 L)
NR
NR
440,498
316,786
188,764
152,705
109,850
92,412
89,515
79,051
53,057
47,168
46,886
37,393
26.184
NA
NA
973,502
700,097
473,799
383,290
268,034
318,820
224,683
198,418
141,131
125,466
67,985
99,464
69,649
1
2
391
529
344
494
634
652
80
34
205
125
85
71
215
NA
NA
1,126.6
598.8
548.7
309.1
173.3
141.7
1,118.9
2,325.0
258.8
377.3
551.6
526.7
121.8
Small Business (36 respondents)
Share of
Total Imputed Number
Sales Sales Volume Revenues of
(10J L) (%) (105 $1991) Products
NR
0.0
4,214
2,802
3,320
4,765
2,780
2,603
19,939
4,450
602
39
5,649
318
3,177
100.0
NA
1.0
0.9
1.8
3.1
2.5
2.8
22.3
5.6
1.1
0.1
12.0
0.9
12.1
NA
0
9,313
6,192
8,334
11,960
6,782
8,980
50,046
11,170
1,601
103
8,192
846
8,452
1
0.0
81
87
92
155
137
78
20
22
18
11
21
14
48
Share of
Total
Products
100.0
0.0
20.7
16.4
26.7
31.4
21.6
12.0
25.0
64.7
8.8
8.8
24.7
19.7
22.3
Average Volume
per Product
(101 L)
NA
0.0
52.0
32.2
36.1
30.7
20.3
33.4
996.9
202.3
33.4
3.5
269.0
22.7
66.2
-------
TABLE 6-1. SMALL BUSINESS PRESENCE IN THE ARCHITECTURAL COATINGS MARKET":
SURVEY POPULATION (CONTINUED)
Total Survey Population
(116 respondents)"
Market
Segment Regulation
Number Category
12
1.2
9
7.8
0\ 10
00
12
5,6
7,8
10
5,6
13
3
7
10
Metallic
pigmented
coatings
High-
performance
Lacquers
Sealers
Clear wood
preservatives
Dry fog
coatings
Quick dry
primers.
undercoaters
waterproofing
sealers , opaque
Semitransp.
wood
preservatives
Undercoaters
Mastic texture
coatings
Quick dry
enamels
Shellacs, clear
& opaque
solventborne
Opaque wood
preservatives
Sales
(105 L)
22,163
19,233
17,351
15,375
15,064
14,107
13,612
10,214
8,902
7,096
6,443
6,390
3,937
2,974
Imputed
Revenues'
<10l $1991)
90,868
48,274
36,090
40,898
21,843
57,837
33,212
27,169
12,909
17,313
22,230
20,321
12,047
4,312
Number
of
Average
Volume
per
Product
Products (10' L)
128
186
65
100
15
68
40
29
32
43
42
85
11
8
173.1
103.4
266.9
153.8
1,004.3
207.4
340.3
352.2
278.2
165.0
153.4
75.2
357.9
371.7
Sales
(101 L)
2,390
145
4,384
350
0
63
0.0
17
0.0
66
2,637
280
0.0
0.0
Small Business (36 respondents)
Share of
Total Sales imputed
Volume Revenues
(%(
10.8
0,8
25.3
2.3
0.0
0.4
0.0
0.2
0.0
0.9
40.9
4.4
0.0
0.0
(101 $1991)
9,798
364
9,118
932
0
259
0
44
0.0
162
9,096
890
0.0
0.0
Number
of
Share of Average
Total volume
Products per Product
Products (%)
50
23
13
9
0
7
0.0
2
0.0
7
10
14
0.0
0.0
39.1
12.4
20.0
9.0
0.0
10.3
0.0
6.9
0.0
16.3
23.8
16.5
0.0
0.0
(101 L)
47.8
6.3
337.2
38.9
0.0
9.0
0.0
8.4
0.0
9.5
263.7
20.0
0.0
0.0
-------
TABLE 6-1. SMALL BUSINESS PRESENCE IN THE ARCHITECTURAL COATINGS MARKET": SURVEY
POPULATION (CONTINUED)
Total Survey Population
(116 respondents )b
Market
Segment
Number
13
12
12
12
12
12
12
10
13
13
12
Regulation
Category
Sanding
sealers
Multicolor
coatings
Concrete
curing
compounds
Form release
compounds
Graphic arts
coatings
Pretreatment
wash primers
Swimming pool
coatings
Below ground
wood
preservatives
High
temperature
coatings
Appurtenances
Antigraffiti
coatings
Sums /averages
Sales
(10' L)
2,796
1,613
1,256
1,026
994
668
656
507
483
205
42
1,853,623"
Imputed
Revenues0
(10' $1991)
9,648
6,614
5,148
4,206
4,075
2,739
2,689
736
1,665
707
172
4,528,916
Number
of
Products
41
4
6
5
30
18
27
8
44
7
8
4,920
Average
Volume
per
Product
(10J L)
68.2
403.3
209.3
205.2
33.1
37.1
24.3
63.4
11.0
29.3
5.3
376.8
Sales
(10> L)
80
0
NR
0
251.3
11.3
126.4
0
4.4
200.5
40.1
65,914
Small Business (36 respondents)
Share of
Total sales
Volume
(%)
2.9
0
NA
0
25.3
1.7
19.3
0
0.9
97.9
95.3
3.6
Imputed
Revenues
(101 $1991)
277
0
NA
0
1,030.5
46.2
518.1
0
15.1
691.6
164.2
166,022
Number
of
Products
8
0
1.0
0
20.0
4.0
14.0
0
9.0
5.0
6.0
992
Share of
Total
Products
(%)
19.5
0
16.7
0
66.7
22.2
51.9
0
20.5
71.4
75.0
20.2
Average
Volume
per
Product
(10' L)
10.0
0
NA
0
12.6
2.8
9.0
0
0.5
40.1
6.7
66.4
See notes at end of table.
(continued)
-------
TABLE 6-1. SMALL BUSINESS PRESENCE IN THE ARCHITECTURAL COATINGS MARKET":
SURVEY POPULATION (CONTINUED)
• Small businesses are defined as producing less than $10 million in architectural coatings products or
less than $50 million in total sales.
b The survey had 116 respondents and 36 of those identified themselves as having under $10 million in
annual sales. Twelve survey respondents did not report company size.
c Revenues were imputed using average prices taken from Section 2. A weighted average was used when the
product category belongs in two market segments.
d The actual total volume reported in the survey is 1,853,658,716 L. The difference here, 35,677 L, is
attributed to bond breakers and magnesite cement coatings, and waterborne quick dry enamels, which are
not reported. The quantity reported here is slightly greater than the total survey quantity used in the
market analysis because quantities in the survey that were identified as either "exempt* or "unknown*
(with respect to solventborne or waterborne) were not included in the total for the market analysis.
i NR = Not reported due to disclosure of individual companies.
»->
0 NA = Not available.
Source: Industry Insights. Architectural and Industrial Maintenance Surface Coatings VOC Emissions
Inventory Survey. Prepared for National Paint and Coatings Association in cooperation with the
AIM Regulatory Negotiation Industry Caucus. Final Draft Report. 1993.
-------
revenue. This is evidence that small businesses tend to
produce lower volumes per product. The average price per
product in the small business segment is $2.52/L, compared to
$2.44/L for the industry. The largest volume category for
small producers is roof coatings, at 19.9 million L/yr. Small
producers comprise just over 22 percent of the volume in that
category. Small businesses produce over 95 percent of the
total volume of antigraffiti coatings, but the volume is quite
low, with six products totaling about 40,060 L.
Other categories in which small producers comprise more
than 20 percent of the market volume are lacquers, mastic
texture coatings, graphic arts coatings, bond breakers, and
appurtenances. In addition to roof coatings, small producers
collectively produce over 4 million L in the following
categories: traffic marking paints, exterior nonflats,
bituminous coatings, lacquers, and interior flats.
6.2.2 VOC Content of Small Business Products: Technology and
Specialization Effects
The extent to which small businesses are affected by the
architectural coatings regulation will depend partly on the
average VOC content of small business products relative to the
industry average. Table 6-2 presents the average baseline VOC
content for products manufactured by small businesses as
compared with those manufactured by the industry as a whole.70
Small business products generate approximately 6.2 percent of
total VOC emissions in the survey, which is substantially
greater than their output share. The average VOC content for
small business products, 325 g/L, is almost 75 percent higher
than the average VOC content for all surveyed products
combined, 186 g/L.
Small business products have a higher VOC content than
the industry average for two possible reasons. First, small
businesses specialize in products that tend to be higher in
6-11
-------
TABLE 6-2. BASELINE VOC CONTENT
Size
Category*
All products
Small business
products
VOC
Emissions
(Mg)
344,059
21,431
Sales
(kL)
1,853,623
65,914
Average VOC
Content
(g/L)
186
325
a The survey had 116 respondents and 36 of those identified themselves as
having under $10 million in annual sales. Twelve survey respondents did
not report company size.
Source: Industry Insights. Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey. Prepared for
National Paint and Coatings Association in cooperation with the
AIM Regulatory Negotiation Industry Caucus. Final Draft Report.
1993.
VOCs because of fundamental performance requirements of the
products. Second, small businesses tend to produce higher
VOC-content products regardless of the product category. The
first reason can be called a specialization effect and the
second reason a technology effect.
Some further clarification may be in order. Many of the
small companies in the architectural coatings industry are
regional firms whose product line is tailored to the region in
which they operate and may tend to focus on smaller "niche"
markets for which larger manufacturers may not choose to
devote manufacturing and marketing resources. Thus small
businesses may "specialize* in higher VOC coatings within
categories. Therefore, what is referred to here as a
technology effect {higher VOC within categories in which small
and large manufacturers compete) may be caused by
specialization strategies. In other words, some technology
effect may actually be due to specialization within a
category. With that caveat in mind, this report refers to
across-category factors as the specialization effect and
within-category factors as the technology effect.
6-12
-------
Distinguishing between specialization and technology
factors underlying small companies' higher VOC content is
important in terms of the scope for regulatory flexibility.
To the extent that the specialization effect dominates, small
business impacts can potentially be addressed by modifying the
VOC limits in the high VOC categories where small companies
specialize. If the technology effect dominates, there is less
scope for modifying category limits to reduce impacts.
The observed difference in average VOC content of small
businesses and all products was separated into the
specialization and technology effects using a simple
procedure. First, a measure of the projected average VOC
content of small business products was computed. The
projected value was based on the distribution of small
business products among the different product groups, weighted
by the average VOC content of each group. This is a measure
of its specialization-based VOC content:
N
Vs = E Vi1 • SiB . (6.1)
i=l
Here, V^ is the industry average VOC content for all products
in product category i, SiB is the share of total small business
product quantity attributable to product category i, and N is
the total number of product categories.8 The separation of
the average VOC content difference into the two component
effects derives from the following equation:
(VB - V1) = (VB - Vs) + (Vs - V1) (6.2)
Difference = Technology + Specialization
in Average Effect Effect
Content
•s^ is not the small business share of total production in category
I, but rather the contribution of category I to total small business
production.
6-13
-------
VB and V1 are, respectively, the small business and
industrywide VOC content averages. The technology effect
quantifies the difference between the actual average VOC
content for small businesses and the specialization-adjusted
average. The specialization effect quantifies the difference
between the specialization-adjusted average for small
businesses and the overall industry average.
Table 6-3 yields the computation of the Vs measure for
the small business products in the survey.71 The computed Vs
value is 261, meaning that one would expect an average VOC
content of 261 g/L for the small business sector, based purely
on the way their products are distributed among product groups
(i.e., their specialization). Placing this value into Eq.
(6.2), along with the values for VB and V1 given above (325 and
186), the breakdown is computed as follows:
{Vs - V1) = (VB - Vs) + (Vs - V1)
(325-186) = (325-261) + (261-186)
139 =64 +75
Approximately 54 percent of the 139 g/L difference
between the small business sector's VOC content average and
the industrywide average can be attributed to greater
specialization in high-VOC product categories (specialization
effect), and the remaining 46 percent can be attributed to the
disproportionate presence of small business products in the
high-VOC end of the respective product categories (technology
effect).
As indicated above, this finding has implications for the
feasibility of designing a TOS to minimize small business
impacts. Since small business producers are somewhat
concentrated in the higher VOC categories, as indicated by the
6-14
-------
TABLE 6-3.
SPECIALIZATION-BASED AVERAGE VOC CONTENT:
SMALL BUSINESS PRODUCTS'
Market
Segment
Number
12
12
1,2
11
1,2
1,2
9
3,4
1,2
7,8
3,4
5,6
13
13
12
7,8
7,8
7,8
3
12
7
13
1,2
Regulation
Category
Bond breakers
Concrete curing
compounds
Roof coatings
Traffic marking
paints
Nonflat,
exterior
Bituminous
coatings and
mastics
Lacquers
Flat, interior
Flat, exterior
Varnishes
Nonflat,
interior
Primers
Mastic texture
coatings
Industrial
maintenance
coatings
Metallic
pigmented
coatings
Stains,
semi transparent
Sealers
Waterproofing
sealers, clear
Quick dry
enamels
Graphic arts
coatings
Shellacs, clear
fc opaque
solventbome
Apurtenances
High performance
All
Products
Average VOC
(g/L)
NA
621
239
369
173
23
657
52
79
474
134
172
146
374
459
475
312
632
461
366
539
411
335
Share of Total
Small Business
Volume
NA
NA
0.3025
0.0857
0.0723
0.0675
0.0665
0.0639
0.0504
0.0482
0.0425
0.0422
0.0400
0.0395
0.0363
0.0091
0.0053
0.0048
0.0042
0.0038
0.0032
0.0030
0.0022
Share-
Weighted
Content
Factor
-------
TABLE 6-3. SPECIALIZATION-BASED AVERAGE VOC CONTENT:
SMALL BUSINESS PRODUCTS' (CONTINUED)
Market
Segment
Number
12
13
5,6
12
12
7,8
7,8
12
13
10
10
10
10
12
12
13
13
5,6
All
Products
Regulation Average VOC
Category (g/L)
Swimming pool
coatings
Sanding sealers
Undercoaters
Dry fog coatings
Antigraf f iti
coatings
Stains, opaque
Waterproofing
sealers, opaque
Pretreatment wash
primers
High- temperature
coatings
Below ground wood
preservatives
Clear wood
preservatives
Opaque wood
preservatives
Semi transparent
wood
preservatives
Form release
compounds
Multicolor
coatings
Fire-resistant/
retardant
coatings
Magnesite cement
coatings
Quick dry
primers,
undercoaters
Sums / averages
552
525
206
300
397
257
239
706
561
541
419
362
548
599
321
16
NA
439
Share of Total
Small Business
Volume
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
.0019
.0012
.0010
.0010
.0006
.0006
.0003
.0002
.0001
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
share -We i ghted
Content Factor
(g/L)
1.06
0.64
0.21
0.29
0.24
0.15
0.06
0.12
0.04
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NA
0.00
260.87"
* Small businesses are defined as producing less than $10 million in architectural
coatings products or less than $50 million in total sales.
b Specialized average VOC content equals the sum of share-weighted content
factors.
NA = Not available
Source: Industry Insights. Architectural and Industrial Maintenance Surface
Coatings VOC Emissions Inventory Survey. Prepared for National Paint and
Coatings Association in cooperation with the AIM Regulatory Negotiation
Industry Caucus. Final Draft Report. 1993.
6-16
-------
empirically sizable specialization effect, the regulation can
be designed to be somewhat less restrictive in categories with
high small business presence. However, the effectiveness of
such an approach in mitigating small business impacts will be
limited by the fact that small business producers are also
concentrated in the high-VOC range of each product category.
An additional approach taken by the EPA was to evaluate
requests for additional categories to determine if a breakout
category for products in the higher-VOC range of a category
was needed.
In 1993, the National Paint and Coatings Association
(NPCA) analyzed the VOC content limits that were under
discussion during the regulatory negotiation and found that
the projected emissions reduction from the small business
sector would be 19.65 percent of baseline emissions, compared
to a projected 25 percent reduction for the industry.72 This
estimate provides some evidence of relief for small business
products under the standards under consideration at the time.
Moreover, the final regulation is less stringent than the form
provided to NPCA in 1993. Unfortunately, data were not
available to recompute these estimates based on the current
content limits to see whether the proportional reduction from
the small business sector is still less than the current
overall reduction target of 20 percent.
6-2.3 Costs Associated With Regulatory Compliance
As discussed in Section 2, compliance options that can be
quantitatively evaluated include product reformulation and the
payment of an exceedance fee. The cost of a typical
reformulation is estimated at $87,000 per reformulation.73
This initial cost is converted to an annualized cost of
6-17
-------
$14,573.£ The per-unit fee that producers can use as an
alternative compliance mechanism is computed as follows:
fee = (VOC content - VOC limit) • rate, (6.3)
VOC content is measured in grams per liter, and the fee rate
is paid on the grams per liter in excess of the limit. The
fee rate is $2,500 per ton or $0.0028 per excess g/L (in .1996
dollars, $0.0024 when converted to 1991 dollars). Total fee
payment per product simply equals the per-liter fee multiplied
by total liters of production.
6.2.4 Reformulation Cost Impact Estimates
Given the data from the survey and the VOC content limits
set by the standard, the number of products produced by small
businesses that exceed the VOC limits were identified. The
number of potential reformulations was estimated by applying
the content limits to the number of products reported by
category and VOC content in the survey to determine the number
exceeding the limit for each category. Results are reported
in Table 6-4.7* An estimated 421 small business products in
the survey (42 percent) exceed the VOC content limits. This
figure is slightly higher than the proportion of all surveyed
products that exceed the limit (36 percent). As established
in Section 2, approximately one-third of products over the VOC
limit can costlessly comply with the regulation because of
their similarity to the remaining over-the-limit products that
are being reformulated. The remaining over-the-limit products
are referred to as "constrained" by the regulation and the sum
of the costless compliance products and under-the-limit
products as "unconstrained" by the regulation.
'Details of the derivation of these estimates are presented in
Section 2 of this report.
6-18
-------
TABLE 6-4. SMALL BUSINESS COSTS BY MARKET SEGMENT: REFORMULATION OPTION ONLY"
Market
Segment Regulation
Number Category
12
12
1,2
11
1,2
1,2
ON 9
^ 3,4
10 1.2
7,8
3,4
5,6
13
13
12
7,8
7,8
Bondbreakers
Concrete curing
compounds
Roof coatings
Traffic marking
paints
Nonflat,
exterior
Bituminous
coatings and
mastics
Lacquers
Flat, interior
Flat, exterior
Varnishes
Nonflat,
interior
Primers
Mastic texture
coatings
Industrial
maintenance
coatings
Metallic
pigmented
coatings
Stains,
semitransparent
Sealers
Sales
(L)
NA
NA
19,938,649
5,649,468
4,764,963
4,450,233
4,383,825
4,214,045
3,320,278
3,177,311
2,801,665
2,779,526
2,636,665
2,602,918
2,389,690
601,978
350,434
Average
Price"
($/L)
$4.10
$4.10
$2.51
$1.45
$2.51
$2.51
$2.08
$2.21
$2.51
$2.66
$2.21
$2.44
$3.45
$3.45
$4.10
$2.66
$2.66
imputed
Revenues
($1991)
NA
NA
$50,046,010
$8,191,729
$11,960,057
$11,170,084
$9,118,356
$9,313,039
$8,333,897
$8,451,646
$6,191,679
$6,782,044
$9,096,494
$8,980,067
$9,797,729
$1,601,261
$932,155
Number of
Products
over
VOC Limit
0
0
13
14
70
1
5
13
34
33
26
79
1
33
19
12
2
Number of
Constrained
Products'
0
0
9
9
47
1
3
9
23
22
17
53
1
22
13
8
1
Total Cost of
Reformulating
Constrained Reform Cost
Products* per Unit
($1991) ($/L)
$0
$0
$126,273
$135,987
$679,933
$9,713
$48,567
$126,273
$330,253
$320,540
$252,547
$767,353
$9,713
$320,540
$184,553
$116,560
$19,427
NA
NA
$0.01
$0.02
$0.14
$0.00
$0.01
$0.03
$0.10
$0.10
$0.09
$0.28
$0.00
$0.12
$0.08
$0.19
' $0.06
Reformulation
Cost /Revenues
(%)
NA
NA
0.3%
1.7%
5.7%
0.1%
0.5%
1.4%
4.0%
3.8%
4.1%
11. 3*
0.1%
3.6%
1.9%
7.3%
2.1%
(continued)
-------
TABLE 6-4.
SMALL BUSINESS COSTS BY MARKET SEGMENT:
(CONTINUED)
REFORMULATION OPTION ONLY"
Market
Segment Regulation
Number Category
7,8
3
12
7
13
? 1.2
^ 12
O
13
5.6
12
12
7.8
7,8
12
13
Waterproofing
sealers, clear
Quick dry
enamels
Graphic arts
coatings
Shellacs, clear
& opaque
solventborne
Apurtenances
High performance
Swimming pool
coatings
Sanding sealers
Undercoaters
Dry fog coatings
Antigraffiti
coatings
Stains, opaque
Waterproof ing
sealers, opaque
Pre treatment
wash primers
High- temperature
coatings
Sales
(L)
318,046
279,893
251,343
210,817
200,473
144,958
126,358
80,170
66,427
63,187
40,060
38,668
16,726
11,272
4,383
Average
Price"
($/L)
$2.66
$3.18
$4.10
$3.06
$3.45
$2.51
$4.10
$3.45
$2.44
$4.10
$4.10
$2.66
$2.66
$4.10
$3.45
Imputed
Revenues
($1991)
$846,
$890,
$1,030,
$645,
$691,
$363,
$518,
$276,
$162,
$259,
$164,
$102,
$44,
$46,
$15,
002
060
506
100
630
844
070
587
081
066
248
856
491
214
121
Number of
Products
over
VOC Limit
9
9
8
0
5
12
0
1
3
6
1
6
2
4
0
Total Cost of
Reformulating
Number of Constrained
Constrained Products'
Products0 ($1991)
6
6
5
0
3
8
0
1
2
4
1
4
1
3
0
$87,
$87,
$77,
$48,
$116,
$9,
$29,
$58,
$9,
$58,
$19,
$38,
420
420
707
$0
567
560
$0
713
140
280
713
280
427
853
$0
Reform Cost
per Unit
$0.27
$0.31
$0.31
$0.00
$0.24
$0.80
$0.00
$0.12
$0.44
$0.92
$0.24
$1.51
$1.16
$3.45
$0.00
Reformulation
Cost /Revenues
10.3%
9.8%
7.5%
0.0%
7.0%
32.0%
0.0%
3.5%
ia.o%
22.5%
5.9%
56.7%
43.7%
•4.1%
0.0%
(continued)
-------
TABLE 6-4.
SMALL BUSINESS COSTS BY MARKET SEGMENT:
(CONTINUED)
REFORMULATION OPTION ONLY8
to
Market
Segment
Number
10
10
10
10
12
12
13
13
5,6
Regulation
Category
Below ground
wood
preservatives
Clear wood
preservatives
Opaque wood
pr eservat ives
Semi transparent
wood
preservatives
Form release
compounds
Multicolor
coatings
Fire-resistant/
retardant
coatings
Hagnesite
coatings
Quick dry
primers,
undercoaters
Total /Average
Sales
(L)
0
0
0
0
0
0
0
0
0
65,914,429
Average
Price"
(S/L)
$1.45
$1.45
$1.45
$1.45
$4.10
$4.10
$3.45
$3.45
$2.44
$2.52
Imputed
Revenues
($1991)
$0
$0
$0
$0
$0
$0
$0
$0
$0
$166,022,123
Number of
Products
over
VOC Limit
0
0
0
0
0
0
0
0
0
421
Number of
Constrained
Products0
0
0
0
0
0
0
0
0
0
281
Total Cost of
Reformulating
Constrained
Products'1
($1991)
$0
$0
$0
$0
$0
$0
$0
$0
$0
$4,089,313
Reform Cost
per Unit
($/L)
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.06
Reformulation
Cost /Revenues
(%)
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
2.5%
NA * not available
* Small businesses ar« defined as producing less than $10 million in architectural coatings products or less than $50 million in
total sales.
b Average prices are taken from Section 2, and a weighted average is used when the product category belongs in two market
segments.
c Number of products over the limit multiplied by two-thirds to represent the portion that potentially undergo major
reformulation. Numbers in table are rounded.
d Annual!zed cost of reformulations is the number of products facing reformulation multiplied by the annualized major
reformulation cost estimate per product of $14,573 (details in Section 2).
Source: Industry Insights. Architectural and Industrial Maintenance Surface Coatings VOC Emissions Inventory Survey. Prepared
for National Paint and Coatings Association in cooperation with the AIM Regulatory Negotiation Industry Caucus. Final
Draft Report. 1993.
-------
Less than 10 percent of the small business products in
the sanding sealers, mastic texture coatings, and bituminous
categories will be constrained by the regulation. Swimming
pool coatings, shellacs, and high-temperature coatings
produced by the small business sector will require no
reformulations. Traffic paints, roof coatings, and varnishes
are all relatively high-volume categories in which over
40 percent of the surveyed small business products are
constrained by the VOC limits.
6.2.4.1 Small Business Impacts Under "Reformulation-
Only* Option. In this section, the estimation of the total
and per-unit annualized compliance costs for small producers
in each product category with reformulation as the only
compliance option is described. As with the impacts presented
in Section 2, the "reformulation-only" scenario gives the
upper bound of regulatory costs. The effect of cost-reducing
strategies (fee and withdrawal) is considered in the next
subsection.
The annualized $14,573 estimate of the cost per
reformulation was multiplied by the number of products
constrained by the regulation (all products over the limit
less the one-third that can costlessly comply). Table 6-4
lists the cost estimates. These costs can be compared with
revenue information to gauge the relative impact of the
regulation on small businesses.
To compute product revenue, the analysis uses average
price per liter for each category (see Sections 2 and 3) for
the market segment in which the category is classified.0 The
cost of reformulation as a percentage of revenues was computed
using the estimated cost of reformulation divided by the
"Where a coating category could not be separated into waterborne and
solventborne market segments (categories in market segments 1 through 8), a
weighted average of the two prices was used.
6-22
-------
imputed revenues for each product category. Ideally, costs
would be calculated for each firm affected by the regulation
and compared to the firm's revenues as a firm-specific measure
of impacts. Then, these measures could be used to determine
the number and percentage of firms exceeding certain
cost/revenue threshold values, e.g., 1 percent or 3 percent.
What constitutes a significant impact varies, depending on
typical profit rates and other industry-specific factors.
Unfortunately, the product-level survey data used to
estimate costs did not identify the firms that produced each
surveyed product. Therefore, it was not possible to estimate
costs at the firm level. In lieu of the firm-level measures,
the analysis calculated cost/revenue affects per market
segment (in Table 6-4) and the average cost/revenue ratio per
small company using summary totals from the small business
component of the survey (in Table 3-5).
6.2.5 Cost Impacts Across Market Segments
The data presented in Table 6-4 illustrate a number of
scenarios pertaining to potential small business impacts of
the regulation under a reformulation-only response scenario.
Key phenomena indicated by the data are examined below.
Based on the survey data, roof coatings is the largest
quantity and highest revenue category for small businesses.
For small business roof coatings, 43 percent of the individual
products will be constrained; however, the cost of
reformulation as a percentage of sales is relatively small,
less than 1 percent.
Categories with cost/revenue ratios in excess of
10 percent are highlighted in bold in Table 6-4. The three
highest impact categories are opaque waterproofing sealers
(43.7 percent), opaque stains (56.7 percent), and pretreatment
6-23
-------
wash primers (84.1 percent).11 In each case, the large impacts
result from the fact that the average product volumes are very
small (e.g., just 2,800 liters per product in pretreatment
wash primers). This provides further evidence of the point
made throughout the report that the impact on small volume
products is potentially large because of the fixed cost nature
of reformulation. Obviously the impacts would be dramatic if
these products were forced to reformulate. However, the fee
option provides relief from these high impacts. Therefore,
the highest proportional impacts estimated in Table 6-4 would
not occur with the fee as a compliance option. If, for
instance, an average size pretreatment wash primer
(2,800 liters) were 100 g/L over the limit for the category,
then the total fee payment would be (100 g/L) • $0.0022/g •
2,800 1 = $616. Clearly the producer's cost-minimizing
compliance option would be to choose the fee rather than incur
the annualized reformulation cost of almost $15,000. As a
result, the 84.1 percent figure greatly overstates the true
cost impact for the prototypical pretreatment wash primer
product. Given the fee amount just computed, the figure would
be closer to 5 percent of revenues for that category. Similar
arguments can be made for the other categories representing
the highest impacts in Table 6-4. Further quantitative
evidence of the cost savings from the fee (and withdrawal)
compliance options is presented below.
Antigraffiti coatings present quite a different small
business impact outcome. Small businesses represent almost
the entire market but produce small quantities in relation to
other coating categories and generate lower revenues. Only
one product requires reformulation under the VOC limits, but
"This analysis is based on the interim standards presented in
Section 2. As indicated in Section 7, the content limit for opaque
waterproofing was raised in the final standards. Thus, the cost impact for
that category would likely be lower than indicated here.
6-24
-------
the cost of reformulating that product would represent about 6
percent of revenues in the category.
6.2.6 Average Cost Impacts for Small Company
For the small business segment of the architectural
coatings industry overall, 42 percent of the products are over
the VOC content limits, and 28 percent are expected to undergo
reformulation, pay an exceedance fee, or exit. The total
annualized cost for the sample of small businesses in the
survey under the reformulation-only scenario is $4.1 million.
The average cost per unit is $0.06 per liter.
Table 6-5 compares small firm and industry averages for
revenues, number of products, and reformulation costs.75 Small
businesses on average manufacture approximately one-third
fewer products than the industry average. On average, small
firms have fewer constrained products than the industry
average, but they comprise a slightly larger percentage of
total number of products, 28 percent, as compared to
23 percent for the industry. Similarly, small business
reformulation costs as a percentage of revenues are higher at
2.5 percent than the industry at roughly 0.4 percent.
In response to concerns expressed in the public comment
period about the limited coverage of firms used to assess
small business impacts, EPA obtained a list identifying small
businesses in the industry and gathered data on total revenues
and employment for these firms. However, without specific
information on the number of products produced and their VOC
content, there is no method to determine the number of
products for each firm that would incur reformulation costs.
Unfortunately, assigning the average costs for a small firm
presented here (based on 7.8 noncompliant products) cannot
produce a meaningful evaluation of the distribution of small
firms' impacts. This occurs because the calculation of
cost/revenue ratios for these firms varies the denominator
6-25
-------
TABLE 6-5. AVERAGE REGULATORY IMPACT BY FIRM SIZE-
*REFORMULATION-ONLY* SCENARIO"
Revenue" ($1991)
Number of products"
Number of products facing major
reformulation0
Annualized reformulation cost* ($1991)
Ratio of annualized reformulation
cost to revenues (percent)
Industry
Average
38,990,000
42.4
9.9
144,272
0.4
Small Firm
Average
4,614,000
27.5
7.8
113,669
2.5
• The survey has 116 respondents and 36 of those identified themselves as
having under $10 million in annual sales. Twelve survey respondents did
not report company size.
b Data for revenues and products per firm were based on data reported in
Table 6-1. The number of products per firm is based on the total number
of products for which quantity data are available.
c This number represents two-thirds of the products over the 1998 TOS.
Industry experts estimate that approximately two-thirds of the products
with VOC contents exceeding the TOS limits face a 'major* reformulation.
d Annualized cost of reformulation is the number of major reformulations
multiplied by the annualized reformulation cost estimate per product of
$14,573 ($1991).
Source: Industry Insights. Architectural and Industrial Maintenance
Surface Coatings VOC Emissions Inventory Survey. Prepared for
National Paint and Coatings Association in cooperation with the
AIM Regulatory Negotiation Industry Caucus. Final Draft Report.
1993.
(revenues) by firm, but the numerator (compliance costs)
remain fixed as those represented by the model (average) firm.
Using this method, the estimated impacts would, by definition,
be relatively larger for firms with smaller revenues.
However, it does not necessarily follow that a firm with low
revenues would have the same level of reformulation costs as a
firm with larger revenues; such an analysis would therefore
overstate impacts on the smallest firms. Therefore, for the
final rule EPA uses the data from the 36 firms in the survey
to provide a representative look at model company small
business impacts as described above.
6-26
-------
6.2.7 Potential Factors Mitigating Small Business Impacts:
Exceedance Fee, Withdrawal, and Tonnage Exemption
6.2.7.1 Fee and Withdrawal Options. As discussed in
Section 2, a product's output level affects the choice between
reformulating the product and paying an exceedance fee. Since
the cost of reformulation is a fixed cost (i.e., it is
independent of output level), the average reformulation cost
per unit of output falls as output levels increase. However,
the exceedance fee per unit of output is constant with respect
to the output levels and the fixed costs of the fee
(recordkeeping) are relatively small. Thus, the fee is more
likely to be chosen by small-volume producers, all else equal.
Because the fee will be more cost-effective only for
lower-volume products and lower-excess VOC categories,
allowing the fee option should have a relatively small impact
on variation from the aggregate emissions reduction targets as
long as the fee assessment rate is not set at an
inappropriately low level. The results presented in Section 2
support this point. Therefore, the fee option provides
increased flexibility for small businesses by placing an upper
limit on the per-unit costs of complying with the regulation,
without significantly jeopardizing VOC emissions reduction
targets.
It is not possible to directly conduct a best-response
(least-cost) analysis of the fee/reformulation decisions for
the small business segment of the survey because of
insufficiently detailed VOC data on small businesses.
However, the results of the best-response analysis in
Section 2 can be employed to indirectly measure the effect of
alternative compliance strategies on the relative size of
small business impacts.
Based on survey data for the small business segment, the
average small firm has 27.5 products, 7.8 of which would be
6-27
-------
constrained by the regulation. Table 6-6 divides the average
small company's number of constrained products into three
compliance categories: reformulation, fee, and withdrawal.
The average number of products selecting each strategy is
based on the average percentage of all constrained products in
the survey (small company and large) that select each option.
TABLE 6-6. AVERAGE REGULATORY IMPACT FOR SMALL
COMPANIES-"BEST-RESPONSE" SCENARIO
Compliance
Strategy
Reformulate
Fee
Withdrawal
Total
Average percent
Percent of
All
Constrained
Survey
Products
Selecting
Option
60.5%
35.5%
4.0%
100.0%
of sales
"Expected"
Number of
Products
Selecting
Strategy3
4.7
2.8
0.3
7.8
Average
Compliance
Cost per
Product
(1991 $)
14,573
7,197b
12,705°
11,879
Compliance
Cost
(1991 $)
68,767
19, 936
3, 955
92, 658
2.0%
a Equals average number of constrained products for small companies (7.!
multiplied by percentage of all constrained products in the survey
selecting each strategy.
b Average fee cost computed by taking the average fee rate ($0.084/L),
multiplying by the average size per small company product (65,914 L),
and adding the recordkeeping cost per product of $590.
c Equals the average value of foregone profits for the 46 surveyed
products that select the fee as the best-response strategy.
This is expected to be a conservative assumption because small
volume products produced by small businesses are more likely
to select the fee option to reduce regulation costs.
Compliance costs were estimated by multiplying the number of
products in each category by the per-product cost of that
strategy. Summed across all products, the per-company
compliance costs fall to about $88,000, which is about
23 percent less than the cost per company under the
reformulation-only scenario. The average cost ratio under the
6-28
-------
best-response scenario is 2.0 percent. Considering that small
companies may be even more likely to select the fee than the
survey population at large, the cost reductions may be even
greater than those estimated in Table 6-6.
The results presented in Tables 6-5 and 6-6 together
indicate that, while the average impact on small companies is
expected to be larger than the average impact on all
producers, the alternative strategies to reformulation,
particularly the fee option, can reduce the small company
impacts substantially.
6.2.7.2 Tonnage Exemption. As an alternative to the fee
options of reformulation, fee, or withdrawal, the EPA will
allow a phased tonnage exemption for architectural producers.
Affected firms will be allowed to exempt a total of 23 Mg of
VOC emissions from control responsibilities through
December 31, 2000, 18 Mg in 2001, and 9 Mg in 2002 and beyond.
These tonnage exemption levels differ from the fee in two
ways. First, the exempt emissions can be applied across all
noncompliant products a firm produces, whereas the fee is
assessed individually for each noncompliant product for which
the fee is selected. Second, the exempt emissions that are
granted are the total emissions of the product rather than
just those in excess of the content limit. Thus, a firm must
coordinate the VOC levels and reguirements of all facilities
and products to determine which ones will be produced under
the tonnage exemption.
The tonnage exemption allows some low-volume products
relief from reformulation costs that can be difficult to
recover from the small amount of revenue generated by a
low-volume product. Both the exceedance fee alternative and
the tonnage exemption are compliance options aimed at
addressing the potential issue of "niche markets" in which
low-volume products exist for which it may not be
6-29
-------
cost-effective for either the manufacturer or resin supplier
to develop a lower VOC formulation.
The EPA lacks data to directly evaluate the economic
impact of the tonnage exemption. It is likely, however, that
many of the products covered under the tonnage exemption might
otherwise be subject to the exceedance fee because both
provisions are most applicable to the smallest volume
products1. Therefore, the tonnage exemption provision is not
likely to further curtail emissions reductions much beyond
what is curtailed by the fee option. However, to the extent
that it supplants the fee as a firm's compliance option, it
will reduce the financial impact of the regulation on that
firm. For example, if 9 Mg of VOCs exempted from regulation
represents 3.6 Mg of exceedance (assuming an exceedance rate
on over-limit products of 40 percent), then the firm subject
to the tonnage exemption can forego 3.6 Mg worth of fee
payments which, at $2,200 per MG (in 1991 dollars), translates
to an impact reduction of $7,920 per firm. If this is applied
to the roughly 500 firms in the architectural coatings
industry, the maximum potential reduction in aggregate
producer impacts is estimated to be about $4 million.
However, it cannot be directly determined whether each firm
would be able to take advantage of the tonnage exemption and
incur these savings. One should also note that, while these
represent potential savings to producers, these are offset by
reductions in fee receipts by the government sector. Thus, to
the extent that the tonnage exemption merely substitutes for
the fee, the substitution has not affected the net social cost
of the regulation.
1 EPA recognizes that a few products on the margin that would be
reformulated if the fee was the only alternative option, may now use a
combination of the tonnage exemption and fee if it is determined to be the
firm's least-cost compliance option. To the extent that this will occur,
there will be a minimal effect on additional foregone emission reductions
when the exemption is considered as a compliance strategy.
6-30
-------
The tonnage exemption may also serve in lieu of small
product withdrawals. In this case, the tonnage exemption
would curtail some emission reductions. However, given the
relatively few products projected for withdrawal and the small
volumes involved, the effect on VOC emissions would likely be
small.
While seeking ways to mitigate the impacts of the
regulation for small manufacturers, the EPA recognizes that
the two different approaches discussed here, the fee option
and small product tonnage exemption, have different
implications for the marginal incentives for VOC reductions.
Although the fee option continues to provide incentive to
reformulate the small niche products because marginal
reductions in VOC content will reduce the per- unit fee paid,
a tonnage exemption would provide no such incentive.
6.3 REGULATORY ALTERNATIVES TO REDUCE IMPACTS
The Agency has engaged in extensive dialogue with both
large and small businesses over the 8-year period of
development of the final rule. The Agency has sought input
from small businesses through a regulatory negotiation,
meetings between EPA and small businesses, and SBA review of
the proposal. Based on this involvement, the EPA incorporated
many of the suggested changes and designed the proposed rule
to address concerns about potential impacts on small
businesses. Specifically, coating categories and VOC content
limits were selected to account for niche products in which
smaller manufacturers have a disproportionate presence. In
addition, to evaluate whether further steps were still needed
to accommodate niche market coatings, the Agency requested
that commenters identify any additional specialty coatings
that could not comply with the proposed VOC content
6-31
-------
requirements. The Agency also requested comment on whether to
include several other compliance options to provide
flexibility and reduce the burden for small businesses. This
section presents a summary of significant issues raised by
public comment on those compliance options and the Agency's
consideration of those compliance options as well as other
provisions in the rule to mitigate rule impacts on small
businesses and preservation of niche markets. The response to
comments document entitled "National Volatile Organic Compound
Emission Standards for Architectural Coatings—Background for
Promulgated Standards," EPA-453/R-95-009b, contains more
detailed summaries of the comments and the EPA's response.
The EPA considered the following compliance options and
other measures to mitigate impacts of the rule on small
businesses:
• selection of VOC content limits and coating
categories;
• low-volume exemption option;
• exceedance fee compliance option;
• extended compliance time for small businesses;
• compliance variance for cases where compliance would
result in economic hardship; and
• selection of recordkeeping and reporting
requirements.
Based on review of comments and further analysis of the
effects of the rule, the EPA has elected to incorporate a
number of the above compliance options and other measures into
the final rule to avoid unnecessary impacts on small
businesses. This section presents the results of the EPA's
final regulatory flexibility analysis, which evaluates the
alternative measures considered to mitigate the impacts of the
rule on small businesses. This discussion incorporates the
6-32
-------
results of the economic impact analysis presented earlier in
this section as well as the Agency's policy considerations and
other information used in selecting the compliance options and
other measures to mitigate the impacts of the rule on small
businesses.
6.3.1 Selection of VQC Content Limits and Coating Categories
In developing the proposed rule, the EPA recognized that
it may not be economical for some manufacturers to reformulate
certain lower-volume products. Rather than exempting these
lower-volume products, the EPA proposed the VOC content limits
in the upper range of VOC content limits in existing state
rules for these categories. For categories for which no state
standards exist, the EPA included the categories in the
architectural coating rule based on discussions with industry
representatives and end-user groups, petitions from
stakeholders prior to proposal, and public comments from
companies providing support for inclusion of the categories
and a suggested VOC content limit. In discussion of the
proposed low-volume exemption, the EPA also requested that
commenters submit detailed information on any specialty
coatings that would not comply with the proposed VOC content
limits and that cannot be cost-effectively reformulated. The
proposal indicated that the EPA would consider whether to
develop additional categories for newly identified niche
categories or to provide a categorical exemption for the
specialty coating.76'77
As a result of information submitted by commenters, the
Agency has added seven new categories to the final rule to
address specific groups of specialty coatings that were
identified through public comment. Also, based on new
information the VOC content limits were increased in the final
rule for four categories. Available information indicates
that the final rule includes VOC content limits at levels that
6-33
-------
recognize the limited potential for reformulation of specialty
niche products and sets VOC contents at the upper range for
the particular type of product. The EPA established special
categories and limits for niche products and established
higher-than-proposed VOC content limits for niche product
categories where commenters submitted sufficient supporting
information. As a result, the final VOC limits for these
categories are unlikely to require manufacturers to
reformulate many products. The specific changes are
identified in Section 7 of this document.
6.3.2 Low-Volume Exemption Option
The Agency requested comment on the concept of a low-
volume compliance exemption option.78 In the proposal preamble
this exemption was described as a compliance option under
which "any manufacturer or importer may request an exemption
from the VOC levels in table 1 for specialized coating
products that are manufactured or imported in quantities less
than a specified number of gallons per year." The Agency
specifically requested comment on exemptions ranging from
1,000 to 5,000 gallons of product per year. The exemption, as
described in the proposal, could be used by a manufacturer for
multiple products, provided that each product was manufactured
in quantities less than the cutoff level. As described in the
proposal preamble, the manufacturer would be required to
submit a request for the exemption and document that the
product(s) for which the exemption was requested "served a
specialized use which cannot be cost-effectively replaced with
another, lower VOC product." The EPA recognizes that small
businesses who produce products with limited volume will
benefit most from an exemption of this type.
Seventeen commenters supported some form of a low-volume
exemption, and four commenters opposed such an exemption.
Commenters supporting the low-volume exemption suggested
6-34
-------
cutoffs ranging from 100,000 gallons per product down to
1,000 gallons per product. Commenters opposed to the
low-volume exemption argued that it was subject to abuse
because of difficulty in defining what is a "product." These
commenters believed that this compliance option would provide
an incentive for companies to develop purportedly "new"
specialty products to keep selling noncompliant coatings.
Based on the arguments presented by commenters about the
need for some type of exemption for very low-volume specialty
products for which it is not cost-effective for either the
manufacturer or the resin supplier to devote time and
resources to reformulation, an exemption is included in the
final rule to accommodate these types of products. Although
in the proposal preamble, the exemption was described in terms
of a per-product exemption at a level between 1,000 and
5,000 gallons annually, commenters highlighted the potential
problems with this type of provision. Therefore, the final
rule contains a variation on the low-volume exemption approach
described at proposal. Specifically, a VOC tonnage exemption
is provided in the final rule. This approach continues to
accommodate the needs of small businesses, niche markets, and
specialty products, as did the proposed low-volume exemption;
but it more effectively limits the VOC emissions resulting
from the exemption. It is expected that this provision will
provide more benefit to small businesses than large
businesses.
Under the VOC tonnage exemption, each manufacturer can
exempt a total of 23 megagrams (25 tons) of VOC in the period
of time from the compliance date through December 31, 2000;
18 megagrams (20 tons) in the year 2001, and 9 megagrams
(10 tons) for the year 2002 and for each year thereafter.
Since some corporations have multiple companies and/or
divisions, an architectural coatings manufacturer or importer
6-35
-------
is defined in the rule to mean the parent company and not each
individual company, subsidiary, or division. Thus, if a
corporation (parent company) has several subsidiaries or
divisions that manufacture coatings, only one exemption per
parent company will be allowed annually. This provision is
structured in this manner to avoid sacrificing VOC emission
reductions and to be equitable to manufacturers. For the
purposes of the tonnage exemption, the manufacturer or
importer calculates VOC tonnage by multiplying the total sales
volume in liters by the "in the can" VOC content of the
coating in grams per liter of coating including any water or
exempt compounds. The "in the can" VOC content must include
consideration of the maximum thinning recommended by the
manufacturer. In the following examples, g/L (or Ib/gal) is
an abbreviation for grams (or pounds) of VOC per liter (or
gallon) of coating, including water and exempt compounds at
the manufacturer's maximum recommendation for thinning. For
example, under this exemption in the second year a
manufacturer could exempt 38,300 liters (8,000 gallons) of a
600 g/L (5 Ib/gal) coating.
51bs/gallon * 8,000 gallons = 40,000 Ibs or 20 tons
Alternatively, a manufacturer could exempt 18,939 liters
(4,000 gallons) of an 800 g/L (6.67 Ib/gal) coating plus
13,731 liters (3,625 gallons) of a 550 g/L (4.58 Ib/gal)
coating.
[(6.67 Ibs/gal * 4,000)+(4.58 Ibs/gal * 2,900)] =
40, 000 Ibs or 20 tons
A manufacturer can exempt any combination of coatings and
volumes as long as the total emissions from these products do
not exceed 23 Mg (25 tons) from the compliance date through
6-36
-------
December 31, 2000; 18 Mg (20 tons) in the year 2001; and 9 Mg
(10 tons) in the year 2002 and each year thereafter.
The tonnage limits would exempt no more than 1.5 to
2 percent of the total expected emission reductions from
architectural coatings in the first year the standard is in
effect. The 9 Mg (10 ton) per-year exemption that goes into
effect in the year 2002 will provide adequate flexibility for
future needs, while effectively limiting emissions due to the
exemption. For firms with VOC content around 600 g/1 (5
Ib/gal), the exemption could apply to 4,000 gallons total
across all of the firm's products. .As is demonstrated in the
calculation of potential cost savings, the exemption can
provide significant relief to small firms or niche market
products by reducing prossible fee payments. However, since
it applies to all products of a firms, it is substantially
lower than the 1,000 to 5,000 gallon per product exemption
considered at proposal.
This exemption differs from the low-volume exemption in
the proposal preamble in the following ways:
(1) The EPA changed the exemption from a per-product
basis to a per-manufacturer basis. This was done to
avoid the difficulty of defining a "product" and to
avoid the related potential for abuse by
manufacturers in designating products for exemption.
(2) The EPA changed the exemption level from gallons of
coating to tons of VOC. This change was made for
two primary reasons. First, it provides an
incentive for manufacturers to reduce the VOC
content of the coatings for which they claim this
exemption. For example, with a 5,000 gallon
exemption, the manufacturer could exempt 5,000
gallons whether the product was 850 g/L or 200 g/L.
With a tonnage exemption, however, the VOC content
in each can of coating counts toward the allotted
exemption. Therefore, if the manufacturer reduces
the VOC content of the coating it wishes to exempt,
more gallons of that coating could be sold under the
exemption. Second, the choice of VOC tonnage
instead of gallons of coating for the exemption
6-37
-------
alters the exemption from an unknown loss of
emission reductions to a cap on tons exempted per
manufacturer. Therefore, this change serves to
place an upper bound on the emission reductions that
are lost through this exemption, which allows the
Agency to better estimate its anticipated impact.
(3) The exemption is reduced over time. The ratcheting
down of the tonnage exemption from 23 Mg (25 tons),
to 18 Mg (20 tons), and then to 9 Mg (10 tons)
provides a strong incentive to manufacturers using
the exemption to continue to seek ways to reduce the
VOC content of their coatings. This exemption is
intended to provide additional time for
manufacturers to reformulate coatings, and provide
some relief in the long run for small volume
producers.
6.3.3 Exceedance Fee Compliance Option
The EPA requested comment on whether to include an
exceedance fee option for use as a compliance alternative to
meeting the VOC content limits in the proposed rule.79 This
option was designed to provide compliance flexibility and set
the fee rate high enough to provide an economic incentive for
reformulation. The proposed fee rate was $0.0028 per gram
($2,500 per ton) of VOC in excess of the applicable VOC
content limit multiplied by the amount of coating produced.
The EPA also requested comment on the appropriateness of the
proposed fee rate and the recordkeeping and reporting
requirements associated with the exceedance fee compliance
option.
Public comment on the concept of this option varied
widely. Some commenters, including small businesses and
national coating manufacturers trade associations, were
supportive of the concept because it provided compliance
flexibility. Some of these commenters supported the concept
under the condition that the option would not be accompanied
by burdensome recordkeeping requirements. Other groups of
commenters opposed inclusion of this option because they
6-38
-------
thought that it could disrupt the market (increase prices),
that it would be difficult to enforce, or that it was
unnecessary because the proposed limits were not hard to
achieve. For a more complete description of the comments on
this option, see Section 2.4.1 of the Architectural Coating
Regulation BID.
Careful evaluation of all of the comments and discussions
with the SBA led the Agency to include the exceedance fee
option in the final rule. Under this approach, manufacturers
and importers have the option of paying a fee, based on the
extent to which VOC content limits are exceeded, instead of
achieving the VOC content limits in the rule. The fee is
calculated at a rate of $0.0028 per gram ($2,500 per ton), in
1996 dollars, of VOC in excess of the applicable VOC content
limit, multiplied by the volume of coating produced. This
option is included in the rule for several reasons. The
exceedance fee option will provide transition time for those
manufacturers that need additional time to obtain lower-VOC
technologies. The exceedance fee option provides long-term
flexibility and a less costly compliance option than
reformulation for both small and large manufacturers selling
very low-volume specialty coatings where the cost of
reformulation may be prohibitive compared to the potential
profit, thus enabling manufacturers to continue to make these
products available to consumers. The exceedance fee option is
significantly less burdensome for manufacturers than the
proposed compliance variance provision, which has not been
retained in the final rule. However, contrary to some
comments received, costs resulting from the exceedance fees
will likely generally motivate manufacturers over time to
develop high performance products with low-VOC content.
Some commenters believed that the exceedance fee will
disrupt the marketplace, shifting business among companies.
6-39
-------
However, since the fee will probably be used primarily for the
manufacture of low-volume specialty coatings, which are driven
by demand from consumers, it is not likely that the demand
from these markets would be significant enough to provide any
incentive for manufacturers to shift to these products. The
impacts to the market are lower with the fee than they would
be if reformulation was the only option available for
producers, because the fee reduces the number of potential
product withdrawals and reduces the net social cost. Raising
the VOC content limits, as suggested by some of the
commenters, in lieu of offering the fee could significantly
undermine the emissions reduction objectives of the rule. The
fee provides some flexibility to producers of low-volume
products, or products that are only slightly above the VOC
content limit of the standard, who may find it prohibitive to
incur the largely fixed cost of reformulation. Because
products for which manufacturers will choose to pay the fee
would tend to represent a small portion of the national VOC
emissions from architectural coatings, the fee option itself
would not significantly undermine emission reduction
objectives. However, raising the VOC content limits in the
rule to accommodate all low-volume products would negate the
VOC emission reductions from all these products. The fee also
provides continued incentive for producers to reduce VOC
content until they achieve the VOC content limits in the rule.
With regard to concerns about enforcement of the
exceedance fee, the recordkeeping and reporting requirements
are designed to ensure compliance with this option. Any
violations of the recordkeeping and reporting or any other
requirements could result in enforcement actions and the
possibility of penalties.
The estimated cost for reporting and recordkeeping of the
fee provision at a small company using the exceedance fee
6-40
-------
provisions for eight products is approximately $5,000 per year
(see Table 6-5). This cost represents the cost to maintain
the records of the VOC content and the total volume
manufactured or imported for which the exceedance fee option
is used as well as the preparation of the annual report for
payment of the exceedance fee. Assuming $5 million of sales
revenue as a midpoint estimate for small companies in the $0
to 10 million range, fee recordkeeping costs would be
approximately 0.1 percent of sales revenue, which is not a
significant burden.
Price increases on fee-paying products will cause some
consumer substitution to nonfee-paying (lower-VOC) products.
For some products, it may not be profitable to reformulate or
pay the fee, so firms may consider withdrawing the product
from the market. These phenomena are explicitly modeled
elsewhere in this document. However, the premise of the fee
is that it internalizes the (public) environmental cost of VOC
emissions into the private cost of the good. Therefore, if
some consumers substitute away from the now higher-priced
fee-paying product, it reflects the fact that they are not
willing to pay the "full" cost of consuming the higher—VOC
products. This is the fundamental purpose of market-based
incentives for environmental protection.
6.3.3.1 Exceedance Fee Rate. Several commenters also
submitted comments on the proposed exceedance fee rate of
$0.0028 per gram of VOC in excess of the applicable VOC
content limit. Some of these commenters thought that the fee
rate was too low to encourage development of compliant
coatings. Other commenters thought that it was too high
relative to the price of some products or in light of the
additional costs associated with recordkeeping for this
option. One commenter suggested a phase-in of the fee. For a
6-41
-------
more complete description of the comments on this option, see
Section 2.4.2 of the Architectural Coatings Regulation BID.
Several factors affected the selection of fee level,
including the benefit per ton of VOC reductions value
historically used in analyses under the Clean Air Act, the
historical range of acceptable cost-effectiveness values for
VOC, the magnitude of the loss in emission reductions, and the
effect on the market model (price and output adjustments,
distribution of welfare impacts across consumers and
producers, and changes in social cost) as well as the effect
of different exceedance fee rates on the industry
cost-to-revenues ratio.
More specifically, the value chosen for analysis at
proposal is slightly higher than the benefit transfer value
(i.e., the benefit value per ton of VOC reduced) historically
used in EPA analyses and is also slightly higher than
historical cost-effectiveness values for VOCs. This was
intended to provide incentive for manufacturers to continue to
strive to find low-cost methods of reducing the VOC content in
their products. Therefore, manufacturers that find the fee to
be the lowest-cost option of compliance with the regulation
(in comparison to reformulation or losing profits from product
withdrawal) would pay the fee, but be encouraged to find an
even lower-cost solution to reduce total production costs in
the long run.
Another consideration was the amount of emission
reductions lost at the selected fee level. This level also
proved to provide only minor adjustments in market price and
quantity in comparison to reformulation by itself, while
providing substantial flexibility to manufacturers of
small-volume products or products that exceed the standards by
a small amount. The Agency also evaluated a higher fee rate
prior to proposal and found that social cost increased with a
6-42
-------
relatively small change in lost emission reductions (as
compared to the lower fee rate). The selected fee rate was
thus set high enough to make reformulation attractive for the
majority of producers, but low enough to allow a small sector
of products to remain on the market in lieu of withdrawal.
Also, the lost emission reductions will be limited and the
impact on the markets will be minor. The Agency also examined
the effect of varying the fee rate on the fee adoption rates,
social cost impacts, foregone emission reduction, and small
business impacts. This analysis showed that at lower fee
rates (e.g., $l,500/ton and $l,000/ton) there was a
significant increase in the amount of foregone emission
reductions and only a small decrease in the average
cost-to-revenues ratio for small businesses.80
Based on the economic analysis, the EPA believes that the
fee is set at an appropriate level. The economic model
compares the cost of paying the fee to the cost of
reformulation for surveyed products. While many products are
projected to opt for the fee, these products are uniformly
small in volume; thus, their contribution to total market
output (and emission reduction) is relatively small. It
generally would not be advantageous for producers of
large-volume products, which generate a disproportionately
large share of emissions, to opt for the fee over
reformulation. Furthermore, the existence of the fee provides
continued incentive for fee-paying firms to reduce VOC
contents on the margin, because this will reduce the amount of
fee they must pay.
Some commenters suggested that the EPA should base the
fee on price, rather than the quantity of VOC emitted by the
product. The premise is that only a large proportional price
effect will induce large changes in behavior. The objective
of a pollution fee, however, is to "charge" for the pollution
6-43
-------
generated. The only consistent way to accomplish this is to
have the fee payment depend on the amount of pollution
generated. It is not clear how a price-based fee would be
tied to the amount of VOC emitted. For instance, a low-priced
high-VOC product could have a fee per unit that is much lower
than a high-priced lower-VOC product. In this case, the fee
mechanism would not work to ensure enough incentive for the
higher-VOC product to reduce VOC content. In other words, a
ton of extra emissions from one product would incur less of a
fee than a ton of extra emissions from the other. For
example, such a mechanism would favor very high-VOC content
products that are very inexpensive. Alternatively, having one
ton of exceeded emissions face the same fee, regardless of
source is more efficient, and seemingly more fair.
The combination of the compliance options in the final
rule provides the phase-in of the fee suggested by some
commenters. Specifically, the phasing of the tonnage
exemption in combination with the exceedance fee provision
will operate to increase the fee for products that exceed the
VOC content limits in the rule. In the time period from the
compliance deadline through the year 2000, manufacturers may
exempt from regulation 25 tons (23 Mg) of VOC, so total fee
payments would be lower than in the second year. The
following year, 2001, has a lower exemption level of 12 tons
(11 Mg) of VOC, so fee payments would be slightly greater for
those manufacturers who choose not to reformulate or otherwise
reduce the VOC content of their products. In the next year
and any subsequent year of compliance, the fee rate would
become level because the exemption level remains the same at
5 tons (4.5 Mg) per year. The fee payments would also provide
incentive for manufacturers to find lower-cost VOC technology
to meet the standard and eliminate or reduce their fee
payments.
6-44
-------
6.3.4 Extended Compliance Time for Small Businesses
At proposal the Agency requested comment on whether the
final rule should include a compliance extension for small
businesses.81 In effect, this extension would have allowed
small businesses 12 additional months to comply. Thirteen
commenters commented on the small business compliance
extension concept. Two-thirds of the commenters providing
comments on this provision were against special treatment for
small businesses. The primary concern was that this provision
would provide small businesses an unfair advantage in the
marketplace. Some of the commenters opposing the extension
noted that an extension should not be necessary because of the
specialized coating categories and the VOC content limits for
these categories, small volume exemption, the potential
exceedance fee compliance option, and the variance provision.
After careful evaluation of the comments, the Agency has
decided not to include a compliance extension specific to
small businesses but has instead lengthened the compliance
period for all regulated entities to 12 months. This time
period was selected to balance the needs of the regulated
entities, both large and small businesses, against the need
for rapid implementation of the rule to achieve the required
reductions of VOC emissions.
6.3.5 Compliance Variances
In the proposal preamble the Agency requested comments
from small businesses on their expected use of a compliance
variance provision.82 The proposed compliance variance
provision would have allowed manufacturers and importers of
architectural coatings to submit a written application to the
Administrator requesting a variance if, for reasons beyond
their reasonable control, they could not comply with the
requirements of the proposed rule. In particular, the
proposed variance provision allowed additional compliance time
6-45
-------
and was developed especially for small businesses, but would
have been available to any size business.
Of the 22 commenters on this provision, only
eight commenters supported the concept. The 14 commenters
opposing the concept included some small businesses. Concerns
expressed by those commenters included concerns that it would
impose such a heavy burden that businesses would choose to
shut down rather than use the variance and that the variance
requirements as proposed are unduly difficult to achieve. For
example, one commenter noted that the variance provision as
proposed required significant expense with little or no
guarantee of approval. The commenter recommended an extended
compliance period as a more effective option to alleviate the
heavy burden upon small businesses.
Based on the comments received, the Agency concluded that
the variance provision may not provide the intended additional
compliance flexibility, especially for small businesses.
Therefore, the variance provision has not been included in the
final rule. Even though the proposed variance requirements
were intended to be the minimum necessary to approve a coating
variance, the requirements may have been burdensome,
particularly for small businesses with limited or no
regulatory compliance staff. It is also possible that the
variance provision could create an uneven playing field
because small businesses would not have the resources needed
to pursue this option, thereby putting small businesses at a
disadvantage compared to large businesses. Also, as one
commenter pointed out, even with the investment of time and
money, the Agency cannot guarantee approval of the variance
application. In addition, review and approval of numerous
variance applications would place a heavy burden on EPA's
staff, thereby delaying implementation of the intended
flexibility to the disadvantage of regulated entities.
6-46
-------
Nevertheless, there is still value in providing
additional compliance flexibility; therefore, new provisions
have been incorporated into the final rule (i.e., the tonnage
exemption that phases down over time and the exceedance fee
option). These provisions provide even greater flexibility
than the variance provision and are less burdensome. Both of
these compliance options are automatically available to all
regulated entities and, thus, do not involve complex
application and approval processes. However, these compliance
options do require some minimal recordkeeping and reporting.
The tonnage exemption will allow each regulated entity to
exempt from the VOC content limit anywhere from 7,000 to
30,000 gallons of coatings the first 15 months; 3,400 to
14,400 gallons the second year; and 1,400 to 6,000 gallons the
third year and beyond (the actual amount exempted depends on
the VOC content of the product(s)). Therefore, this exemption
is ideal for low-volume products that cannot be reformulated
in the foreseeable future.
The exceedance fee option is designed to give
manufacturers additional time to develop lower-VOC
technologies, if necessary. This option allows regulated
entities to continue to sell coatings that exceed the VOC
content limits in addition to the coatings for which they are
claiming the low-volume exemption, provided they pay an
exceedance fee. The amount of the fee is based on the volume
of the product sold, the VOC content of the product, the VOC
content applicable to the product, and the fee rate.
In addition to these provisions, the compliance time,
which concerned some commenters, has been extended to
12 months, and the EPA added seven new specialty coatings
categories (e.g., zone markings, concrete curing and sealing,
conversion varnishes) to the final rule and increased the VOC
content limits for four coating categories.
6-47
-------
6.3.6 Selection of Recordkeeping and Reporting Requirements
The EPA also selected the recordkeeping and reporting
requirements of the rule, taking into consideration the
impacts of the rule on small businesses. The EPA designed the
proposed rule to require only those recordkeeping and
reporting requirements necessary to allow determination of
compliance and enforcement, if necessary. The proposed rule
required an initial report and labeling of containers for
manufacturers who choose to demonstrate compliance by meeting
the VOC content limits in the standard. There were no
additional reports or records required from these
manufacturers. Additional recordkeeping and reporting
requirements were proposed for the recycled coatings option,
the exceedance fee option, and the low-volume exemption
option.
Two industry commenters requested even more limited
recordkeeping and reporting requirements in the rule and
several industry commenters noted the need to correct dates
and clarify some of the labeling requirements in the proposed
rule. In the final rule, the EPA has maintained the proposed
recordkeeping and reporting requirements for manufacturers who
choose to demonstrate compliance by meeting the VOC content
limit in the standard. The EPA has also clarified the
container labeling requirements and provided additional
flexibility for labeling of VOC content of the coating as well
as for placement of the date codes. In the final rule, the
EPA required only those records and information necessary to
determine compliance with the compliance alternatives of the
exceedance fee, the tonnage exemption, and the credit for
recycling of coatings. Specifically, the final rule only
requires semiannual reporting from manufacturers who elect to
use the exceedance fee compliance option and annual reporting
from manufacturers who elect to use the tonnage exemption or
6-48
-------
the recycled coatings provision. These records and reports
are essential for enforcing these provisions and the EPA
believes that these records and reports do not represent an
undue burden on manufacturers or importers who elect to use
these optional compliance provisions. For example, as noted
earlier, the estimated cost for reporting and recordkeeping of
the exceedance fee provision at a company with an average of
eight reformulations would be approximately 0.1 percent of
sales revenue, which is not a significant burden.
6.4 SMALL BUSINESS IMPACT SUMMARY
The potential for significant impacts on small businesses
of the regulation arise from two primary sources:
• Products made by small producers, on average, have a
higher VOC content than the industry average.
• The costs of reformulating products to comply with the
regulation are independent of product volume and
thereby impose higher average costs per unit of product
on small volume coatings.
The first problem is related to small producers' tendency
to specialize in coatings categories that are naturally higher
in VOC content and to their tendency to concentrate in the
"high-VOC" end of the distribution of products within a given
category. Thus the potential for disproportionate impacts of
VOC reduction regulation on small businesses follows partly
from the fact that small businesses contribute a
disproportionate amount of the aggregate VOC emissions that
are targeted for reduction.
The second problem follows from the nature of
reformulation costs. A coating's formula is the product of an
intellectual capital investment, much like the development of
6-49
-------
a drug or a computer software product. The cost of the
investment follows directly from the level of effort necessary
to revise the formula to meet both the VOC standards imposed
by the regulation and performance standards imposed by the
marketplace. This level of effort is essentially independent
of the quantity of the product that is eventually sold.
Therefore, the relative impacts on smaller volume products is,
by definition, greater.
The data used in this analysis suggest that these two
primary factors are relevant in the case of small
architectural coatings producers. The average VOC content of
the products made by the small business producers in the
survey is 75 percent higher than the average VOC content of
all products combined. A little over half of the difference
in the averages is attributed to the specialization of small
producers in high-VOC content product categories, with the
remainder attributed to the tendency for small businesses to
produce higher VOC products within each product group.
Moreover, the average product volume of products made by small
businesses is less than 20 percent of the average product
volume for the entire survey population, implying much larger
average reformulation costs. Thus, without mitigating
factors, the impacts on small businesses are potentially
significant.
The regulation has been designed to mitigate small
business impacts. Despite their inherently higher VOC
content, the proportion of small business products exceeding
the regulatory standards is not much higher than the
corresponding proportion for the survey population at large
(42 percent vs. 36 percent). In addition, the availability of
the exceedance fee option is beneficial to small business
producers because it places an upper bound on the per-unit
costs of compliance. Data analyzed in this study indicate
6-50
-------
that small business producer costs are reduced by nearly
one-quarter when the exceedance fee is introduced and the
possibility of product withdrawal is considered in lieu of
reformulation. The cost/revenue ratio exemplifies the
advantages of the lower-cost compliance options (the fee and
withdrawal) in that the ratio for small businesses drops from
2.5 percent to 2.0 percent.
In addition to adding the exceedance fee and the tonnage
exemption to the final rule, the EPA also increased the
compliance time to 12 months and added seven new product
categories and increased the VOC content limits for four
categories. These changes were made in response to public
comments to further mitigate the rule's small business
impacts. The analysis of the impacts of the final rule shows
that these provisions are likely to be used by small entities
and the impact on a typical small firm is reduced without
significant reduction in the emission reductions achieved by
the rule. The EPA believes that these measures adopted in the
final rule represent a significant mitigation of the economic
impacts on small businesses compared to the impacts that might
otherwise have occurred.
6-51
-------
SECTION 7
EPILOGUE
Because regulatory development is an evolving process,
the final Table of Standards for VOC content limits differs
slightly from the interim Table of Standards used in the
analysis reported here. The main difference between the two
sets of standards (see Table 7-1) is the addition of new
categories in the final standards and the revision of content
limits for other categories. These two dimensions of change
are evaluated in turn below.
7.1 NEW PRODUCT CATEGORIES
The final standards added seven product categories not
included in the interim standards. These are:
• calcimine recoaters
• concrete curing and sealing compounds
• concrete surface retarders
• conversion varnish
• faux finish/glazing
• stain controllers
• zone marking coatings
7-1
-------
TABLE 7-1. TABLE OF STANDARDS: INTERIM VS. FINAL
Architectural Coatings
Category
VOC Content Limit
(g/L)
Interim
(see Table 2-1) Final
Difference
Antenna coatings
Anti fouling coatings
Antigraffiti coatings
Bituminous coatings and mastics
Bond breakers
Calcimine recoater
Chalkboard resurfacers
Concrete curing compounds
Concrete curing and sealing
compounds
Concrete protective coatings
Concrete surface retarders
Conversion varnish
Dry fog coatings
Extreme high-durability
coatings
Faux finishing/glazing
Fire-retardant /resistive
coatings
Clear
Opaque
Flat coatings, N.O.S.
Exterior
Interior
Floor coatings
Flow coatings
Form release compounds
Graphic arts coatings (sign
paints)
Heat reactive coatings
High- temperature coatings
Impacted immersion coatings
Industrial maintenance coatings
500
450
600
500
600
NA
450
350
NA
400
NA
NA
400
800
NA
850
450
250
250
400
650
450
500
420
650
780
450
530
450
600
500
600
475
450
350
700
400
780
725
400
800
700
850
450
250
250
400
650
450
500
420
650
780
450
Limit increased
New category
New category
New category
New category
New category
(continued)
7-2
-------
TABLE 7-1. TABLE OF STANDARDS: INTERIM VS. FINAL (CONTINUED)
VOC Content Limit
(g/D
Architectural Coatings
Category
Lacquers (including lacquer
sanding sealers)
Magnesite cement coatings
Mastic texture coatings
Metallic pigmented coatings
Multicolor coatings
Nonferrous ornamental metal
lacquers
Nonflat coatings, N.O.S.
Exterior
Interior
Nuclear coatings
Pretreatment wash primers
Primers and undercoaters,
N.O.S.
Quick dry coatings
Enamels
Primers, sealers, and
undercoaters
Repair and maintenance
thermoplastic coatings
Roof coatings
Rust preventive coatings
Sanding sealers
Sealers
Shellacs
Clear
Opaque
Stains
Clear and semi transparent
Opaque
Waterborne low solids
Stain controllers
Interim
(see Table 2-1)
680
600
300
500
580
870
380
380
450
780
350
450
450
650
250
400
550
400
650
550
550
350
120
NA
Final Difference
680
600
300
500
580
870
380
380
450
780
350
450
450
650
250
400
550
400
730 Limit increased
550
550
350
120
720 New category
(continued)
7-3
-------
TABLE 7-1. TABLE OF STANDARDS: INTERIM VS. FINAL (CONTINUED)
Architectural Coatings
Category
VOC Content Limit
(g/L)
Interim
(see Table 2-1) Final
Difference
Swimming pool coatings
Thermoplastic rubber coatings
and mastics
Traffic marking paints
Varnishes
Waterproofing sealers and
treatments
600
550
150
450
600
550
150
450
Clear
Opaque
Wood preservatives
Below ground
Clear and semi transparent
Opaque
Low solids
Zone marking coatings
Total New Categories
Total Limit Changes
600
400
550
550
350
NA
NA
7
4
600
600
550
550
350
120
450
Limit increased
New category
By and large, new categories were added to accommodate
specialty products that were previously included in other
categories with lower (more stringent) VOC limits. As a
result, some products that would be over the limit in the
previous category, thereby necessitating a compliance action
(reformulate, fee payment, withdrawal), are no longer
constrained by the regulation. In these cases, the addition
of the new categories reduces the number of required
compliance actions, as a result, also cuts compliance costs
and the quantity of emission reductions.
However, one of the new product categories, concrete
curing and sealing (CCS) compounds, applies to products that
were considered outside of the regulated universe in the
economic analysis presented in this report. Therefore, the
7-4
-------
compliance actions required for those products are not
estimated in this analysis. The potential cost implications
of that omission are discussed below.
Supplemental data could be obtained for only two of the
seven new product categories. These data were gathered after
proposal and are used here to estimate the likely impact of
these new categories on regulatory costs.
One of the categories for which supplemental data were
obtained is the zone markings category. First, we note that
46 products from the original survey data in the traffic
paints category have VOC contents that are greater than
150 g/L (the final traffic marking paints content limit) and
450 g/L (the zone markings limit). These 46 products
constitute the entire list of surveyed products that could
potentially be relieved from compliance by the addition of the
higher zone markings limit. According to data from the state
of Texas, zone markings constitute approximately 9 percent of
all traffic coatings.83 We use this percentage to estimate the
number of those 46 products that are zone markings, yielding
an estimate of 4.1 (decimals are used to reflect an averaging
effect). Using an expansion factor of 3.0 to reflect the
scale of the national estimate of traffic coatings to the
survey estimate, we estimate that 12.3 products nationwide can
avoid compliance action due to the addition of the new zone
markings category.
Data were gathered for 77 CCS products with a total
product volume of 11.2 million liters.84 Of these 77 products,
38 were determined to exceed the content limit of 700 g/L. As
described in Section 2 of this report, the number of
noneompliant coatings is reduced by a factor of one-third to
estimate the total number of noncompliant coatings needing a
compliance action (reformulation, fee, or withdrawal). After
this adjustment, 25 of the 77 CCS products surveyed are
7-5
-------
estimated to require compliance action. The CCS data also
indicate an estimate of 37.8 million liters of CCS products
nationwide. Taking the ratio of national CCS volume to the
volume captured in the supplemental data collection
(37.8/11.2) and multiplying by the 25 surveyed products
needing compliance action yields a national estimate of CCS
compliance actions of 85.6 products.
Taking the 85.6 additional compliance actions due to the
new CCS category together with the 12.3 fewer compliance
actions due to the zone markings category yields a net
increase of 73.3 compliance actions. To approximate the
social cost implications, we take the ratio of the total
social costs from the architectural coatings market analysis
($20.2 million in Table 3-2) and divide by the total number of
compliance actions in the analysis (2,345 products in
Table 2-2) to get a social cost per compliance action of
approximately $8,600. Multiplying this number by
74 compliance actions gives a social cost estimate of
approximately $632,000 ($1991).
7.2 CATEGORIES WITH HIGHER VOC CONTENT LIMITS
Besides the additional categories, VOC content limits
were higher (less stringent) in the final standards than in
the interim standards for the following categories:
• antenna coatings
• shellacs, clear
• waterproofing sealers and treatments, opaque
The survey data indicate that nine products in these
three product categories would have been noncompliant under
the interim standards but are compliant under the final
7-6
-------
standards. Reducing the nine otherwise noncompliant products
by one-third yields an estimate of six compliance actions
within the survey population that are avoided by the higher
content limit in the final standards. Because the volume of
surveyed products in these categories roughly equals the
national sales estimates, the estimated number of avoided
compliance actions nationwide is also six. Multiplying this
number by the social cost estimate of $8,600 yields an
estimate for the reduction of social costs caused by the new
content limits of approximately $52,000. Subtracting this
from the net cost increase quantified for the new product
categories reduces the cost estimate to about $580,000
($1991) .
7.3 SUMMARY
The VOC content standards included in the final rule
differ from the limits analyzed in this report. The difference
between the two sets of standards are the inclusion of seven
new product categories and an increase in the content limits
(reduction in stringency) for three product categories.
Because of data limitations, only a subset of these
changes lend themselves to quantification of potential costs
impacts. The net quantified effect is a $580,000 increase in
the estimate of annual social costs. However, this increase
in cost must be considered against the unquantified decrease
in costs from the expected fall in compliance due to the five
other new categories. Without additional data, it is
difficult to conclude whether the cost reductions from those
categories will together outweigh the net cost increases
quantified. Given that the social cost effects quantified
here are less than 3 percent of the total estimated social
costs of the regulation, factors that reduce (or reverse the
7-7
-------
sign) of these costs lead to the conclusion that the total
social cost estimate is not greatly affected by the
differences between the interim standards used in the analysis
and the final standards issued in the rule.
7-8
-------
REFERENCES
1. Stigliani, William M. Chemical Emissions from the
Processing and Use of Materials: the Need for an
Integrated Emissions Accounting System. Ecological
Economics 2^(4) :325-341. 1990. (Adapted from Figure 2).
2. Industry Insights. Architectural and Industrial
Maintenance Surface Coatings VOC Emissions Inventory
Survey. Prepared for the National Paint and Coatings
Association in cooperation with the AIM Regulatory
Negotiation Industry Caucus. Final Draft Report. 1993.
3. U.S. Department of Commerce. 1987 Census of
Manufactures, Industry Series: Paints and Allied
Products. Washington, DC, Government Printing Office.
1990.
4. Ref. 2.
5. Ref. 2.
6. National Paint and Coatings Association. U.S. Paint
Industry Data Base. Menlo Park, CA, SRI International.
1992.
7. Whittington, Trevellyan V. Paint Fundamentals. In Paint
Handbook, Guy E. Weismantel (ed.). New York, McGraw-
Hill. Pp. 1-1 to 1-23. 1981.
8. Beno, J., W. Brown, and P.P. Obst. Formulating and Using
Water Based Thermoplastic Resins for Wood Finishing. In
Proceedings of the Nineteenth Water-Borne, Higher-Solids,
and Powder Coatings Symposium. Robson F. Storey and
Shelby F. Thames (eds.). University of Southern
Mississippi, Department of Polymer Science. Pp. 626-638.
1992.
9. Singer, Elias. Raw Materials. In Paint Handbook, Guy E.
Weismantel (ed.). New York, McGraw-Hill. Pp. 3-1 to
3-22. 1981.
R-l
-------
10. Rauch Associates, Inc. The Rauch Guide to the U.S. Paint
Industry. Bridgewater, NJ, Rauch Associates, Inc. 1990.
11. Ref. 9.
12. Kemezis, Paul. Wait-and-See Stance Taken on Zero-VOC
Architectural Paints. Chemical Week. Pp. 52-53.
October 1992.
13. Ref. 8.
14. Ref. 10.
15. Ref. 7.
16. Ref. 7. (Adapted from Figure 1.4).
17. Ref. 10.
18. Bakke, Timothy 0. Clean Air Paints. Popular Science.
23J7:85. August 1990.
19. D'Amico, Esther. Waterborne Systems Gaining Niche By
Niche. Chemical Marketing Reporter. 238(18);SR20-SR28.
1990.
20. Ref. 18.
21. Ref. 18.
22. Ref. 18.
23. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1982. Washington, DC,
Government Printing Office. 1983.
24. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1983. Washington, DC,
Government Printing Office. 1984.
25. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1984. Washington, DC,
Government Printing Office. 1985.
26. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1985. Washington, DC,
Government Printing Office. 1986.
27. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1986. Washington, DC,
Government Printing Office. 1987.
R-2
-------
28. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1987. Washington, DC,
Government Printing Office. 1988.
29. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1988. Washington, DC,
Government Printing Office. 1989.
30. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1989. Washington, DC,
Government Printing Office. 1990.
31. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1990. Washington, DC,
Government Printing Office. 1991.
32. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1991. Washington, DC,
Government Printing Office. 1992.
33. O'Reilly, Richard. Product Development Drives Market
Growth. In: Standard and Poor's Industry Surveys:
Chemicals Current Analysis. Pp. C40-C43. New York,
Standard and Poor's Corporation. 1991.
34. Ref. 33.
35. Ref. 33.
36. Johnson, Duane. The Best Paint for the Job. Family
Handyman. 42^:18. June 1992.
37. Consumer Reports Magazine. Interior Latex Paints,
p. 333. May 1991.
38. National Paint Coatings Association. The Household Paint
Selector. New York, Barnes and Noble Books. 1975.
39. Ref. 3.
40. U.S. Department of Commerce. The 1982 Benchmark Input-
Output Accounts of the United States. Washington, DC,
Government Printing Office. 1991.
41. U.S. Department of Commerce. Census of Manufactures,
Subject Series: Type of Organization. Washington, DC,
Government Printing Office, February 1991.
R-3
-------
42. Source for Percentage Accounted for Data: U.S.
Department of Commerce. 1987 Census of Manufactures,
Concentration Ratios of Manufacturers. Washington, DC,
Government Printing Office. 1992.
43. Telecon. Lindsay, Sam. Small Business Administration,
with Dempsey, Jenny. Research Triangle Institute,
July 6, 1993.
44. U.S. Department of Commerce. 1987 Economic Censuses.
Volume 1, Report Series, Release ID. Census of
Manufactures: Location of Manufacturing Plants. file
MC87LMCO. 1991.
45. Ref. 10.
46. Ref. 3.
47. Loesel, Andrew. Coatings Industry Faces New Mix. In
Chemical Marketing Reporter. 238(18);SR3-SR8. 1990.
48. Ref. 47.
49. Finishers' Management. The U.S. Paint and Coatings
Industry. Pp. 23-25. April 1991.
50. U.S. Department of Commerce. U.S. Industrial Outlook
'92, Business Forecasts for 350 Industries. Washington,
DC, Government Printing Office. 1992.
51. AIM Coatings Regulatory Negotiation Committee meeting.
July 28-30, 1993, Washington, DC. Meeting Summary.
52. Ref. 51.
53. Office of Management and Budget. Guidelines and Discount
Rates for Benefit-Cost Analysis of Federal Programs.
Memorandum for Heads of Executive Departments and
Establishments. Circular No. A-94 Revised (Transmittal
Memo No. 64). October 29, 1992.
54. ICF Consulting Associates, Incorporated. Small Business
Economic Impact Study. Prepared for the South Coast Air
Quality Management District. Final Report. June 17,
1988.
55. Ref. 2.
56. Ref. 2.
R-4
-------
57. Telecon between Chris Sarsony, Eastern Research Group to
Brian Murray, Research Triangle Institute, December 15
1997.
58. U.S. Department of Commerce. 1991 Annual Survey of
Manufactures: Statistics for Industry Groups and
Industries. Washington, DC, Government Printing Office.
1992.
59. Ref. 2.
60. Memorandum. Seagroves, Monica, Eastern Research Group,
to Ducey, Ellen, U.S. Environmental Protection Agency.
January 7, 1998. Architectural coatings certification
and registration information from Massachusetts
Department of Environmental Protection.
61. Ref. 58.
62. Ref. 58.
63. Ref. 58.
64. Eastern Research Group. "Traffic Coating Analysis.*
Prepared for the U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards.
Morrisville, NC: Eastern Research Group. 1998.
65. Ref. 64.
66. Eastern Research Group. "Emission Reduction from the
Final Architectural Coatings VOC Rule.* Prepared for the
U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards. Morrisville, NC:
Eastern Research Group. 1998.
67. Ref. 3.
68. Ref. 2.
69. Ref. 2.
70. Ref. 2.
71. Ref. 2.
72. Memorandum. Nelson, Robert, National Paint & Coatings
Association, to Madariaga, Bruce, EPA/OAQPS. October 14,
1993.
R-5
-------
73. Ref. 51.
74. Ref. 2.
75. Ref. 2.
76. 61 FR 32740.
77. 61 FR 32741.
78. Ref. 77.
79. Ref. 76.
80. Memorandum from Murray, Brian, Research Triangle
Institute, to Evans, Ron, Environmental Protection
Agency. July 1, 1998. AIM SBREFA analysis.
81. 61 FR 32732.
82. 61 FR 32743.
83. Telecon. Seagroves, Monica, Eastern Research Group, with
Turner, Mel, Standard Paints. April 22, 1997. Comment
clarification—zone category.
84. Facsimile. Sarsony, Chris, Eastern Research Group, with
Murray, Brian, Research Triangle Institute. June 29,
1998. Calculation sheet: concrete curing and sealing
compounds.
R-6
-------
APPENDIX A
MARKET DEFINITION,
DEMAND ESTIMATION, AND DATA
-------
A.I PRODUCT/MARKET CROSS-REFERENCE METHOD
Data on coating prices, quantities, average VOC contents,
and VOC content limits are necessary to estimate the effect of
VOC content limits on architectural coatings products. Price
and quantity data were taken from the 1991 Current Industrial
Reports: Paint and Allied Products.1 The Architectural and
Industrial Maintenance Surface Coatings Survey (the survey)2
provided the sales-weighted average VOC emissions, which
represent VOC content. VOC content limits were from the TOS
developed by EPA.
Census data are organized according to product codes,
which define product categories; however, these Census product
categories differ from the product categories in the survey.
Furthermore, the TOS (see Table 2-1) gives VOC content limits
for product categories that differ slightly from those
categories for which data are provided in the survey. Data
from all three sources are necessary to conduct the economic
impact analysis. Therefore, a fourth product categorization
was constructed, which is called market segments, that
aggregates the categories so that data may be used from all
three sources to provide the necessary level of resolution for
market analysis. Table A-l illustrates the individual product
categories represented by each data source and how they map
into the market segments used in the analysis.3'4
The mapping in Table A-l proceeds from the most
aggregated category to the least aggregated category. In some
cases, however, the survey provides more detail than the TOS.
A-l
-------
TABLE A-l. PRODUCT/MARKET CROSS-REFERENCE
Market Segment
Exterior solventborne
Exterior waterborne
Interior solventborne
Interior waterborne
Solventborne primers
and undercoaters
group
Current Industrial Report Census Category
Exterior solvent paints and tinting
bases, including barn and roof paints
Exterior solvent enamels and tint.
including ex. -in floor enamels
Other exterior solvent coatings.
including bituminous paints
Exterior water pelnts end tinting bases.
including barn and roof paints
Exterior water exterior- -interior deck
end floor enamels
Other exterior water coatings
Interior flat solvent wall paints and
tinting bases
Interior solvent gloss and quick dry
enamels and other solvent paints and tint
Interior semigloss. eggshell, satin solvent
paints and tinting bases
Interior flat water paints and tinting
bases
Interior semigloss, eggshell, aatln and
other water paints and tints
Exterior solvent undercoaters and primers
Interior solvent undercoaters and primers
Census
Product
Code
2851112
2851115
2851139
2851141
2851142
2851155
2851163
2851165
2851169
2851181
2851183
2851125
2851171
VOC Emission Inventory
Survey Category
Roof coatings- -
solventborne
Flat, exterior.
solventborne
High performance arch.
coatings- -solvent
Bituminous coatings--
solvent
Nonflat, exterior--
solvent
Plat, exterior.
water borne
Roof coatings--
water borne
Nonflat, exterior--
waterborne
High performance arch.
coatings- -water
Bituminous coatings--
water
Plat, tnterlor--
solventborne
Nonflat, interior--
solventborne
Quick dry enamels
Plat, interior--
waterborne
Nonflat, interior--
water borne
Primers, solventborne
Undercoaters,
solventborne
Q.D. primers, sealers,
undercoaters - - sol vent
Table of Standards Proposed
Regulation Limits Category
Roof coatings*
Flat, exterior*
High performance, floor coatings*
Bituminous coatings and mastics*
High performance, rust preventive*
High performance, concrete
protective*
Nonflat, exterior*
Flat, exterior*
Roof coatings*
Nontlat, exterior*
High performance, floor coetings*
Bituminous coatings and mastics*
High performance, rust preventive*
High performance, concrete
protective*
Plat, interior.*
Nonflat, interior*
Quick dry enamels
Flet, interior*
Nonflat, interior*
Primers and undercoaters*
Quick dry primers, sealers.
undercoaters*
See notes at end of table.
(continued)
-------
TABLE A-l. PRODUCT/MARKET CROSS-REFERENCE (CONTINUED)
Market Safltsant
Currant Industrial Report Census Category
Census
Product
Code
VOC Emlnion Inventory
Survey Category
Table of Standard! Proposed
Regulation Limits Category
Waterborne primera
and undercoatera
group
Exterior water undercoetera and primers
Interior water undercoatera and prlaiara
2851144 Priatara, waterborne
2851186 Undercoaters, waterborne
Q.D. primers, sealers,
undercoaters--water
Primera and undercoaters*
Quick dry primers, sealers,
undercoatera*
Solventborne clear
coating, aealer,
and atain
Exterior aolvent clear finishes and aeelera 28511J5
Interior aolvent clear finishes and aealers 2851175
Exterior aolvent stains 1851137
Interior aolvent atains 1851177
Sealers, solventborne
Shellacs
Waterproofing sealers
w/pigment, solvent
Waterproofing sealers--
clear, solvent
Varnishes, solventborne
Stains, opaque, solventborne
Stains, semitransparent,
solventborn*
Seal era*
Sealera--shellacs, clear
Sealera--shellacs, opaque
Waterproofing aeelers.
opaque*
Materproofing sealers, clear*
Varnishes*
Staina, opaque*
Stains, clear and
semi transparent*
Waterborne clear
coatings and
stains group
Exterior water atains and sealers
Other interior water coatinga, ataina, and
aealers
2851149 Staina, opaque, waterborne
Sealers, waterborne
Staina, semi transparent,
waterborne
2851188 Waterproofing sealers
w/pigment. water
Waterproofing sealera--
clear, water
Varnishes, waterborne
Staina, opaque*
Sealers'
Staina, clear, and
seaU t ranaparent*
Waterproofing aaalers,
opaque*
Waterproofing aealera, clear*
Varnishes*
Staina, low solids
Lacquers
Architectural lacquers
2851193
Lacquers
Lacquers
Wood preservatives
group
Other miscellaneous allied paint producta,
including brush cleanera, nonpreaaur* wood
preservatives, putty, and glazing compounds,
etc.
2851598 Wood preservatives, below-
ground
Clear wood preservatives
feed transplant wood
preservatives
Wood preservatives, opaque
Wood preservetives, below-
ground
Wood preservetives, cleer,
and semitransparent
Wood preservetives, opaque
Traffic marking
painta
Traffic marking painta
2851311 Traffic paints
Traffic narking painta
See notes at end of table.
(continued)
-------
TABLE A-l. PRODUCT/MARKET CROSS-REFERENCE (CONTINUED)
Market Segment
Current Industrie! Report Census Cat•gory
Census
Product
Cod*
VOC Emission Inventory
Survey Category
Table of Standard* Proposed
Regulation Limits Category
Special purpose group Special purpose coatings, n.s.k.
2851300 Dry fog coatings
Metallic pigmented coatings
Antigraffiti coatings
Concrete curino compound*
ratm release compounds
Graphic arts coatings
Multicolor coatings
Pretreatmant wash primers
Swimming pool coatings
Dry fog coating*
Metallic pigmented coatings
Antifouling coatings
Antigraffiti coating*
Bond breekers
Chalkboard resurfacers
Concrete curing compounds
Form release compounds
Graphic arts coating*
Impacted imerslon coating*
Multicolor coating*
Pretreatment waah primers
Swinming pool coating*
Flow coatings
Industrial
Maintenance
group
Interior industrial new construction and
Maintenance paints
Exterior industrial new construction and
maintenance paint*
2851301
2851305
Antenna coating*
Extreme high durability
coatings
Fire-retardant/resistive
coatings
High temperature coatings
Magnesite cement coatings
Mastic texture coatings
Sanding sealer*
Industrial maintenance
coating*
Appurtenance*
Fir*-ret*rdant/re*i*tive
coatings
High temperature coating*
Hagnesit* cement coating*
Mastic texture coating*
Nonferrous ornamental metal
lacquer and surface
protectants
Nuclear power plant coatings
Repair and maintenance
thermoplastic
Sanding sealer*
Thermoplastic rubber coating*
and mastics
Industrial maintenance
coatings
Heat reactive coating*
• In the TOS, only one limit for both solvent and waterborne* 1* givani (or presentation the product 1* listed under both.
Source*; U.S. Department of Commerce. Current Industrial Report*: Paint and Allied Product*, 1991. Washington, D.C. Government Printing Office.
1993.
Industry Insights. Architectural and Industrial Maintenance Surface Coating* VOC Emission* Inventory Survey. Prepared for the National
Paint and Coating* Association in cooperation with the AIM Regulatory Negotiation Industry Caucus. Final draft report. 1993.
-------
Where possible, the market segments were paired as solvent
borne and waterborne coating categories. Separate market
segments could not be created for flat and nonflat coatings in
the interior and exterior segments because the Census data do
not differentiate between exterior flats and nonflats.
The necessary data were developed for each of the
13 market segments using the mapping scheme presented in
Table A-l. Data for individual Census product codes were
summed where necessary to compute prices and quantities.
A.2 ESTIMATING DEMAND ELASTICITIES FOR COATINGS
To perform the market analysis, own- and cross-price
elasticities of demand were estimated for four broad coating
categories: exterior solventborne and interior solventborne
and their two respective substitutes, exterior waterborne and
interior waterborne. The variables used in estimation are
domestic consumption quantity; real value of domestic
consumption; real consumption price; national income; a
housing variable; and the real price of alkyd resins, acrylic
resins, and titanium dioxide. Complete data for these
variables were collected for the years 1981 through 1991.
Justification of these variables and their data sources is
given below.
A.2.1 Estimation Procedure and Results
Econometric estimation of the interrelated demand system
for interior solventborne, interior waterborne, exterior
solventborne, and exterior waterborne architectural coatings
generated estimates of own-price demand elasticities for each
of the four groups and cross-price demand elasticities between
the solventborne and waterborne segments of each interior
(exterior) pair.
A-5
-------
The quantity demanded of a commodity is a function of its
price, the price of any substitutes and other factors, such as
income, that affect aggregate demand. Estimating the demand
function, however, is more complicated than just running
regressions of observed market quantities on observed market
prices and other demand variables. One must account for the
fact that the observed prices and quantities are equilibrium
values, which are simultaneously determined by both demand and
supply factors.
Variables that are determined within a system (such as
prices and quantities in a market equilibrium system) are
endogenous to that system, whereas those variables determined
outside of the particular system (e.g., income, housing
activity) are termed exogenous. In simultaneous equations
models, endogenous variables are correlated with the error
terms through solution of the system. As a result of the
interdependence of the endogenous variables and the error
terms, the application of standard regression techniques is
modified to estimate the effect of an endogenous right-hand
side variable (i.e., equilibrium price) on the endogenous
left-hand dependent variable (equilibrium quantity). In
general, ordinary least squares estimation of the individual
demand equations leads to biased and inconsistent parameter
estimates when a regressor is endogenous.
Endogeneity bias is corrected by applying the two-stage
least squares (2SLS) regression procedure for each estimated
equation (see, for example, Pindyck and Rubinfield5) . In the
first stage of the 2SLS method, the price observations were
regressed against all exogenous demand and supply variables in
the system. This regression produced fitted (predicted)
values for the price variables that are, by definition, highly
correlated with the true endogenous variable (the observed
A-6
-------
equilibrium price) and uncorrelated with the error term. In
the second stage, these fitted values were employed as
observations of the right-hand side price variables in the
demand equations. This procedure can also be used to estimate
the underlying structural supply equations; however, because
of the poor performance of various specifications in the
supply estimations, only demand estimates are reported here.
The 2SLS procedure was used to estimate the four demand
functions. Both linear and double-log regressions were
estimated. The double-log specifications are presented here
because of slightly better statistical fit and because the
parameter estimates are directly interpretable as point
elasticities.
For the two exterior categories, housing completions are
included as an exogenous demand determinant. Exogenous supply
factors incorporated into the first-stage regressions include
the prices of various raw material inputs and a price index
for substitute outputs, which captures the effect of non-
exterior coatings prices on the supply of exterior coatings.
For the two interior categories, U.S. domestic GNP is included
as a proxy for the exogenous effect of aggregate income on the
demand for interior coatings. Exogenous supply factors
incorporated into the first-stage regressions also include the
prices of various raw material inputs and a price index for
substitute outputs, which in this case captures the effect of
noninterior coatings prices on the supply of interior
coatings. The results of the demand estimations are shown in
Table A-2.
Unfortunately, sufficient data to estimate the demand
parameters for the other market segments were unavailable.
For the other two solvent/water-paired segments—clear coatings
and primers/undercoaters—the mean of the respective own- and
A-7
-------
TABLE A-2. DEMAND CURVE ESTIMATES
Adjustable Elasticity
Variable F£ F-Value Estimate t-statistic
Exterior solventborne 0.94 50.52
demand
Log-housing 0.17 3.30
completions
Log exterior -1.43 -1.89
solventborne price
Log exterior 0.20 0.36
waterborne price
Exterior waterborne 0.92 39.36
demand
Log-housing -0.05 -0.62
completions
Log exterior 0.51 0.42
solventborne price
Log exterior -1.89 -2.17
waterborne price
Interior solventborne 0.69 8.49
demand
Log GNP
Log interior
solventborne price
Log interior
waterborne price
Interior waterborne 0.99 588.90
demand
Log GNP
Log interior
solventborne price
Log interior
waterborne price
1.01
-1.50
1.43
1.00
0.36
-1.39
1.67
-1.74
1.28
5.07
1.28
-3.80
A-8
-------
cross-price elasticities from the interior and exterior
estimation process were used as proxies for the elasticities.
The other five segments—special purpose, industrial
maintenance group, traffic marking paints, lacquers, and wood
preservatives—are specialty groups whose demand is assumed to
be fairly inelastic and not dependent on prices in the other
segments. Therefore, a value of -0.5 for the own-price
demand elasticity and zero for all cross-price elasticities
were assigned to each of these categories. Table A-3 provides
the matrix of own- and cross-price elasticities for all
13 market segments.
A.2.2 Data Used in Demand Estimation
Domestic consumption quantities and values were
calculated using data from U.S. Department of Commerce
publications Current Industrial Reports: Paint and Allied
Products6'7-8'9-10-11-12-13-14 and U.S. Exports Schedule B Commodity by
Country.15-16-17'18'19-20-21-22 Domestic quantity and value of
shipments figures were used, which include exports. Exports
were then subtracted to estimate domestic consumption
(architectural coatings imports are negligible and are not
included in the consumption variable). Consumer price indexes
from the U.S. Department of Labor's Handbook of Labor
Statistics23 and the U.S. Department of Commerce's Survey of
Current Business24-25-26 are used to adjust the current figures
to real values. Real consumption price was imputed for each
product by dividing real value of domestic consumption by the
quantity of domestic consumption.
A-9
-------
TABLE A-3. DEMAND ELASTICITY MATRIX FOR ARCHITECTURAL COATINGS MARKET SEGMENTS
I
M
O
Percentage
Change in
Quantity Demand
of Product
Market aegnent
1
2
3
4
5
6
7
8
9
10
11
12
13
With Respect to
1
-1.43
0.51
0
0
0
0
0
0
0
0
0
0
0
2
0.20
-1.89
0
0
0
0
0
0
0
0
0
0
0
3
0
0
-1.50
0.36
0
0
0
0
0
0
0
0
0
4
0
0
1.43
-1.39
0
0
0
0
0
0
0
0
0
5
0
0
0
0
-1.47
0.44
0
0
0
0
0
0
0
Percentage Change in Price of
6
0
0
0
0
0.82
-1.64
0
0
0
0
0
0
0
7
0
0
0
0
0
0
-1.47
0.44
0
0
0
0
0
8
0
0
0
0
0
0
0.82
-1.64
0
0
0
0
0
9
0
0
0
0
0
0
0
0
-0.50
0
0
0
0
Product
10
0
0
0
0
0
0
0
0
0
-0.50
0
0
0
11
0
0
0
0
0
0
0
0
0
0
-0.50
0
0
12
0
0
0
0
0
0
0
0
0
0
60
-0.50
0
13
0
0
0
0
0
0
0
0
0
0
0
0
-0.50
-------
The GNP in constant 1987 dollars from 1981 through 1991
was used as an aggregate income measure.*'27-28'29 Housing
completions for 1981 through 1991 were obtained from the U.S.
Department of Commerce's Current Construction Reports.30
Prices for alkyd and acrylic resins are obtained from the U.S.
International Trade Commission publication Synthetic Organic
Chemicals, U.S. Production and sales .31'32'33'34'35-3S'37'39'39'40 Prices
for titanium dioxide were imputed using quantity and value of
shipment data for U.S. production from the Current Industrial
Reports, Inorganic Chemicals.41 Real prices for these raw
materials were calculated by deflating normal values using
CPIs. Alkyd and acrylic resins were used to represent raw
materials for the nonvolatile vehicle portion of the coatings,
which are found mainly in solventborne and waterborne
coatings, respectively. Titanium dioxide was used to
represent a raw material in the pigment portion of the
coating, which is found in both types of coatings. A
Laspeyres price index was constructed to incorporate the price
of substitute outputs as a supply-side effect in the first
stage regressions of the 2SLS procedure. Let the price and
quantity of commodity n in period t be pnc and q^,
respectively for n - 1, . .., N and t = 0, 1, ...,T. Then the
Laspeyres price index of the N commodities for period t
(relative to the base period 0) is defined as
N
*L'E
N=l
N
£
N=l
0 0
•All constant values were converted to 1982-1984 dollars for the
analysis to be consistent with the consumer price index (CPI), which has
1982-1984 as a base.
A-ll
-------
Real domestic prices and quantities of nonexterior
coatings were used to construct the price index for the
exterior coatings equations and real domestic prices, and
quantities of noninterior coatings were used to construct the
index for the interior coatings equations. Each index is
computed for the years 1981 through 1991, with 1981 serving as
the base year.
A.3 EVALUATION OF DATA QUALITY
The Current Industrial Report series is generally
considered a reliable source for quantities and values of
products shipped. Monthly and annual data were estimated from
a sample designed to measure activities of the entire paints
and allied products industries. Each annual report provides
data for 2 years, and figures from the 1991 report were used
for the coatings analyses. In addition to the four
representative coatings products, the architectural coatings
Census category includes two other products: architectural
lacquers and architectural coatings, not elsewhere
classified. These categories were not included in the
estimates because of insufficient data. However, in 1991,
these two product categories combined represented only 1.3
percent of the total value of shipments for the architectural
coatings market.42 Statistics reported in the Current
Industrial Reports at the seven-digit SIC product level are
based on Annual Surveys of Manufactures and represent about 95
percent of total shipments in the paint industry (SIC 2851)."
To produce estimates for the entire industry, the Census
Bureau inflates the quantity and value figures reported in the
annual survey by a factor based on data reported by all
A-12
-------
establishments in the 1987 Census of Manufactures.1* The
inflation factors for architectural coating product categories
are as follows: 1987 through 1991, 1.00; 1982 through 1986,
1.004; and for 1981, 1.04.c Quantity and value figures for
the four product categories used in the demand estimation are
inflated using these factors. Prior to 1981, data were not
collected at the more specific seven-digit SIC level. Using
the longer time series would provide more data points but
would also preclude analysis of the individual product
categories, and representativeness would be lost.
The export data used are the best publicly available;
however, combining export and domestic data to estimate
domestic consumption poses some problems. The classification
systems used to gather both types of data are different, and
the corresponding product categories used cannot always be
compared. For example, data from the U.S. Department of
Commerce publication U.S. Imports for Consumption and General
Imports, TSUSA Commodity by Country of Origin were not used
because the imported commodity classifications had no
comparable domestic output classification. Exclusion of
imports from the estimate of domestic consumption does not
pose a problem because in 1991 the value of imports for
architectural, OEM, and special purpose coatings (SIC 28511,
28512, 28513) combined represented less than 0.9 percent of
the total domestic value of shipments.44 Data from U.S.
Exports Schedule B Commodity by Country were available for
1981 through 1991, and the export categories correspond well
"The inflation factor for 1981 is based on 1977 Census relationships
and for 1982 through 1986 on 1982 Census relationships.
"The 1991 quantities and values used in the model (values to impute
price) also include products in the special purpose and miscellaneous
allied paint products categories. The special purpose inflator for 1991 is
1.06, and the miscellaneous inflator in 1991 is 1.18.
A-13
-------
with the four domestic product categories except for 1989
through 1991.*5-«• «7-*8-49-«• "•« In 1989, the export codes and
categories changed and are no longer compatible with the
domestic categories. In addition, quantities are reported in
kilograms rather than gallons, as they were in previous years.
For these reasons, export data were not used to adjust
domestic consumption after 1988. The GNP data typically
represent income for the entire nation including income
generated from American businesses located overseas. The
current price data for the paint products and raw materials
should be considered reliable, though their accuracy may be
affected by the exclusion of imports for the coatings products
and of exports and imports for the raw materials prices. CPIs
for all urban consumers with a base of 1982 through 1984 were
used in calculating real prices.
The raw material prices used are representative of the
entire U.S. and export market for these products, rather than
just the U.S. supply to the paints and coatings industry. The
alkyd resins were used in this estimation to represent an
input found only in solventborne coatings and acrylic resins
to represent an input found only in waterborne coatings.
However, some acrylic resin materials are used in some
solventborne coatings and alkyd resins are used as modifiers
in waterbornes. Exports and imports were not considered when
computing raw material supply prices because foreign trade
data were not available for alkyd and acrylic resins. In
1991, exports of titanium dioxide represented 17.9 percent of
the total domestic value shipped and imports were 10.9
percent."
A-14
-------
A.4 REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10
11
12.
U.S. Department of Commerce. Current Industrial Reports:
Paint and Allied Products, 1991. Washington, DC:
Government Printing Office. 1992.
Industry Insights. Architectural and Industrial
Maintenance Surface Coatings VOC Emissions Inventory
Survey. Prepared for the National Paint and Coatings
Association in cooperation with the AIM Regulatory
Negotiation Industry Caucus. Final draft report. 1993.
Ref. 1.
Ref. 2.
Pindyck, Robert S., and Daniel L. Rubinfield.
Econometric Models and Economic Forecasts. 2nd Ed.
York, McGraw-Hill, Inc. 1981.
New
U.S. Department of Commerce. Current Industrial Reports
Paints and Allied Products, 1982. Washington, DC:
Government Printing Office. 1983.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1983. Washington, DC:
Government Printing Office. 1984.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1984. Washington, DC:
Government Printing Office. 1985.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1985. Washington, DC:
Government Printing Office. 1986.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1986. Washington, DC:
Government Printing Office. 1987.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1987. Washington, DC:
Government Printing Office. 1988.
U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1988. Washington, DC:
Government Printing Office. 1989.
A-15
-------
13. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1989. Washington, DC:
Government Printing Office. 1990.
14. U.S. Department of Commerce. Current Industrial Reports:
Paints and Allied Products, 1990. Washington, DC:
Government Printing Office. 1991.
15. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1981. 1982.
16. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1982. 1983.
17. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1983. 1984.
18. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1984. 1985.
19. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1985. 1986.
20. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1986. 1987.
21. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1987. 1988.
22. U.S. Department of Commerce. U.S. Exports Schedule B
Commodity by Country. FT446/Annual 1988. 1989.
23. U.S. Department of Labor. Handbook of Labor Statistics.
Bulletin 2340. 1989.
24. U.S. Department of Commerce. Survey of Current Business.
v. 70, no. 7. July 1990.
25. U.S. Department of Commerce. Survey of Current Business.
v. 72 no. 7. July 1992.
26. U.S. Department of Commerce. Survey of Current Business.
v. 72, no. 12. December 1992.
27. Ref. 24.
28. Ref. 25.
29. Ref. 26.
A-16
-------
30. U.S. Department of Commerce. Current Construction
Reports, Housing Completions, November 1992. Washington,
DC: Government Printing Office. 1993.
31. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1981. USITC
Publication 2470. 1982.
32. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1982. USITC
Publication 2470. 1983.
33. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1983. USITC
Publication 2470. 1984.
34. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1984. USITC
Publication 2470. 1985.
35. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1985. USITC
Publication 2470. 1986.
36. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1986. USITC
Publication 2470. 1987.
37. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1987. USITC
Publication 2470. 1988.
38. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1988. USITC
Publication 2470. 1989.
39. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1989. USITC
Publication 2470. 1990.
40. U.S. International Trade Commission. Synthetic Organic
Chemicals, U.S. Production and Sales, 1990. USITC
Publication 2470. 1991.
41. U.S. Department of Commerce. Current Industrial Reports:
Inorganic Chemicals, 1991. Washington, DC: Government
Printing Office. 1992.
42. Ref. 1.
A-17
-------
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
Ref. 1.
Ref. 1.
Ref. 15
Ref. 16
Ref. 17
Ref. 18
Ref. 19
Ref. 20
Ref. 21
Ref. 22
Ref. 41
A-18
-------
APPENDIX B
SUMMARY OF REFORMULATION COST ESTIMATES FROM
PUBLIC COMMENTS
-------
At proposal, EPA's estimate for per product reformulation
cost was based on an estimate for a hypothetical new coating
included in a presentation to the Regulatory Negotiation
committee (July 28, 1993). This lump-sum cost estimate was
$250,000, implemented over three years at $83,333 per year.
During the public comment period, EPA solicited public
input regarding the size and nature of reformulation costs to
gauge the reasonableness of (and potentially modify) the
estimate used in the EIA. The public comments on costs were
reviewed for this purpose. Costs were organized along the
following dimensions:
• technical staff training
• prioritization of products needing reformulation
• survey available materials
• reformulate to desired properties
• performance tests
• field tests
• marketing costs
• production costs (labels)
• sales training
• executive expenses
B-l
-------
Upon review of the public comments on costs, eleven of
the responses appeared to provide comparable information for
estimating lump-sum reformulation costs per product. Other
responses presented costs for all of the company's products,
but did not provide information on the number of products to
enable computation of cost per product. Other responses
could not be used either because of incompleteness or lack of
clarity about the information provided. A list of and summary
statistics for the eleven potentially comparable responses
plus the original Regulatory Negotiation committee estimate
are presented in Table B-l. Note that two of the estimates
are alternative interpretations of the same estimate. One
interpretation estimates per-product cost by dividing
the company's total cost estimate by all noncompliant
formulas. The other interpretation is that the total cost
number is divided by the subset of formulas that are most
feasible to reformulate. It was unclear from the comment,
which number the company used to estimate its total compliance
costs, so both interpretations were used to provide a range.
Cost per product estimates (in 1991 dollars) range from
$576 to $272,000, with a mean value of $86,326. The mean
value was rounded up to $87,000 to provide the model product
cost estimate used throughout the analysis. As the summary
statistics in Table 2-1 indicate, the central tendency cost
estimates (mean and median) are well-below the $250,000 lump-
sum cost per product estimate used in the EIA at proposal,
ranging anywhere from 20 to 35 percent of that estimate.
In summary, a review of the public comments related to
reformulation costs suggests that EPA may have significantly
overestimated the per-product costs by a factor of three to
five times at proposal. Because it is based on information
B-2
-------
TABLE B-l. REFORMULATION-RELATED COST ESTIMATES
Public Comment Docket Number
Estimated Cost
per Product
(current $)
Estimated Cost
per Product
($ 1991)'
IV-D-217 (Interpretation 1,
Total cost divided by all noncompliant
products)
IV-D-217 (Interpretation 2,
Total cost divided by most feasible
reformulations)
IV-D-108
IV-D-110
IV-D-130
IV-D-93
IV-D-152
IV-D-36
IV-D-38
IV-F-le
IV-D-182
II-E-52
Summary statistics
N =
Min
Max
Mean
Median
15,764
48,220
63,500
13,000
20,300
656
122,417
51,210
310,000
150,000
96,000
267,000
13,832
42,311
55,719
11,407
17,812
576
107,416
44,935
272,013
131,619
84,236
254,038
12
576
272,013
86,326
50,327
a Converted from year in which estimate is given (usually 1996) to 1991
using the Gross Domestic Product Price Deflator.
Source: U.S. Department of Commerce, Bureau of Economic Analysis, August
1997.
provided in the public comment period, the revised estimate
used in this analysis should provide a more valid estimate of
reformulation-related costs than the estimate used at
proposal. Alternative methods for annual!zing the lump-sum
cost estimate of $87,000 are presented in the main text.
B-3
-------
APPENDIX C
CALCULATION OF REGULATION-INDUCED COSTS WHEN REFORMULATION
NORMALLY OCCURS AT FIXED TIME INTERVALS
-------
One complicating factor in estimating the cost of the
regulation is the fact that product reformulation is a normal
business activity in the architectural coatings industry.
Therefore, rather than viewing the regulation as creating
reformulation responsibilities (the maintained assumption
throughout the analysis) , one might take the alternative view
that a different time pattern of reformulation is created,
thereby leading to a lower estimate of regulatory costs. This
appendix presents the issue analytically and develops a
numerical example to quantify the difference in costs under
the alternative assumptions.
Suppose a company routinely reformulates products every
eight years. If the average product is product midway through
its reformulation cycle, it will be reformulated four years in
the future in the absence of the regulation. However, the
regulation requires them to do the reformulation now rather
than four years in the future and this acceleration imposes
costs on the firm. To estimate the costs of this
acceleration, assume the initial reformulation cost of $87,000
occurs in the first year. Then the net present value, today,
of a cost otherwise deferred four years into the future is
NPV(-4) = $87,000/1.07* = $66,372
Instead, the company is required to reformulate today at a
cost of
C-l
-------
NPV(O) = $87,000
The net effect on the company of accelerating the next
formulation is then
Initial Net effect = NPV(-4) - NPV(O) = -$20,628
Thus, if the regulation just accelerates the next
reformulation, the one-time cost of that acceleration is
approximately $20,000. This is substantially below the one-
time cost of $87,000 currently assumed in the EIA. However,
if it is assumed that this requirement also forces all future
reformulations to be moved up four years, then the computation
must be expanded to measure the present value of the current
and all future adjustments. To start, the present value of an
initial $87,000 cash expenditure repeated every eight years
thereafter can be written
V(0) = $87,000 + $87,000*(l/( (1.07)" - 1))
= $208, 139
Without the regulation, this stream of costs would be deferred
four years into the future. Evaluating this in present value
terms gives
V(-4)= V(0)/1.07* = $158,788
Thus, the difference in present value between the two
reformulation cost streams is the total net effect of
accelerating this and all future reformulations.
Total net effect = V(-4) - V(0) = $49,351
C-2
-------
This can be viewed as conceptually equivalent to a one-
time cost of the regulation for an average product that is
over-the-limit. This explicitly accounts for the net present
value of the regulation's affect on all future formulations.
This one-time cost is substantially below the $87,000 one-time
cost assumed in the analysis.
By comparison, if the product were otherwise to be
reformulated one year in the future without the regulation,
the present value of this cost acceleration can be computed is
a similar fashion as $13,617 (16 percent of $87,000). If the
previous reformulation had been implemented just one year
before the regulation, then the present value of accelerating
the future reformulation cycle by seven years would be $78,520
(90 percent of $87,000).
In summary, the one-time cost estimate of an accelerated
reformulation schedule ranges from a small fraction to a large
fraction of the reformulation cost estimate used in the EIA.
In this example, the average product's one-time cost
equivalent is less than 60 percent of the estimate used in the
EIA. Thus, EPA contends that it has provided a conservatively
high estimate of the true incremental cost of reformulating a
product subject to the regulation.
C-3
-------
APPENDIX D
METHODOLOGY FOR COMPUTING MARKET AND
WELFARE ADJUSTMENTS
-------
D.I METHODOLOGY FOR COMPUTING SUPPLY EFFECTS
For the purposes of modeling the regulatory effects in
each market, products are separated into four categories,
based on their producers' response to the regulation:
• products slated for withdrawal,
• products on which exceedance fees are paid,
• products slated for reformulation, and
• products unconstrained by the regulation.
The baseline (preregulatory) quantities from these groups
are denoted as follows: Q*, Qr, Q*, and Qa for groups 1, 2, 3,
and 4, respectively. Total baseline market output equals the
sum of the four components:
Q = Q* + QF + Q* + Qu. (D.I)
Figure D-l depicts the aggregation of these subgroups
into a market supply function. The regulation causes a shift
in the aggregate supply function depicted in Figure D-l as a
result of two phenomena: an inward supply shift due to
eliminating Group 1 through product withdrawals (e.g., the
shift from S° to S1), and an upward supply shift due to
imposing per-unit fees on the products from Group 2 (the shift
from S1 to S1') . There is no supply shift emanating from
D-l
-------
P1
Po
Q- QO
Q/t
Figure D-l. Single market effects of VOC content regulation,
Group 3 because the reformulation is assumed not to affect
marginal production costs, and there is no shift from Group 4
because the unconstrained products experience no regulation-
induced change in their cost structure. So the full
regulation-related shift is from S° to S1', which leads to a
new market equilibrium. At the new equilibrium, price rises
to P' and quantity falls to Q' .*
"This graphical analysis demonstrates that the post-regulatory market
effects are uncertain if the analysis were to consider the possibility that
the reformulation process changes the marginal cost of producing the
coating as a result of changes in material or labor costs, for example.
This empirical issue can be resolved given sufficient data on the effect of
VOC content on production costs for all affected products. Unfortunately,
these data were not available for this study, so the appropriate empirical
analysis could not be conducted to draw such conclusions.
D-2
-------
D.2 DEMAND EFFECTS
Figure D-l depicts a partial equilibrium view of the
short-run effect of imposing content limits in one market.
One must also consider the role of substitute products in
determining the equilibrium adjustments, which suggests a
multimarket perspective. Figure D-2 depicts the markets for
two products (A and B) that are demand substitutes. The price
of product B factors into product A's demand function and vice
versa:
DA = DA(PA, P.) (D.2)
DB = D^PB, PA) . (D.3)
Given that A and B are substitutes implies
6DA / 5PB > 0 (D.4)
/ 5PA > 0 . (D.5)
Suppose the supply of A is affected by the content limits
in the manner described above, but that the supply of B is
unaffected. This initiates a supply shift in market A from SA°
to SAR. Holding the initial demand function constant, this
shift would generate an equilibrium quantity of QA* and price
of PA". However, the associated price increase in market A
induces an outward shift in the demand for product B, which
raises the price of product B. Likewise, the increase in B's
price leads to an outward shift in the demand for product A,
which raises its price and so on. This interaction continues
until post-regulatory equilibrium is established at (PA*, Q/) ,
(PBR, QB*)-
D-3
-------
Dj (Pg)
o; Q;
QAA
$/Oi
Product A
08
D(P)
Product B
Figure D-2. Multiple market effects of VOC regulations
D-4
-------
D.3 COMPUTING CHANGES IN EQUILIBRIUM PRICES AND QUANTITIES
The change in equilibrium prices and quantities for the
products affected by the content limits and their substitutes
can be numerically computed by adjusting the equations in the
multimarket supply and demand system to reflect the imposition
of these limits. For each market, i, the equilibrium change
in quantity supplied of each product affected by the
regulations equals the sum of the supply changes from each of
the producer subgroups:
AQiS = AQiX + AQ/ + AQi" + AQi". (D.6)
The change (from baseline) in quantity supplied by the
withdrawal sector is simply the negative of the quantity
originally supplied by that group:
AQiX = - Qix. (D.7)
The change in quantity supplied from the fee-paying
sector is specified as follows:
AQ/ = e/JQi'/Pi) (APt - Ft) (D.8)
where ej is the supply elasticity of the fee producers in
market i, APt is the change in equilibrium market price, and
other terms are as previously defined (without the
subscripts). APA- Fx is the change in "net price' for the fee-
paying producers (i.e., the change in unit process less the
unit fee).
The changes in quantity supplied from the reformulating
group and unconstrained groups, respectively, are
D-5
-------
(D.9)
AQiU = elu(Qi°/P)AP1. (D.10)
These producers respond to the increase in price with no
counteracting effect on costs. Given the higher price in the
post-regulatory equilibrium, output will increase from these
two groups of producers.
The aggregate change in equilibrium supply quantity can
now be restated by combining the preceding five equations:
AQ,S = - Q/ + e/tQ/XPi) (AP.-FJ + e/fQjVPJ APi
(D.ll)
The change in market demand for each product is given by
AQiD = Eu(Qi /PJAP,. + Ei-jCQi/PjjAPj (D.12)
where Eu is the own-price demand elasticity for product, i and
Eij is the associated cross-price demand elasticity between
products i and j . Consumer demand theory supports the
assertion that own-price elasticities are negative and that
cross-price elasticities of substitutes are positive. To
attain equilibrium, the change in quantity demanded must equal
the change in quantity supplied in both markets:
This provides a system of M*3 equations in M*3 unknowns,
where M equals the number of markets affected by the
regulation. This can be reduced to an M*2 equation system,
simply by substituting AQiD = AQiS = AQi. This system can be
solved simultaneously to compute the change in equilibrium
price and change in equilibrium quantity for each market. To
D-6
-------
do this, baseline market data, model parameters (supply and
demand elasticities), and an empirical characterization of the
various supply shocks alluded to above are needed.
D.4 COMPUTING WELFARE EFFECTS
Changes in the market equilibrium cause changes in
resource allocation, which, when quantified, provide measures
of how the welfare costs of the regulation are distributed
across groups affected by the regulation. The groups focused
upon here are architectural coatings producers and consumers,
because the changes in prices and quantities directly affect
their welfare. Since fee payments are considered, the
government sector is also included in the welfare analysis
because they collect the fee revenues. This study does not
measure the welfare benefits of reductions in VOC emissions, a
value against which these costs may be measured to determine
the net value to society of the proposed regulatory structure.
D.4.1 Effects on Architectural Coatings Producers
The profits earned at the new equilibrium to the profits
earned at the old equilibrium can be compared as a measure of
effects of the regulation on the individual producer.
Foregone baseline profits (n°) provide a measure of the loss
to producers that choose to exit rather than reformulate:
An = nR* - n° = -n°. (D.14)
For the remaining producers, the change in profits is affected
by several factors, including the incurrence of the fixed
reformulation cost and any associated changes in price,
quantity, and marginal cost.
D-7
-------
The remaining firms' costs may be affected through either
the reformulation cost or the fee payment. The effect of the
content limit on producers is generally not uniform and thus
raises some distributional considerations. As indicated
above, shifts in the aggregate supply function will cause the
market price to rise. For some producers, the benefits of the
price increase may outweigh the net costs of compliance. This
is certainly the case for producers of coatings with VOC
content below the regulatory standards, because they incur no
reformulation costs but would gain from the rise in market
price sparked by the compliance costs and/or product
withdrawals incurred by their competitors. Alternatively,
fixed reformulation costs may be substantial for some
producers, outweighing the positive price effect. The profit
effect will be negative for those producers. Other producers
may fall in the midrange, where the price benefits and cost
effects essentially offset each other.
Changes in producer welfare are generally reported as
changes in producer surplus. The aggregate change in producer
surplus for the withdrawn-product producers equals the sum of
forgone profits from all withdrawn products in market i:
The j subscript indicates forgone profits from the j'th
product in market i. NAA is the number of withdrawn products
in market i. The change in producer surplus from the
reformulating sector can be approximated as follows:
D-8
-------
APi is the change in equilibrium price, AQ/ is the change in
equilibrium quantity from the reformulating producers, Qi* is
the initial quantity of the reformulating producers, R,c is the
annualized reformulation costs, and NtR is the number of
products needing reformulation.
The change in producer surplus for the fee-paying
producers is initially computed as follows:
APS/1 = (APi-FJMQ/ + AQ/J-O.S'AQ/MAPi-Fi) . (D.17)
The first term reflects the net revenue effects of the price
rise less the fee payment and the second term reflects changes
in deadweight loss. To this term we must add the fixed cost
(per product) associated with fee recordkeeping requirements
so that the full welfare effect is
APS/ = APS/1 - FF • Ni (D.18)
where FF is the fixed cost per product of fee recordkeeping
and Ni equals the number of products subject to the fee in
market i .
Finally, the change in producer surplus for unconstrained
producers is
= APi • Qiu + 0.5 • AQiU • APi (D.19)
with the Qi° reflecting the quantity supplied by these
producers. Total (net) producer surplus effects is simply the
sum of the terms above:
i = APSiX + APS^ + PS/ + APSt°. (D.20)
D-9
-------
D.4.2 Effects on Architectural Coatings Consumers
Changes in consumer welfare are measured by the change in
consumer surplus, which quantifies losses due to a combination
of the higher price and reduced consumption quantity. This
change can be approximated as follows:
ACSj. = -APi • (Qi + AQi) + 0.5 • APi • AQi. (D.21)
D.4.3 Effects on the Government Sector
The transfer of fees from the fee-paying producers to the
recipient of those fees must be considered. For the purposes
of the welfare analysis, the government is identified as the
"recipient" of the fees.
AGSi = Fi • (Q/ + AQ/) . (D.22)
Ultimately, the government may choose to redistribute
those fees back to affected producers or consumers or back to
other members of society via the Treasury; however, for
purposes of quantifying these distributional flows, they are
assigned as gains to the government sector.
D.4.4 Net Welfare Effects
The net welfare effects are computed by taking the sum of
producer, consumer, and government surplus:
+ ACSj. + AGSi. (D.23)
This calculation nets out any transfers from one group to
another within society (e.g., transfers from consumers to
producers through higher prices and transfers of fee revenues
D-10
-------
from producers to the government) because these transfers do
not affect the total sum of resource costs, just how they are
distributed within society. AWFi provides an estimate of the
net social costs of the regulation.
D-ll
------- |