                               Chapter 5:  Cost Estimates

Synopsis

This chapter summarizes the data sources and methodology used to
estimate the costs of attaining the alternative more stringent levels
for the ozone primary standard analyzed in this RIA (bounds of the
proposed range, 0.075 and 0.070 ppm, and the more stringent alternative
of 0.065 ppm.).  First, the chapter presents cost estimates for the
hypothetical control strategy outlined in Chapter 3 (which uses
currently available known and supplemental controls).  Second, the
chapter presents an analysis of the economic impacts of the hypothetical
national control strategy using currently available controls.  Finally,
the chapter presents a range of estimates for the costs of the
additional tons of emissions that need to be extrapolated to move from
predicted ozone concentrations following the known control strategy to
attainment of the alternate standards analyzed (methodology and numbers
discussed in Chapter 4).  

As noted in Chapter 3, EPA first modeled a hypothetical control strategy
aimed at attaining a tighter standard of 0.070 ppm in 2020.  These
controls were insufficient to bring all areas into attainment with 0.070
ppm, and EPA then developed methodology to estimate additional tons of
emissions needed to attain the bounds of the proposed range, 0.075 and
0.070 ppm, and the tighter alternative of 0.065 ppm.  This chapter
presents the costs associated with each portion of the control analysis,
clearly identifying the relative costs of modeled versus extrapolated
emissions reductions as well as providing an estimate of the total cost
of reaching attainment nationwide.    Section 5.1 summarizes the
methodology and the engineering costs associated with applying known and
supplemental controls to partially attain a 0.070 ppm alternative
standard, incremental to reaching the current baseline (effectively
0.084ppm) in 2020.  Section 5.2 describes the full economic impact that
could be expected to result from the application of these known and
supplemental controls.  This economic impact analysis (EIA) is the study
of the way in which the direct benefits and costs of a regulation affect
the local, regional, or national economy. It attempts to measure the
consequences that the regulation will have on considerations such as
local or regional employment patterns, wage levels, business activity,
tourism, housing, and even migration patterns.

 (compared to the 0.070 ppm case).   

Section 5.4 then combines the results from Sections 5.1, 5.2, and 5.3 to
describe the total estimated cost of reaching attainment nationwide,
including both the full economic costs of modeled controls for reaching
partial attainment (including engineering costs plus social costs),and
the additional costs of tons of extrapolated emissions reductions needed
to reach attainment.  

The costs described in this chapter generally include the costs of
purchasing, installing, and operating the referenced technologies.  For
a variety of reasons, actual control costs may vary from the estimates
EPA presents here.  As discussed throughout this report, the
technologies and control strategies selected for analysis are
illustrative of one way in which nonattainment areas could meet a
revised standard.  There are numerous ways to construct and evaluate
potential control programs that would bring areas into attainment with
alternative standards, and EPA anticipates that state and local
governments will consider programs that are best suited for local
conditions.  Furthermore, based on past experience EPA believes that it
is reasonable to anticipate that the marginal cost of control will
decline over time due to technological improvements and more widespread
adoption of previously niche control technologies.  Also, EPA recognizes
the extrapolated portion of the cost estimates reflects substantial
uncertainty about which sectors, and which technologies, might become
available for cost-effective application in the future.

It is also important to recognize that the cost estimates are limited in
their scope.  Because we are not certain of the specific actions that
states will take to design State Implementation Plans to meet the
revised standards, we do not present estimated costs that government
agencies may incur for managing the requirement and implementation of
these control strategies or for offering incentives that may be
necessary to encourage or motivate the implementation of the
technologies, especially for technologies that are not necessarily
market driven.  This analysis does not assume specific control measures
that would be required in order to implement these technologies on a
regional or local level. 

5.1	Modeled Controls 

5.1.1 Sector methodology

5.1.1.1 Non-EGU Point and Area Sources:  AirControlNET

After designing a national hypothetical control strategy to meet an
alternative standard of 0.070 ppm using the methodology discussed in
Chapter 3 (see sub-section 3.2.1), EPA used AirControlNET to estimate
engineering control costs.  AirControlNET calculates costs using three
different methods:  (1) by multiplying an average annualized
cost-per-ton estimate against the total tons of a pollutant reduced to
derive a total cost estimate; (2) by calculating cost by using an
equation that incorporates information regarding key plant information;
or (3) by using both cost per ton and cost equations.  Most control cost
information within AirControlNET has been developed based on the
cost-per-ton approach. This is because estimating cost using an equation
requires more data, and parameters used in other non-cost per ton
methods may not be readily available or broadly representative across
sources within the emissions inventory.  The costing equations used in
AirControlNET require either plant capacity or stack flow to determine
annual, capital and/or operating and maintenance (O&M) costs.  Capital
costs are converted to annual costs, in dollars per ton, using the
capital recovery factor.  Applied controls and their respective costs
are provided in Appendix 6.

for NOx reductions from these sectors.  There were two controls whose
cost per ton was greater than this cap, due to the large capital
component of installing these controls.  A similar process was followed
for reductions from VOCs.  The marginal cost curve was analyzed, and
there was a clear break in the curve at approximately $6,000/ton.  At
this cap, over sixty percent of the possible reductions are being
controlled at less than thirty percent of the total cost of the VOC
reductions.  

5.1.1.2 EGU Sources:  The Integrated Planning Model

Costs for the electric power sector are estimated using the Integrated
Planning Model (IPM).  The model determines the least-cost means of
meeting energy and peak demand requirements over a specified period,
while complying with specified constraints, including air pollution
regulations, transmission bottlenecks, fuel market restrictions, and
plant-specific operational constraints.   IPM is unique in its ability
to provide an assessment that integrates power, environmental, and fuel
markets.  The model accounts for key operating or regulatory constraints
(e.g. emission limits, transmission capabilities, renewable generation
requirements, fuel market constraints) that are placed on the power,
emissions, and fuel markets.  IPM is particularly well-suited to
consider complex treatment of emission regulations involving trading and
banking of emission allowances, as well as traditional
command-and-control emission policies.  Applied controls and their
respective costs are provided in Appendix 6.?

5.1.1.3 Onroad and Nonroad Mobile Sources 

Cost information for mobile source controls was taken from studies
conducted by EPA for previous rulemakings and studies conducted for
development of voluntary and local measures that could be used by state
or local programs to assist in improving air quality.  Applied controls
and their respective costs are provided in Appendix 6.?  

5.1.2 Known Controls— Cost by Sector

In this section, we provide engineering cost estimates of the control
strategies identified in Chapter 3 that include control technologies on
non-EGU stationary sources, area sources, EGUs, and onroad and nonroad
mobile sources.  Engineering costs generally refer to the capital
equipment expense, the site preparation costs for the application, and
annual operating and maintenance costs.  The economic impact analysis,
following in section 5.2, also provides a more in-depth evaluation of
how these engineering costs will impact society through a distributional
analysis of changes in price and production levels in affected
industries, and who will bear the burden of the regulatory costs
(consumers or suppliers).  

The total annualized cost of control in each sector in the control
scenario is provided in Table 5.1.  These numbers reflect the
engineering costs across sectors annualized at an interest rate of 7
percent for control measures applied to non-EGU point, area, and mobile
sources.   This interest rate is consistent with the guidance provided
in the Office of Management and Budget’s (OMB’s) (2003) Circular
A-4.  Also consistent with that guidance, we provide annualized control
costs for non-EGU point, area, and mobile sources at a 3 percent
interest rate to show the sensitivity of our annualized control costs to
the choice of interest rate.



		Total	$200

 B. Onroad	$920

 C. Nonroad	$160

		Total	$1,080

D.  Non-EGU Sector

	   Point Sources (Ex: Pulp & Paper, Iron & Steel,  

   Cement, Chemical Manu.)	$2,700

E.  Area Sector	

	   Area Sources (Ex: Res. Woodstoves, Agriculture)	$320

		Total	$3,020

	Total Annualized Costs 

(using a 7% interest rate)	$4,300

	Total Annualized Costs 

(using a 3% interest rate)	$3,565



5.1.3 Limitations and Uncertainties Associated with Engineering Cost
Estimates

EPA bases its estimates of emissions control costs on the best available
information from engineering studies of air pollution controls and has
developed a reliable modeling framework for analyzing the cost,
emissions changes, and other impacts of regulatory controls.  The
annualized cost estimates of the private compliance costs are meant to
show the increase in production (engineering) costs to the various
affected sectors in our control strategy analyses.  To estimate these
annualized costs, EPA uses conventional and widely-accepted approaches
that are commonplace for estimating engineering costs in annual terms. 
However, our cost analysis is subject to uncertainties and limitations.

There are some unquantified costs that are not adequately captured in
this illustrative analysis. These costs include the costs of federal and
State administration of control programs, which we believe are less than
the alternative of States developing approvable SIPs, securing EPA
approval of those SIPs, and Federal/State enforcement.  Additionally,
control measure costs referred to as "no cost" may require limited
government agency resources for administration and oversight of the
program not included in this analysis; those costs are generally
outweighed by the saving to the industrial, commercial, or private
sector.  The Agency also did not consider transactional costs and/or
effects on labor supply in the illustrative analysis.  

The direct engineering costs estimated in this RIA do not reflect the
actual impact of these illustrative controls on consumers.  Given some
price elasticity of demand for products whose consumption is affected by
the implementation of these illustrative controls, the actual impact to
consumers will be less than that implied by the direct engineering
controls.  The greater the price elasticity of demand for a given
affected product, the more rising costs will reduce demand for that
product by a consumer. 

From another vantage point, the illustrative analysis does not take into
account the potential for advancements in the capabilities of pollution
control technologies as well as reductions in their costs over time. 
This is discussed further later in this chapter.

5.2 Economic Impact Analysis 

This section presents the economic impact results of the control
strategy developed by EPA for the purpose of providing an approach of
actions that could be taken to meet attainment of 8-Hour 0.070 ppm ozone
alternate standard.  Given the possible impacts of this guidance on
manufacturing industries, the transportation sector, electricity
generators, consumers, and U.S. Gross Domestic Product (GDP) as a whole,
we believe it is important to gauge the extent to which other parts of
the economy might also be affected by the implementation of the 0.070
ppm ozone standard.  Therefore, an analysis of the economy-wide effects
of implementing the alternate standard is conducted by applying
estimated direct costs to EPA’s computable general equilibrium model
(EMPAX-CGE).  As the chapter will show, the social costs for this
standard are slightly greater than the engineering costs applied to the
CGE model.  

This section contains four considerations that assist in interpreting
the economic impacts and relating these impacts to the attainment costs
presented in Section 5.1.

The selection criteria for the 0.070 ppm ozone control strategy, and its
related compliance costs, is designed to select the least cost controls,
from an engineering cost standpoint, that generate the highest ozone
reductions, but not necessarily the lowest economic impact.  Therefore,
although the control strategy is selected to reduce ozone at the lowest
engineering cost, it does not represent the lowest impact strategy from
a social cost standpoint.  Thus, while this economic impact analysis
presents results for the control strategy approach detailed in Chapter 3
of the RIA, it should not be viewed as the only economic impact estimate
of the ozone 0.70 ppm standard or even as the approach with the lowest
social cost.  Instead, the results should be viewed as guidance or
useful information for states preparing their implementation plans.  It
is likely that states will design implementation plans that present an
alternative control strategy and in some cases design plans that take
into account secondary impacts to industries and consumers within their
borders.  In such a case, the end result would be a set of SIPs that are
more economically optimal and may have lower industry impacts than those
described below.  

As noted previously, we are extrapolating the costs of meeting the 0.075
and 0.065 ppm alternatives.  Extrapolating does not allow us to directly
predict the control strategy used to attain these levels as was done for
the 0.070 ppm alternative.  Thus, it is not possible to calculate the
economic cost for these options.  

and productivity expected to result from air quality improvements.  If
these labor productivity improvements were included, the small
production output decreases projected by the model might be partially or
entirely offset.  EPA continues to investigate the feasibility of
incorporating labor productivity gains and other beneficial effects of
air quality improvements in CGE models.  

5.2.1	Background

To complement the analysis of effects on specific manufacturing sectors
as described in the control strategy in chapter 3, the macroeconomic
implications of the ozone 0.070 ppm standard has been estimated using
EPA’s EMPAX-CGE model.  The focus of this component of the analysis is
on examining the sectoral and regional distribution of economic effects
across the U.S. economy.  This section briefly discusses the EMPAX model
and the approach used to incorporate findings from other models in
EMPAX-CGE.  

EMPAX was first developed in 2000 to support economic analysis of
EPA’s maximum achievable control technology (MACT) rules for
combustion sources (reciprocating internal combustion engines, boilers,
and turbines).  The initial framework consisted of a national
multimarket partial-equilibrium model with linkages only between
manufacturing industries and the energy sector.  Modified versions of
EMPAX were subsequently used to analyze economic impacts of strategies
for improving air quality in the Southern Appalachian mountain region as
part of efforts associated with the Southern Appalachian Mountain
Initiative (SAMI).

Recent work on EMPAX has extended its scope to cover all aspects of the
U.S. economy at a regional level in either static or dynamic modes. 
Although major regulations directly affect a large number of industries,
substantial indirect impacts can also result from changes in production,
input use, income, and household consumption patterns.  Consequently,
EMPAX now includes economic linkages among all industrial and energy
sectors as well as households that supply factors of production such as
labor and purchase goods (i.e., a CGE framework).  This gives the
version of EMPAX called EMPAX-CGE the ability to trace economic impacts
as they are transmitted throughout the economy and allows it to provide
critical insights to policy makers evaluating the magnitude and
distribution of costs associated with environmental policies.  The
dynamic version of EMPAX-CGE employed in this analysis, and its data
sources, is described in past analyses of the CAIR and PM NAAQS, and
also in publicly available documentation on the EPA website (see Ross et
al., 2005). 

EMPAX-CGE can be used to analyze a wide array of policy issues and is
capable of estimating how a change in a single part (or multiple parts)
of the economy will influence producers and consumers across the United
States.  However, some types of policies, including the Ozone National
Ambient Air Quality Standard, are difficult to capture adequately within
a CGE structure because of the boiler- and firm-specific nature of
emission reduction costs.  Consequently, an interface has been developed
that allows linkages between EMPAX-CGE and the detailed technology
models discussed in Chapter 3.  These linkages give the combined
modeling system the advantages of technology detail and broad
macroeconomic coverage, thereby permitting EMPAX-CGE to investigate
economy-wide policy implications. 

As discussed in section 5.1, the models used in developing the control
strategy estimate cost changes by industry and region of the United
States for the sectors of the economy affected by the proposed alternate
ozone standard.  In order for EMPAX-CGE to effectively incorporate these
additional costs, they have to be expressed in terms of the productive
inputs used in CGE models (i.e., capital, labor, and material inputs
produced by other industries).  Rather than assume the costs represent a
proportional scaling up of all inputs, Nestor and Pasurka (1995) data on
purchases made by industries for environmental-protection reasons are
used to allocate these additional expenditures across inputs within
EMPAX-CGE.  Once these expenditures are specified, the incremental costs
from the technology models can be used to adjust the production
technologies in the CGE model.  Additional linkages are made between
EMPAX-CGE and IPM to handle specific IPM findings related to resource
costs and fuel consumption in electricity generation.

5.2.2 Results for Ozone Alternate Standard (0.070 ppm)

This section compares attainment of the ozone 0.070 ppm standard
incremental to the current 8-Hour standard (0.084 ppm, effectively….)
baseline.  Impacts are measured assuming a 2020 implementation year and
are the result of engineering costs described in section 5.1  Thus, the
following graphs compare the ozone 0.070 ppm standard to an economic
growth path that incorporates impacts from our baseline (including CAIR,
CAMR, CAVR, PM2.5 NAAQS 15/35), and ozone current standard through the
year 2020.

Projected Impacts on U.S. Industries of Incremental Costs of Reaching
Tighter Standard

Impacts of the ozone 0.070 ppm standard on manufacturing costs can
affect output and prices of all industries in the EMPAX-CGE model. 
These effects may increase or decrease output and/or revenue, depending
on their implications for production costs and technologies and shifts
in household demands.  In general, the impacts on industries will be
dependent on the control strategy and follow a pattern similar to the
stringency of the ozone standard.  

As shown in Figure 5.1, impacts on industrial output quantities are
generally small across all industries for ozone 0.070 ppm.  Outside of
the energy-intensive sectors (EIS), estimated changes in output of
manufactured goods are less than two one-hundredths of one percent
(0.02%).  Effects on energy producers are somewhat higher and can be
positive or negative, which limits any spillover effects to other
businesses and households.  

As described in chapter 3, selected control options for the ozone 0.070
ppm standard involve additional actions by electric utilities, which
tend to slightly increase coal consumption (influencing U.S. coal
production) and reduce natural gas use.  Other energy industries also
engage in additional measures, which can affect energy users such as the
EIS sectors.  Cement, chemicals and glass production are influenced by
direct control costs on their respective industries and any changes in
energy markets.  Note, however, that across energy-intensive industries
as a group, output quantities decline on average by less than a quarter
of a percent (<0.25%).

Figure 5.2 shows how these changes in output quantities (or units)
compares to changes in gross output revenues, where revenue changes
include the effects of changes in both quantity and output prices (which
reflect changes in production costs).  While additional gross revenues
may not imply that net revenues have increased for a given industry,
Figure 5.2 is useful in illustrating the overall changes occurring in
the economy in dollar terms.  What about transportation, given the
amount of controls being applied.

Revenues in energy-producing industries follow changes in output
quantity relatively closely as there are only limited changes in energy
prices, which are not sufficient to reverse any quantity/price effects. 
There are some minor differences between declines in energy-intensive
production and potential increases in gross revenues.  Particularly for
cement, output declines are accompanied with a rise in production costs
that results in gross revenues increasing in the industry, although this
effect across the entire economy is on the order of $250 million.  At
the other end of the spectrum, the sheer size of the services industry
in the U.S. means that, even though there are essential no changes in
output (-0.01%), when expressed in revenue terms there is a $5 billion
decline in gross revenues.  This occurs because the base output of the
services industry against which these percentage declines are measured
is more than $18 trillion in 2020.

Fig 5.1 Ozone 0.070 ppm Standard Impacts on U.S. Domestic Output
Quantity, 2020

Source:	EMPAX-CGE

Fig 5.2 Ozone 0.070 ppm Standard Impacts on U.S. Domestic Gross Output
Revenues, 2020

Source:	EMPAX-CGE

Projected Impacts on Regional Energy-Intensive Industries

Regional effects will tend to show variation that does not appear at the
national level.  To examine how such variations might occur in response
to the ozone alternate standard; this analysis presents findings for an
East-West split of the United States.  Since changes in output for most
industries are essentially unaffected, Figures 5.3 and 5.4 focus on
regional results for the energy-intensive industries in EMPAX-CGE.  

As with the U.S. average results from Figure 5.4, even though the
energy-intensive sectors show more regional variation, based on
differences in production methods and changes in manufacturing costs,
the majority of the impacts are less than one tenth of one percent
(<0.10%).  However, there are measurable impacts in the output of
specific industries.  Under the 0.070 ozone Standard, energy-intensive
output tends to be redistributed slightly from East to West as decreases
in cement and glass manufacturing output in the East are offset by
increases in the West.  

Fig 5.3 Ozone 0.070 ppm Standard Impacts on Regional Energy-Intensive
Output Quantities, 2020

Source:	EMPAX-CGE

Expressing the findings in dollar terms helps normalize the changes
across the country by implicitly including a measure of the absolute
size of each industry in each region of the country.  Thus, the decrease
in the quantity of cement shown in the East in Figure 5.3 is larger than
a relative increase in the West, leading to an overall decline for the
U.S.  Figure 5.4 expresses these changes in dollar, or gross revenue,
terms, which help offset these differences in industry size across
regions (however, the revenue terms also combine quantity and price
effects as discussed above).  In gross revenue terms, the changes in
Figure 5.4 show somewhat less variation than the quantity changes in
Figure 5.3.  For cement, gross revenues are somewhat higher in both the
East and West, while output quantities have declined slightly in the
East as the result of changes in production costs.  The additional
production in the West leads to additional revenues as some production
moves around the nation, although this effect is reduced for cement in
particular as its interregional trade is limited by transportation
costs.  

Fig 5.4 Ozone 0.070 ppm Standard Impacts on Regional Energy-Intensive
Gross Output Revenues, 2020

Source:	EMPAX-CGE

Similar to the energy-intensive sectors, energy production shows more
regional variation than can be seen in the U.S. results in Figures 5.1
and 5.2.  However, all impacts are less than one half of one percent
(<0.50%).  Under the 0.070 ozone Standard, coal consumption by electric
utilities tends to increase slightly in 2020, while natural gas use
falls, leading to comparable adjustments in regional production.  The
crude oil and petroleum refining industries react to the alternative
standard by minor reductions in output, although refining in the West
rises since it has a small comparative advantage over the East as fewer
refiners need to install additional controls.  In gross revenue terms,
the changes in Figure 5.6 show somewhat less variation than the quantity
changes in Figure 5.5.  Coal revenues rise with the increase in
production, and natural gas falls.  Along with adjustments in the
production technologies used in electricity (based on the IPM findings),
which affect generation costs and output levels, overall demand for
electricity rises as businesses purchase additional electricity to run
the controls installed to meet the ozone alternative standard.  This
leads also leads to an increase in electricity output revenues.  

 

 Fig 5.5 Ozone 0.070 ppm Standard Impacts on Regional Energy Output
Quantities, 2020

Source:	EMPAX-CGE

Fig 5.6 Ozone 0.070 ppm Standard Impacts on Regional Energy Gross Output
Revenues, 2020

Source:	EMPAX-CGE

When examining such findings, however, it is important to note that
these impacts and redistributions are directly related to the specific
control option assumed in this illustrative analysis.  As previously
stated, these results represent the impact of an approach presented by
EPA that could make progress towards attainment under the alternate
standard.  While EPA is providing this analysis as guidance for States,
it is expected that States will evaluate the best strategies for
achieving compliance and may choose options that could significantly
alter these regional effects.  Therefore, SIPs will most likely be
different than the strategy developed in this RIA and could be designed
to alleviate any disproportionate impacts on sensitive industries.  For
example, given the impact on glass and cement production, assumed with
this scenario, affected States may well design SIP strategies that
mitigate the impact on these particular industries, perhaps distributing
costs more uniformly among all sectors.  

Projected Impacts on GDP

The combination of economic interactions affecting business and
household behavior will be reflected in the changes in GDP estimated by
a CGE model.  Given that this cost-based approach to analyzing the ozone
0.070 standard does not reflect its benefits to the environment, public
health, and labor productivity, CGE models (including EMPAX-CGE) will
tend to over estimate declines in total production in the United States.
 Potentially offsetting these benefits are attainments costs that have
not been included in this analysis mainly due to their lack of direct
industry cost information. Consequently, these results can be considered
incomplete because they do not reflect potential productivity benefits
of the ozone Standard or the full cost of attainment.  The impacts on
GDP should be viewed as an approximation of the social costs of the
modeled controls applied for the alternate standard and are provided
here for illustration.  

Figure 5.7 illustrates GDP in the EMPAX-CGE model’s baseline forecast
and the 0.070 ozone policy case.  As shown, the estimated GDP impact is
negligible and, in fact, it is not possible to adjust the scale of the
graph to the point where the two lines do not overlap.  Projected
decrease in GDP for the ozone 0.070 standard is roughly 0.04 percent
(0.04%), respectively, for the year 2020.  This is equivalent to a $7.1
billion decrease in GDP during the implementation year.  In absolute
terms, these estimated implications for U.S. GDP are extremely small
relative to the total size of the economy.  Even these small costs could
be reduced if the CGE analyses were extended to include benefits
associated with the ozone standard such as improvements in labor
productivity from environmental improvements. I do not think this is
correct if labor mobility is assumed.

Fig 5.7 Change in U.S. GDP Compared to EMPAX-CGE Baseline

Source:	Department of Energy, Energy Information Administrations;
EMPAX-CGE

.  Furthermore, the model does not incorporate productivity benefits
resulting from air quality improvements.  Therefore, as a result of
these two potentially offsetting conditions, it is difficult for EPA to
determine if the results presented here overstate or understate the
impacts on industry output and U.S. GDP.   

5.3 Extrapolated Costs

This section presents the results and methodology behind the
extrapolated cost calculations of reaching full attainment of the 0.075,
0.070, and 0.065 ppm ozone alternate standards.  As discussed in Chapter
3, the application of 0.070 ppm control strategies was not successful in
reaching nationwide full attainment of the alternate ozone standards. 
Many areas remained in non-attainment in all three alternate standards;
therefore the engineering costs detailed in Section 5.1 represent only
the costs of partial attainment.  

It is important to emphasize the challenge EPA faced in estimating costs
of controls that do not yet exist.  The estimation of the costs of
unidentified controls needed to reach attainment is inherently a
difficult issue.  In many cases, the feasibility of a region reaching
attainment of these standards is questionable for a number of reasons,
including transboundary ozone contributions, background ozone levels in
higher elevations, and also due to our experience with the difficulty of
some regions attaining the current standard of 0.084 ppm. The degree to
which unspecified controls are needed to achieve attainment depends upon
other variables in the analysis, such as attainment date assumptions. 
We will better understand the true scope of the issue in the future as
states do the detailed area-by-area analysis to determine available
controls and attainment dates that are appropriate under the Clean Air
Act, which we do not attempt to determine in this analysis.

In this draft RIA we use three different approaches to estimating the
costs of unspecified control measures.  This reflects the difficulty in
defining a “best” approach to this issue.  Our approaches have yet
to be peer reviewed. Our approaches reflect a range of views about the
likely cost of future technologies and strategies that reduce air
pollutant emissions.  [The higher-cost estimation approaches are
implicitly more pessimistic about prospects for  technological advances
that avoid large increases in the cost per ton of emission reduction
relative to controls employed in the past.]  A separate section
discusses historical experience which has shown numerous technological
advances in emission reduction technologies, and provides a few examples
of today’s emerging technologies.  (See Section 5.5.)  We will
continue to consider these issues between now and the publication of the
final RIA for the final ozone NAAQS rule.

Due to the level of uncertainty related to the extrapolated costs, two
approaches were applied.  This section provides the additional costs of
reaching nationwide full attainment of the alternate ozone standards
utilizing three approaches: a lower bound fixed cost per ton approach
based on where the majority of the cost of the known control measures
where found, an upper bound fixed cost per ton approach based on the
cost of the last few known control measures used and an increasing
marginal cost approach similar to that used in the PM NAAQS RIA.   Prior
to presenting the aforementioned full attainment costs, a detailed
description of the methodology of each approach is provided.

5.3.1 Increasing Marginal Cost Methodology

This approach stems from the assumption that each unit of incremental
reduction in non-attainment areas will result in an increase in cost per
ton or marginal cost of abatement.  Therefore, EPA estimated constantly
increasing marginal cost curves for emission reductions using cost per
ton values from control strategy data in representative non-attainment
areas.  These curves were then used to estimate a cost of full
attainment using the emission reduction targets detailed in Chapter 4 of
this report.  

5.3.1.1 Marginal Cost Regions

EPA grouped the non-attainment areas described in Chapter 4 along with
their emission reduction targets into six regions of the country (Table
5.2) in order to acquire sufficient and representative data for deriving
the slopes of the marginal cost curves.  As a way of maintaining some
consistency with the modeled controls and the economic impact analysis,
the six regions were loosely based on EMPAX-CGE regions with a few
exceptions.  

Nonattainment areas in Virginia were grouped with the Northeast due to
the fact that Northern Virginia is part of the Ozone Transport Region
(OTR) which makes up the Northeast.  Resources were not available to
disaggregate states by counties. 

Nonattainment areas in Louisiana were grouped with Texas and Oklahoma
(Plains region) due to the similarity in industry mix among those
states.

California was separated from the rest of the west due to the severity
of the ozone problem in the state, the glide path targets unique to the
state, and because EPA determined the rest of the west was not an ideal
representation of California.    

Table 5.2  Regions and Slopes for Extrapolated Costs

Region	Marginal Cost Slope

Northeast (OTR)	0.035

Midwest	0.045

Southeast	0.036

Plains (TX/LA)	0.033

West (Not CA)	0.152

CA	0.211



5.3.1.2 Derivation of the Marginal Cost Slopes

Due to the efficaciousness and efficiency of NOx controls compared to
VOC controls, control strategy cost per ton data was acquired for each
region using a selection criteria defined in Table 5.3 and applied in
Ordinary Least Squares regression equations.  Results of these equations
provided the slope for the marginal cost curves.  For each equation, the
dependant variable (Y = cost/ton) and was regressed conditional to (X =
cumulative emissions reductions).  

 = c + βX + е   

			c = constant

			β = slope

			е = residual

The intent of the regression equations was not necessarily to accurately
capture the relation between cost/ton and cumulative reductions but
instead it was to identify a slope that would provide a rough
approximation of an increasing cost/ton rate as related to cumulative
emission reductions.  This slope would then provide an increasing
cost/ton rate in the extrapolated portion of the costs that was
equivalent to the rate observed under the modeled costs.  

Table 5.3  Data Selection Criteria for Extrapolated Costs

1)  Determine if area has sufficient NOx emissions remaining to reach   
attainment

2)  If area has sufficient NOx remaining to reach attainment, then use
NOx cost/ton data due to their cost effectiveness compared VOC controls

3)  If area does not have sufficient NOx emissions to reach attainment,
then include VOC controls in the data set if:

VOC controls were part of the control strategy for the area in question

VOC controls would significantly alter the value of a NOx based slope
for the marginal cost curve

Note: Data analysis demonstrated that VOC control data would only be
needed for California.  Due to 

lack of available ozone data, NOx controls from California also include
control cost from the PM NAAQS 

RIA control strategies. 

5.3.1.3 Calculating Extrapolated Costs Using Marginal Cost Approach

Once the slope of the marginal cost curve was derived, the extrapolation
was calculated by multiplying that slope with the emission reduction
target and adding that value to the highest of the observed cost/ton
value (Figure 5.8).  For this illustrative analysis, the highest of the
observed cost/ton values was roughly $15,267/ton which represented the
intercept of the marginal cost equation.  Total costs could then be
estimated by adding the area under the marginal cost curve in Figure 5.8
or by taking the integral of the marginal cost function and inputting
the emission reduction target into the equation for total cost.  

 fix this figure

Fixed Cost per Ton Values

Similar to the 1997 Ozone NAAQS RIA, a fixed cost/ton value was also
applied to estimate the extrapolated costs of nationwide full
attainment.  Total costs for each non-attainment area was calculated by
multiplying the fixed cost/ton value with the emission reduction targets
for each region.  For this particular illustrative analysis, a pair of
fixed cost/ton values was used to calculate costs.  

Fixed Cost/Ton Methodology

NOx control strategy data for the East and West were examined for
‘clustering’ within their individual distributions.  Cost/ton data
for the east and west were stratified into thousands (Ex. $0-$1000,
$1000-$2000,…) with individual source counts aggregated within each
interval.  California was separated from the west so source cost/ton
counts were conducted separately for the state.  This was the result of
limited ozone NOx data availability for the state, the low number of NOx
emissions remaining for CA, and the inclusion of VOC controls as well as
NOx controls from PM NAAQS RIA control strategies which were required to
resolve these data and emissions issues.  

For the East, 90% of the controls were below $6,000/ton.  As a result,
the control cost closest to $6,000/ton ($6,012) was selected to
represent to lower estimate of the Eastern fixed cost/ton approach.  For
the West, 94% of the controls had a cost/ton value below $4,000/ton. 
Therefore, $4,213 was selected as the lower estimate for the western
fixed cost/ton approach.  For California, the lower estimate was $9,035
using the same method but including VOC and PM NAAQS NOx control data
from the PM NAAQS RIA hypothetical control scenario. 

In addition to a lower fixed cost/ton estimate, an upper fixed cost/ton
value was used for calculating extrapolated costs.  This upper value as
estimated at $15,267 for all regions for the following reasons.

This value represented the highest, in terms of cost/ton, of the
controls applied in the east which made up the majority of the modeled
controls for the ozone standard.

In the case of the West, the next highest controls were roughly $35,000
and $39,000 per ton.  Controls with these costs were determined to be
significantly less feasible to implement compared other controls.

This value provides a consistent platform from which to incorporate and
compare marginal cost values derived using the increasing marginal cost
approach. 

Results

Tables 5.4 to 5.6 provide the extrapolated cost values for full
attainment of the 0.075, 0.070, 0.065 ppm standards in each area
applying the increasing marginal cost approach as well as the two fixed
cost/ton values.   The reader should be aware of the following
stipulations prior to making inferences from the extrapolated costs
presented in the following tables.

The two extrapolated cost approaches provide nothing more than three
rough estimates of potential costs with the marginal cost approach
providing the highest value.  Neither result includes a probability or a
link to sectors where reductions will be attained.  Therefore, there are
no expected values within this range of outcomes and no assumptions made
about the types of controls that would be applied in 2020. 

0.070 ppm extrapolated costs were estimated using data from the 0.070
ppm control strategy.  Therefore, although the degree of uncertainty is
still significant, these results can be expected to have a higher level
of confidence than results of the 0.075 and 0.065 ppm alternate
standards.

The use of the 0.070 ppm control strategy as a starting point for
extrapolating the 0.075 ppm standard resulted in over attainment of
0.075 ppm in some areas.  For over attaining areas, cost savings and
emission increases were extrapolated using the impact/ton estimates
derived in Chapter 4 and their appropriate emission targets until
reaching the 0.075 ppm standard.

.  Most of these counties were in states within the 0.070 ppm control
strategy region described in Chapter 3.  Therefore, no additional
controls were available and costs had to be extrapolated using the same
impact/ton estimates applied in the 0.070 ppm estimates.  Two new states
were added to the non-attainment region (KS and AL).  Since controls
were available for these states, AirControlNET was used to identify
controls that would achieve the required emission reduction targets.

  

Table 5.4  Extrapolated Costs of Meeting the 0.075 ppm Standard ($ M)

Extrapolated Costs for 075 Standard	MC Curve Estimate	Lower Fixed
Cost/Ton Estimate	Upper Fixed Cost/Ton Estimate*

Extrapolated Costs	 	 	 

CA – Los Angeles**	$0 	$0 	$0 

CA – Kern County**	$0 	$0 	$0 

Houston / Dallas	$1,254 	$400 	$1,008 

Ozone Transport Region	$1,307 	$443 	$1,114 

Lake Michigan region	$1,310 	$449 	$1,130 

Richmond / Norfolk	$790 	$297 	$748 

Detroit	$396 	$152 	$382 

Phoenix	$282 	$72 	$260 

Denver	$160 	$42 	$153 

Cleveland/Columbus/Cincinnati	$92 	$36 	$92 

Atlanta	$15 	$6 	$15 

 	 	 	 

Total Cost	$5,606 	$1,896 	 

 	 	 	 

Extrapolated Cost Savi+ngs	 	 	 

Baton Rouge, LA	($225)	($91)	 

Indianapolis, IN	($209)	($85)	 

Louisville, KY-IN	($284)	($115)	 

St. Louis, MO-IL	($30)	($12)	 

 	 	 	 

Total Cost Savings	($748)	($303)	 

 	 	 	 

Total Extrapolated Cost	$4,858 	$1,593 	 

*Due to the limited amount of controls in the modeled control strategy
which had this value, deducting this amount would likely result in an
over estimate of the savings.

** Los Angeles and Kern Counties have expected attainment dates after
2020.  This analysis counts the portion of reductions expected by 2020
or earlier.  



Table 5.5 Extrapolated Costs of Meeting the 0.070 ppm Standard ($ M) 

Extrapolated Costs for 0.070 Standard	MC Curve Estimate	Lower Fixed
Cost/Ton Estimate	Upper Fixed Cost/Ton Estimate

CA – Los Angeles **	$0 	$0 	$0 

CA – Kern County**	$823 	$181 	$305 

Houston / Dallas	$2,299 	$703 	$1,771 

Ozone Transport Region	$2,310 	$746 	$1,878 

Lake Michigan region	$2,334 	$752 	$1,893 

Richmond / Norfolk	$1,683 	$600 	$1,511 

Detroit	$1,272 	$455 	$1,145 

Phoenix	$1,364 	$282 	$1,023 

Denver	$1,190 	$253 	$916 

Cleveland/Columbus/Cincinnati	$926 	$339 	$855 

Atlanta	$825 	$309 	$779 

St. Louis	$785 	$291 	$733 

Indianapolis	$579 	$218 	$550 

Baton Rouge	$555 	$212 	$534 

Louisville	$491 	$188 	$473 

Memphis	$313 	$121 	$305 

Charlotte	$217 	$85 	$214 

Salt Lake City	$211 	$55 	$198 

Las Vegas	$177 	$46 	$168 

Tampa	$77 	$30 	$76 

Total Extrapolated Cost	$18,435 	$5,867 	$15,328 



**Los Angeles and Kern Counties have expected attainment dates after
2020.  This analysis counts the portion of reductions expected by 2020
or earlier.  



Table 5.6  Extrapolated Costs of Meeting the 0.065 ppm Standard ($ M)

Extrapolated Costs for 065 Standard	 	 	 

 	MC Curve Estimate	Lower Fixed Cost/Ton Estimate	Upper Fixed Cost/Ton
Estimate

CA – Los Angeles**	$0 	$0 	$0 

CA – Kern County**	$2,230 	$452 	$763 

Houston / Dallas	$3,427 	$1,006 	$2,534 

Ozone Transport Region	$3,401 	$1,049 	$2,641 

Lake Michigan region	$3,471 	$1,055 	$2,656 

Richmond / Norfolk	$2,663 	$903 	$2,275 

Detroit	$2,260 	$758 	$1,908 

Phoenix	$2,827 	$493 	$1,786 

Denver	$2,599 	$463 	$1,679 

Cleveland/Columbus/Cincinnati	$1,871 	$643 	$1,618 

Atlanta	$1,726 	$612 	$1,542 

St. Louis	$1,712 	$594 	$1,496 

Indianapolis	$1,479 	$521 	$1,313 

Baton Rouge	$1,417 	$515 	$1,298 

Louisville	$1,355 	$491 	$1,237 

Memphis	$1,157 	$424 	$1,069 

Charlotte	$1,051 	$388 	$977 

Salt Lake City	$1,263 	$265 	$962 

Las Vegas	$1,214 	$257 	$931 

Tampa	$894 	$333 	$840 

Jackson, MS	$757 	$285 	$718 

New Mexico areas (Farmington / Las Cruces)	$819 	$185 	$672 

OK areas (Tulsa, Marshall)	$704 	$267 	$672 

Huntington, WV-KY	$639 	$242 	$611 

El Paso, TX	$538 	$206 	$519 

Kansas City, MO/KS	$325 	$142 	$317 

Little Rock, AR	$442 	$170 	$427 

Mobile AL	$70 	$70 	$70 

Columbia, SC	$154 	$61 	$153 

 	 	 	 

Extrapolated Total	$42,465 	$12,851 	$33,684 

**Los Angeles and Kern Counties have expected attainment dates after
2020.  This analysis counts the portion of reductions expected by 2020
or earlier.  



5.4 Summary of costs

Table 5.7 presents a summary of the total cost of attaining 0.075,
0.070, and 0.065 ppm standards.  This summary includes the costs
presented above from the modeled controls, the economic impact analysis
and the extrapolated costs.  The range presented in the extrapolated
costs and the grand total costs indicate the upper and lower bound cost
estimates. 

Table 5.7 Total Costs of Attainment in 2020 for Different Levels of the
Ozone Standard

	Level of Standard in 2020 

	0.065 ppm 	0.070 ppm 	0.075 ppm 

Modeled Social Costs* ($B)	$7.1	$7.1	$7.1

Extrapolated Costs ($B)	$13 to $42	$5.9 to $18	$1.6 to $4.9

Grand Total Costs ($B)	$20 to $50	$13 to $26	$8.5 to $12

*Modeled social costs includes both the modeled costs presented in table
5.1, as well as the economic impact costs presented in section 5.2.

We should point out that the modeled social costs of the known and
supplemental controls was approximately $4 billion, which resulted in
Modeled social costs of approximately $7.1 billion. Because the
extrapolated costs shown here are much greater than $4 billion in most
cases, it is likely that the modeled social costs including extrapolated
costs could be as much as $10 billion to $16 billion annually for a
0.075 ppm standard; $18 billion to $39 billion annually for a standard
of 0.070 ppm; or as high as $30 billion to $81 billion annually for a
standard of 0.065. {plus California}.  This rests on the assumption that
extrapolated costs would be distributed similarly to known control costs
across different economic sectors.  However, we do not believe that this
would be the case, as the majority of extrapolated costs would likely
fall on the mobile source sector.  Because we do not know how these
costs will be distributed across regions and sectors, we do not model
them with EMPAX.(more discussion)

5.5 Technology Innovation and Regulatory Cost Estimates

During the history of the Clean Air Act, technological innovation and
“learning by doing” often have made it possible to achieve greater
emissions reductions that once were thought infeasible, or have reduced
the costs of emission control in relation to original estimates. 
Innovative companies have successfully responded to the regulatory
challenges and market opportunities provided by the Act, producing
breakthrough technologies for multiple sectors.  Studies have suggested
that costs of a number of EPA programs have been less than originally
estimated due in part to inadequate inability to predict and account for
future technological innovation in regulatory impact analyses.

Although we have not seen this occur in California and other ozone
nonattainment areas for the current 0.084 standard. Add discussion of
feasibility of achieving current standard here.

A constantly increasing abatement cost curve similar to the one utilized
for estimating extrapolated costs in this RIA is likely to induce the
type of innovation that would result in lower costs than estimated early
in this chapter.  Breakthrough technologies in control equipment would
result in an outward shift in the marginal cost curve, reducing marginal
costs, and thus deviate from the assumption of one constantly increasing
marginal cost curve.  In addition, elevated abatement costs may result
in significant increases in the cost of production and would likely
induce production efficiencies, in particular those related to energy
inputs, which would lower emissions from the production side.   

Examples of Technological Advances in Pollution Control

There are numerous examples of low-emission technologies developed
and/or commercialized over the past 15 or 20 years, such as:

Selective catalytic reduction (SCR) and ultra-low NOx burners for NOx
emissions

Scrubbers which achieve 95% and even greater SO2 control on boilers

Sophisticated new valve seals and leak detection equipment for
refineries and chemical plans

Low or zero VOC paints, consumer products and cleaning processes

CFC-free air conditioners, refrigerators, and solvents

Water and power-based coatings to replace petroleum-based formulations

Vehicles far cleaner than believed possible in the late 1980s due to
improvements in evaporative controls, catalyst design and fuel control
systems for light-duty vehicles; and treatment devices and retrofit
technologies for heavy-duty engines

Continued development of activated carbon injection (ACI) technology for
control of mercury from electric generating units

Development of integrated gasification combined cycle (IGCC) and
ultra-super critical pulverized coal technologies for electricity
generation

Idle-reduction technologies for engines, including truck stop
electrification efforts

Market penetration of gas-electric hybrid vehicles, biodiesel and other
clean fuels

Influence on Regulatory Cost Estimates

Studies indicate that it is not uncommon for pre-regulatory cost
estimates to be higher than later estimates, in part because of
inability to predict technological advances. 

Multi-rule study:  Harrington et al. of Resources for the Future
conducted an analysis of the costs of 25 EPA and OSHA rules in 2000, and
found a tendency for predicted costs to overstate actual implementation
costs.  Costs were overpredicted in 12 cases and underpredicted in 6
cases, with overestimates occurring more frequently in the larger rules.
 Difficulty of predicting technological innovation was one of several
reasons cited for the overestimates. 

Acid Rain SO2 Trading Program:  Recent cost estimates of the Acid Rain
SO2 trading program by Resources for the Future (RFF) and MIT have been
as much as 83 percent lower than originally projected by EPA.  Note that
the original EPA cost analysis also relied on an optimization model like
IPM. The ex ante numbers in 1989 were an overestimate in part because of
unforeseen technological improvements.

Phase 2 Cost Estimates

Ex ante estimates	$2.7 to $6.2 billion

Ex post estimates	$1.0 to $1.4 billion



EPA Fuel Control Rules:  A 2002 study of cost estimates for EPA vehicle
and fuels rules conducted by OTAQ found a general pattern that “all ex
ante estimates tended to exceed actual price impacts, with the EPA
estimates exceeding actual prices by the smallest amount.” An example
focusing on fuel rules is provided:

Table 5.8 Comparison of Inflation-Adjusted Estimated Costs and Actual
Price Changes

for EPA Fuel Control Rules

 	Inflation-adjusted Cost Estimates (c/gal)	Actual Price Changes
(c/gal)

	EPA	DOE	API	Other	 

Gasoline	 	 	 	 	 

Phase 2 RVP Control (7.8 RVP - Summer) (1995$)	1.1	 	1.8	 	0.5

Reformulated Gasoline Phase 1 (1997$)	3.1-5.1	3.4-4.1	8.2-14.0	7.4 (CRA)
2.2

Reformulated Gasoline Phase 2 (Summer) (2000$)	4.6-6.8	7.6-10.2
10.8-19.4	12	7.2 (5.1, when corrected to 5yr MTBE price)

30 ppm sulfur gasoline (Tier 2) 	1.7-1.9	2.9-3.4	2.6	5.7 (NPRA), 3.1
(AIAM)	N/A

Diesel	 	 	 	 	 

500 ppm sulfur highway diesel fuel (1997$)	1.9-2.4	 	 	3.3 (NPRA)	2.2

15 ppm sulfur highway diesel fuel	4.5	4.2-6.0	6.2	4.2-6.1 (NPRA)	N/A



CFC Phase-Out:    SEQ CHAPTER \h \r 1 EPA used a combination of
regulatory, market based (i.e., a cap-and-trade system among
manufacturers), and voluntary approaches to phase out the most harmful
ozone depleting substances.  This was done more efficiently than either
EPA or industry originally anticipated.  The phaseout for Class I
substances was implemented 4-6 years faster, included 13 more chemicals,
and cost 30 percent less than was predicted at the time the 1990 Clean
Air Act Amendments were enacted.

Comment:  Spell out and identify the supplement controls, as this would
define what control technologies are being analyzed.

Page 1,Synopsis:

As noted in Chapter 3, EPA first modeled a hypothetical control strategy
aimed at attaining a tighter standard of 0.070 ppm in 2020.

Comment:  This reads as if EPA conducted its first model in 2020. 
Reword as ‘Initially, EPA modeled…..for 2020.’

Page 5, 5.1.3, Limitations and Uncertainties Associated with Engineering
Cost Estimates:

EPA bases its estimates of emissions control costs on the best available
information from engineering studies of air pollution controls and has
developed a reliable modeling framework for analyzing the cost,
emissions changes, and other impacts of regulatory controls.  

Comment:  Identify the air pollution controls that are analyzed.

Page 11, Projected Impacts on Regional Energy-Intensive Industries:

…this analysis presents findings for an East-West split of the United
States

Comment:  Again, if there is a way to incorporate at North-South split
(i.e., four quadrants) this would provide greater depth of analysis.  

Page 26, Examples of Technological Advances in Pollution Control: 

Comment:  This listing of air pollution control technologies needs to be
presented in Chapter 1, where ‘emissions controls scenarios’ are
presented.

Question:  On page 3: “The marginal cost curve was analyzed, and there
was a clear break in the curve at approximately $6,000/ton.”  Does
“clear break” mean the marginal cost curve was discontinuous or
kinked?

Question:  On page 4 the reference to Appendix 6 – is that supposed to
be a reference to the appendix to chapter 5?

Comment: On table 5.1 EPA refers to annualized costs.  MAS assumes this
means that the costs for the summer months were extrapolated to the
whole year, but this is not made clear in the text, please clarify. 
Also, A-4 interest rates are usually used for making present discounted
value calculations, for example for estimating the present discounted
value of complying with the rule over the next ten years.  Obviously
this is not the case in this exercise.  To avoid confusion, EPA should
clarify in what context they are using A-4 interest rates.

Comment:  The first sentence of section 5.2 is very awkwardly phrased
and very difficult to understand.  

Comment:  In bullet (b) on page 6, the exclusion of extrapolated costs
from estimates of economic impacts invalidates the entire exercise.  EPA
needs to find a way to include them.  They claim that “Since the
extrapolated cost estimates are not associated with specific controls on
specific sectors it is not possible to estimate the economic impacts of
this portion of the cost estimate.”  That is simply not true.  While
it would be helpful to know what sectors bear the burdens, and even more
helpful to know what specific controls would be used, it is by no means
impossible to carry out a reasonably defensible estimate of economic
impacts with an overall estimate of the extrapolated costs.  EPA knows
that lowering ozone standards will directly affect certain manufacturing
sectors that emit ozone precursors, electricity generating units, and
mobile transportation.  Presumably investing in new technology to curb
emissions in these sectors will involve increased capital investment
that does not increase the industry’s output.  Therefore an increase
in extrapolated costs can be modeled as increasing the capital
requirements of the industries for a given unit of output.  The amount
of increased capital requirement for the manufacturing industry can be
set, for example, equal to the total estimated extrapolated cost
multiplied by manufacturing’s share in total value added of the three
affected sectors.

Comment:  On page 7, in bullet (d), EPA speculates that including the
increase in worker productivity resulting from lower ozone exposure
could possibly offset the small decrease in output.  This is extremely
unlikely.  Reductions in ozone exposure will not be limited to the
immediate environment surrounding manufacturing facilities that emit
ozone precursors – the entire surrounding urban area will be affected.
 While it is possible the overall increase in worker output nationwide
will, at the macro level, offset the decrease in output of the affected
industries, the production declines of the affected industries relative
to the rest of the economy would still hold.

Comment:  Throughout the chapter, the effects on the economy are
characterized as “small” compared to the overall level of economic
activity.  This is a counterfactual analysis – the analysts construct
a snapshot of the base scenario and the policy scenario and compare the
changes between the two – and the appropriate comparisons are
therefore on the margin.  How do the changes in output compare to
historic output growth?  EPA has used the same framework before, most
recently in the PM NAAQS analysis.  How do the outcomes of the analysis
compare to these other studies?

Comment:  EMPAX is not a general equilibrium model because it ignores
international effects.  If the United States imposes strict ozone
precursor emission requirements on U.S. manufacturing firms only, one of
their options is to move operations abroad where they will not face such
requirements.  Presumably all firms have already taken environmental
compliance costs into account when making their location decisions. 
Unilaterally imposing additional costs will then cause manufacturing
facilities that are on the margin to change their location decisions. 
Excluding the extrapolated costs from the economic impact analysis is
particularly problematic if international effects are incorporated.

Comment:  EPA goes to great lengths to explain that technological
advances have occurred in the past and their compliance costs may
therefore be overestimated.  Nevertheless, it would still be useful to
provide a scenario where no additional technological advances occur, at
least to provide an upper bound for compliance cost scenarios.  If there
are no technological solutions to lowering ozone precursor emissions,
then the only alternative is to lower economic activity that results in
emissions.  On mobile sources, appropriate “control technologies”
would then include taxes on fuel that are high enough to discourage
transportation.  For EGU’s, taxes on electricity consumption, and on
manufacturing, output taxes.  It would be very informative to see how
much economic activity would have to be reduced to achieve the ozone
targets set by EPA, assuming no technological advance, particularly if
the costs of this approach are less than estimates of extrapolated
costs.

Comment:  Page 18, section 5.3.1.2.  It would be helpful to have access
to the data used for these regressions.  During our telephone
conversation the EPA expert who carried out the regressions claimed that
the estimated elasticity for the pooled data was very small.  I would
like to see the data and method used for the two estimates of marginal
costs, because I suspect some sort of econometric error may have
occurred.

Comment: (Chapter 5) Cost Estimates:  Has any consideration been given
to the impact on competitiveness of attaining the alternative more
stringent levels for the ozone primary standard, especially for
manufacturing companies?  U.S. manufacturers compete in a global economy
and the potential increased costs of complying with the more stringent
standard could make them significantly less competitive than some of
their foreign competitors.  In the short run, some U.S. manufacturers
could be driven out of business by foreign competition or forced to
relocate their plants to other countries.  The economic impact on local
economics of a plant shut down could far exceed the benefits accrued
from cleaner air.  The potential cumulative economic impact on the U.S.
economy, of a large number of plants or companies shutting down, needs
to be factored into the calculations.   In the long run technological
innovations could significantly reduce the cost of attaining the
alternative more stringent levels for the ozone primary standard, but it
will not help companies which in the short run are forced out of
business. 

It is very reasonable to believe that in the long run technological
advances in pollution control would significantly reduce the costs of
complying with alternative more stringent levels for the ozone primary
standard.   Unfortunately technological innovation have research and
development costs associated with it.  But if all parties were forced to
invest in technological innovations as a means of complying with the
standard; individual companies, cities and states would not be put at a
significant disadvantage verses their competitors by having to invest in
measures to comply with the standard.

 

Question:  Would the costs of complying with a more stringent standard
have such an adverse impact on U.S. cement and glass manufacturers so as
to force them out of business as a consequence of increased competition
from foreign manufacturers?

Question:  What higher costs of production would manufacturers possibly
incur that would impact their competitiveness vis-à-vis foreign
manufacturers?

References 

Anderson, J.F., and Sherwood, T., 2002. “Comparison of EPA and Other
Estimates of Mobile Source Rule Costs to Actual Price Changes,” U.S.
Environmental Protection Agency.

Babiker, M.H., and T.F. Rutherford.  1997.  “Input Output and General
Equilibrium Estimates of Embodied CO2:  A Data Set and Static Framework
for Assessment.”  University of Colorado at Boulder, Working Paper 97
2.  Available at   HYPERLINK
"http://debreu.colorado.edu/papers/gtaptext.html" 
http://debreu.colorado.edu/papers/gtaptext.html .

Babiker, M.H., J.M. Reilly, M. Mayer, R.S. Eckaus, I.S. Wing, and R.C.
Hyman.  2001.  “The MIT Emissions Prediction and CO2 Policy Analysis
(EPPA) Model:  Revisions, Sensitivities, and Comparisons of Results.” 
MIT Joint Program on the Science and Policy of Global Change, Report No.
71.  Available at   HYPERLINK
"http://web.mit.edu/globalchange/www/eppa.html" 
http://web.mit.edu/globalchange/www/eppa.html .

Bovenberg, L.A., and L.H. Goulder.  1996.  “Optimal Environmental
Taxation in the Presence of Other Taxes:  General Equilibrium
Analysis.”  American Economic Review 86(4):985-1000.  Available at
<http://www.aeaweb.org/aer/>.

Brooke, A., D. Kendrick, A. Meeraus, and R. Raman.  1998.  GAMS:  A
User’s Guide.  GAMS Development Corporation.  Available at   HYPERLINK
"http://www.gams.com"  http://www.gams.com .

Carlson, Curtis, Dallas R. Burtraw, Maureen, Cropper, and Karen L.
Palmer. 2000. “SulfurDioxide Control by Electric Utilities: What Are
the Gains from Trade?” Journal of Political Economy 108(#6):1292-1326.

Ellerman, Denny. January 2003. Ex Post Evaluation of Tradable Permits:
The U.S. SO2

Cap-and-Trade Program. Massachusetts Institute of Technology Center for
Energy and Environmental Policy Research.

Feenberg, D., and E. Coutts.  1993.  “An Introduction to the TAXSIM
Model.”  Journal of Policy Analysis and Management 12(1):189 194. 
Available at http://www.nber.org/~taxsim/.

Fullerton, D., and D. Rogers.  1993.  “Who Bears the Lifetime Tax
Burden?”  Washington, DC:  The Brookings Institute.  Available at
http://bookstore.brookings.edu/ book_details.asp?product%5Fid=10403.

Goulder, L.H., and R.C. Williams.  2003.  “The Substantial Bias from
Ignoring General Equilibrium Effects in Estimating Excess Burden, and a
Practical Solution.”  Journal of Political Economy 111:898 927.
Available at   HYPERLINK
"http://www.journals.uchicago.edu/JPE/home.html" 
http://www.journals.uchicago.edu/JPE/home.html 

Harrington, W., R.D. Morgenstern, and P. Nelson. 2000. “On the
Accuracy of Regulatory Cost Estimates.” Journal of Policy Analysis and
Management 19(2):297-322.

Minnesota IMPLAN Group.  2003.  State Level Data for 2000.  Available
from   HYPERLINK "http://www.implan.com/index.html" 
http://www.implan.com/index.html .

Nestor, D.V., and C.A. Pasurka.  1995.  The U.S. Environmental
Protection Industry:  A Proposed Framework for Assessment. U.S.
Environmental Protection Agency, Office of Policy, Planning, and
Evaluation.  EPA 230-R-95-001.  Available at
http://yosemite.epa.gov/ee/epa/eermfile.nsf/11f680ff78df42f585256b45007e
6235/41b8b642ab9371df852564500004b543/$FILE/EE 0217A 1.pdf. 

U.S. Department of Energy, Energy Information Administration.  Undated
(b).  State Energy Price and Expenditure Report.  Washington DC. 
Available at http://www.eia.doe.gov/emeu/states/price_multistate.html. 

U.S. Department of Energy, Energy Information Administration.  2001. 
Manufacturing Energy Consumption Survey 1998.  Washington DC.  Available
at http://www.eia.doe.gov/emeu/mecs/. 

U.S. Department of Energy, Energy Information Administration.  January
2003.  Annual Energy Outlook 2003.  DOE/EIA 0383(2003).  Washington DC. 
Available at
http://www.eia.doe.gov/oiaf/archive/aeo03/pdf/0383(2003).pdf.

U.S. Department of Energy, Energy Information Administration.  January
2004.  Annual Energy Outlook 2004.  DOE/EIA-0383(2003).  Washington, DC.
 Available at http://www.eia.doe.gov/oiaf/aeo/pdf/0383 (2001).

 The less stringent alternative analyzed, which is the current standard,
or baseline, would not require additional costs and therefore costs for
that level are not presented in this RIA.

 For more information on this cost methodology and the role of
AirControlNext, see Section 6 of the 2006 PM RIA, AirControlNET 4.1
Control Measures Documentation (Pechan, 2006b), or   HYPERLINK
"http://www.epa.gov/ttn/catc/products.html#cccinfo" 
http://www.epa.gov/ttn/catc/products.html#cccinfo 

 Detailed information on IPM is available in Section 6 of the 2006 PM
RIA or at http://www.epa.gov/airmarkets/epa-ipm

 A different plant-specific interest rate is applied in estimating
control costs within IPM.  See PM RIA for details.

 All estimates provided reflect the cost of a control strategy for 0.070
ppm, incremental to a 2020 baseline of compliance with the current
standard of  0.084 ppm.  

    Total annualized costs are 3% are calculated for controls where
there is a capital component.

 See Appendix E in the RIA for the Final CAIR rule for additional
discussion of these IPM-EMPAX linkages (  HYPERLINK
"http://www.epa.gov/interstateairquality/technical.html" 
http://www.epa.gov/interstateairquality/technical.html ).

 For more detailed regional impact figures, please see the appendix for
this chapter.  See the CAIR and PM NAAQS analyses, and Ross et al.
(2005), for discussions of regional definitions.

 Redistribution of production will also tend to occur among states in
each region, with some states’ increasing 

output to offset any declines in neighboring states.

 For more information on EMPAX-CGE regions, see appendix NEED NUMBER. 
Data sets used to calculate the slopes of the marginal cost curve were
not based on EMPAX-CGE modeling.  The only similarity with EMPAX-CGE is
the regional breakdown.  Therefore, the extrapolated costs do not
represent social costs in any way nor are they linked to the CGE
baseline data, structure, or sector detail.  Linking extrapolated costs
to individual sectors is beyond the scope of this analysis.

 EPA recognizes that these regression equations may be misspecified.  As
stated above, the objective was not to accurately capture the relation
between control cost/ton and emission reduction for statistical or
economic inference purposes.  These equations represent the most
statistically adequate models that could be specified given the data,
time, and resource constraints. 

 Total Cost = $15,267x + (β/2)x2, where x = emission reduction target

 Harrington et al., 2000.

 Carlson et al., 2000; Ellerman, 2003.

 2010 Phase II cost estimate in $1995.

 Anderson et al., 2002.

 Holmstead, 2002.

 PAGE   

 PAGE   

 PAGE   9 

	

Make sure that you have discounted the benefit estimates in these
regions to 0.075 as well.

 PAGE \# "'Page: '#'

Is this an increasing MC function or constant?

 PAGE \# "'Page: '#'

In how many regions?

 PAGE \# "'Page: '#'

Please provide the split for known and supplemental

 PAGE \# "'Page: '#'

 PAGE \# "'Page: '#'

What are the resulting permit prices?

 PAGE \# "'Page: '#'

á

â

ᘀ䔂　ᙊ唀Ĉ

ñÿ\

	k

ñÿ\

	k

ñÿ\

	k

ñÿ\

	k

ñÿ\

	k

ñÿ\

	k

ñÿ\

	k

&

&

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

	,

kd

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

kd

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

ñÿô

栖˧E

阂l혈\鐄䃿鐋們ఙ!갆

阂l鐇á혈\鐄䃿鐋們ఙ!갆



Please discuss the importance of California in the economy and how
results might be different with California added.

 PAGE \# "'Page: '#'

What is the labor supply assumption in EMPAX-CGE?  Does it restrict
transboundary migration?  If not then the comments about improved labor
supply need to be better justified.

Please provide a reference to the value of benefits you have calculated
for labor productivity in the benefits chapter.

 PAGE \# "'Page: '#'

Please explain how this fits into your 200km assumption.

