Chapter 8:   Conclusions and Implications of the Illustrative
Benefit-Cost Analysis

Overview

EPA has performed an illustrative analysis to estimate the costs and
human health benefits of nationally attaining alternative ozone
standards. We have considered 4 alternative standards incremental to
attaining the current ozone standard:  0.079 ppm, 0.075 ppm, 0.070 ppm,
and 0.065 ppm. This chapter summarizes these results and discusses the
implications of the analysis. This analysis serves both to satisfy the
requirements of E.O. 12866 and to provide the public with an estimate of
the potential costs and benefits of attaining alternative ozone
standards. The benefit and cost estimates below are calculated
incremental to a 2020 baseline that incorporates air quality
improvements achieved through the projected implementation of existing
regulations and full attainment of the current standards for ozone and
PM NAAQS (including the hypothetical control strategy developed in the
RIA for full attainment of the PM NAAQS 15/35 promulgated in September,
2006).  This RIA presents two sets of results:  The first reflects full
attainment in all locations except two areas of California, which are
planning to meet the current standards after 2020, and so have estimated
costs and benefits for the analyzed standards for partial attainment in
2020 (their “glidepath” targets). The second estimate, for
California only, presents the additional costs and benefits that might
result from California fully attaining the standards in a year beyond
2020.  Finally, this chapter provides additional context for the RIA
analysis and a discussion of limitations and uncertainties.  In
addition, given the technological limitations associated with reducing
ozone precursors, we provide estimated cost and benefit numbers based on
both partial attainment (manageable with current technologies) and full
attainment (manageable in some locations only with hypothetical
technologies).

8.1	Results

Presentation of Results

There are two sets of results presented below.  The first set of results
is for 2020.  For analytical purposes explained previously, we assume
that almost all areas of the country will meet each alternative standard
in 2020 through the development of technologies at least as effective as
the hypothetical strategies used in this illustration.  It is expected
that benefits and costs will begin occurring earlier, as states begin
implementing control measures to attain earlier or to show progress
towards attainment. Some areas with very high levels of ozone do not
plan to meet even the current standard until after 2020; specifically,
two California areas have adopted plans for post-2020 attainment as
noted above.  In these locations, we provide estimates of the costs and
benefits of attaining a “glidepath” target in our 2020 analysis
year.   The 2020 results thus do not represent a complete “full
attainment” scenario for the entire nation, particularly for more
stringent alternative new standards examined.  In order to gain an
understanding of the possible additional costs and benefits of fully
attaining in California, we provide an additional set of results
focusing on California.  

By the year 2030, various mobile source rules, such as the onroad and
nonroad diesel rules, among others, would be expected to be fully
implemented.  Because California will likely not have to attain until
closer to 2030, it is important to reflect the impact those rules might
have on the emissions that affect ozone nonattainment.  To reflect the
emission reductions that are expected from these rules, we subtract
those tons from our estimates of the emissions reductions that might be
needed for California to fully attain in 2020, thus making our analysis
more consistent with full attainment later than 2020.  EPA did the
analysis this way because to force full attainment in California in an
earlier year would not be consistent with the CAA, and would likely lead
to an overstatement of costs because those areas might benefit from
these existing federal or state programs that would be implemented
between 2020 and the attainment year (see detail in Chapter 4); because
additional new technologies may become available between 2020 and the
attainment year; and/or the cost of existing technologies might fall
over time due to economic factors such as economies of scale or
improvements in the efficiency of installing and operating controls
(‘learning by doing’).  On the other hand, it is also possible that
new technologies might not meet the specifications, development time
lines, or cost estimates provided in this analysis.  

It is not appropriate to add together the 2020 national attainment,
California glidepath estimate and the estimate of California full
attainment as an estimate of national full attainment in 2020  It is not
appropriate to do this because each estimate is based on different
baseline conditions for emissions and air quality.  In addition, both
estimates include estimates of California glidepath results, leading to
the potential for double counting if added together. 

The following set of tables summarizes the costs and benefits of the
scenarios analyzed, and shows the net benefits for each of the scenarios
across a range of modeling assumptions concerning the calculation of
costs and benefits.  Tables 8.1a-c present benefits and costs of
national attainment in 2020, including the “glidepath” targets for
California. Companion Table 8.2 provides the estimated reductions in
premature mortality and morbidity for national attainment in 2020,
including the “glidepath” targets for California.  Tables 8.3a-c
present the additional costs and benefits of full attainment for
California (“glidepath” plus future year attainment added together
into one total); Table 8.4 is the companion table showing estimated
reductions in premature mortality and morbidity. 

The individual row estimates for benefits reflect the variability in the
functions available for estimating the largest source of benefits –
avoided ozone premature mortality.  Ranges within the total benefits
column reflect variability in the estimates of PM premature mortality
co-benefits across the available effect estimates.  Ranges in the total
costs column reflect different assumptions about the extrapolation of
costs.  The low end of the range of net benefits is constructed by
subtracting the highest cost from the lowest benefit, while the high end
of the range is constructed by subtracting the lowest cost from the
highest benefit.  Following these tables is a discussion of the
implications of these estimates, as well as the uncertainties and
limitations that should be considered in interpreting the estimates.



Table 8.1a  National Annual Costs and Benefits:  0.075 ppm Standard in
2020

(including California glidepath )

Premature Mortality Function or Assumption	Reference	Mean Total
Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$2.7 to $13	$5.5 to $8.8	-$6.2 to $7.2

Meta-analysis	Bell et al. 2005	$7 to $17	$5.5 to $8.8	-$1.9 to $12

	Ito et al. 2005	$7.5 to $18	$5.5 to $8.8	-$1.3 to $12

	Levy et al. 2005	$8.4 to $19	$5.5 to $8.8	-$0.4 to $13

Assumption that association is not causal***	$1.2 to $11	$5.5 to $8.8
-$7.6 to $5.8



Table 8.1b  National Annual Costs and Benefits:  0.070 ppm Standard in
2020

(including California glidepath)

Premature Mortality Function or Assumption	Reference	Mean Total
Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$4.1 to $24	$10 to $22	-$18 to $14

Meta-analysis	Bell et al. 2005	$9.5 to $29	$10 to $22	-$13 to $19

	Ito et al. 2005	$10 to $30	$10 to $22	-$12 to $20

	Levy et al. 2005	$11 to $31	$10 to $22	-$11 to $21

Assumption that association is not causal***	$2.3 to $22	$10 to $22	-$20
to $12



Table 8.1c  National Annual Costs and Benefits :  0.065 ppm Standard in
2020

(including California glidepath)

Premature Mortality Function or Assumption	Reference	Mean Total
Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$7.7 to $45	$17 to $46	-$38 to $28

Meta-analysis	Bell et al. 2005	$18 to $55	$17 to $46	-$28 to $38

	Ito et al. 2005	$19 to $56	$17 to $46	-$27 to $39

	Levy et al. 2005	$19 to $56	$17 to $46	-$27 to $39

Assumption that association is not causal***	$4.3 to $41	$17 to $46	-$42
to $24

*Includes ozone benefits, and PM 2.5 co-benefits

**Range was developed by adding the estimate from the ozone premature
mortality function to both the lower and upper ends of the range of the
PM2.5 premature mortality functions characterized in the expert
elicitation

***Range reflects lower and upper bound cost estimates

****Total includes ozone morbidity benefits only

Table 8.2  National Annual Number of Ozone and PM2.5-Related Premature
Mortalities and Premature Morbidity Avoided in 2020 (including
California glidepath)



Combined Estimate of Mortality

Standard Alternative and 

Model or Assumption*	Combined Range of Ozone Benefits and

 PM2.5 Co-Benefits



0.075 ppm 	0.070 ppm	0.065 ppm

NMMAPS 	Bell (2004)	390 to 2,100	650 to 4,000	1,200 to 7,600

Meta-Analysis	Bell (2005)	1,100 to 2,800	1,500 to 4,900	2,800 to 9,200

	Ito (2005)	1,200 to 2,900	1,600 to 5,000	3,000 to 9,400 

	Levy (2005)	1,300 to 3,000	1,800 to 5,100	3,100 to 9,400 

No Causality	190 to 1,900	370 to 4,000	690 to 7,000



	Combined Estimate of Morbidity



	Acute Myocardial Infarction	 1,100	  2,100	     3,900

Hospital and ER Visits	30,000	55,000	 100,000

Chronic Bronchitis	     380	     740	     1,400

Acute Bronchitis	  1,000	   1,900	     3,700

Asthma Exacerbation	  7,600	15,000	    28,000

Lower Respiratory Symptoms	   8,300	16,000	    30,000

Upper Respiratory Symptoms	   6,100	12,000	     22,000

School Loss Days	610,000	780,000	1,300,000

Work Loss Days	   53,000	100,000	   190,000

Minor Restricted Activity Days	2,000,000	2,700,000	4,700,000



	

***Range was developed by adding the estimate from the ozone premature
mortality function to both the lower and upper ends of the range of the
PM2.5 premature mortality functions characterized in the expert
elicitation



Table 8.3a California:  Annual Costs and Benefits of Attaining 0.070 ppm
Standard (beyond 2020)

Premature Mortality Function 

or Assumption	Reference	Mean Total Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$0.8 to $3.8	$2 to $13	-$12 to $1.8

Meta-analysis	Bell et al. 2005	$2 to $5	$2 to $13	-$11 to $3

	Ito et al. 2005	$2.1 to $5.1	$2 to $13	-$11 to $3.2

	Levy et al. 2005	$2.1 to $5.1	$2 to $13	-$11 to $3.2

Assumption that association is not causal***	$0.4 to $3.4	$2 to $13	-$13
to $1.4



Table 8.3b   California:  Annual Costs and Benefits of  Attaining 0.075
ppm Standard (beyond 2020)

Premature Mortality Function 

or Assumption	Reference	Mean Total Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$0.3 to $1.1	$1.1 to $6.2	-$5.9 to $0.04

Meta-analysis	Bell et al. 2005	$1 to $1.8	$1.1 to $6.2	-$5.2 to $0.7

	Ito et al. 2005	$1.1 to $1.9	$1.1 to $6.2	-$5.1 to $0.8 

	Levy et al. 2005	$1.1 to $1.9	$1.1 to $6.2	-$5.1 to $0.8

Assumption that association is not causal***	$0.1 to $0.9	$1.1 to $6.2
-$6.1 to -$0.2



Table 8.3c   California:  Annual Costs and Benefits of Attaining 0.065
ppm Standard (beyond 2020)

Premature Mortality Function 

or Assumption	Reference	Mean Total Benefits, in Billions of 1999$



Total Benefits*	Total Costs**	Net Benefits

NMMAPS	Bell et al. 2004	$1.2 to $5.6	$2.9 to $21	-$19 to $2.7

Meta-analysis	Bell et al. 2005	$3.2 to $7.6	$2.9 to $21	-$17 to $4.7 

	Ito et al. 2005	$3.4 to $7.8	$2.9 to $21	-$17 to $5

	Levy et al. 2005	$3.4 to $7.8	$2.9 to $21	-$17 to $5

Assumption that association is not causal***	$0.6 to $5	$2.9 to $21	-$20
to $2.1





* Tables present the total of CA glidepath in 2020, plus the additional
increment needed to reach full attainment in a year beyond 2020

** Includes ozone benefits and PM 2.5 co-benefits

*****Range was developed by adding the estimate from the ozone premature
mortality function to both the lower and upper ends of the range of the
PM2.5 premature mortality functions characterized in the expert
elicitation

****Range reflects lower and upper bound cost estimates

*****Total includes ozone morbidity benefits only



Table 8.4 California:  Annual Number of Ozone and PM2.5-Related
Premature Mortalities and Premature Morbidity Avoided (attainment beyond
2020)

Combined Estimate of Mortality

Standard Alternative and 

Model or AssumptionA	Combined Range of Ozone Benefits and

 PM2.5 Co-Benefits



0.075 ppm 	0.070 ppm	0.065 ppm

NMMAPS 	Bell (2004)	49 to 180	120 to 640	190 to 950

Meta-Analysis	Bell (2005)	160 to 290	310 to 830	500 to 1,300

	Ito (2005)	170 to 300	320 to 850	540 to 1,300

	Levy (2005)	170 to 300	330 to 850	530 to 1,300

No Causality	14 to 150	56 to 580	82 to 840



	Combined Estimate of Morbidity



	Acute Myocardial Infarction	80	320	460

Hospital and ER Visits	2,600	8,700	13,000

Chronic Bronchitis	28	110	170

Acute Bronchitis	75	300	440

Asthma Exacerbation	570	2,300	3,300

Lower Respiratory Symptoms	630	2,500	3,600

Upper Respiratory Symptoms	460	1,800	2,700

School Loss Days	120,000	210,000	340,000

Work Loss Days	3,900	16,000	23,000

Minor Restricted Activity Days	320,000	610,000	970,000



	

Note something in wrong with the California mortality/morbidity table,
the morbidity estimates may not appear.  We will fix in next edition. 
It is identical to ES-8.8.2 	Discussion of Results

Relative Contribution of PM benefits to total benefits

To be added, could not find e-mail with language from Bryan Hubbell.

Impact of Uncertainty in the Magnitude of ozone benefits.  

The degree to which net benefits are positive depends largely on the
size of the effect estimate used for the relationship between premature
mortality and ozone and to a lesser extent on the cost extrapolation
methodology and the magnitude of the relationship between ozone and
premature mortality.  In the cases where net benefits are negative, the
magnitude of the economic loss depends largely on the extrapolation
method used to calculate the costs of full attainment.  Because of the
high degree of uncertainty in these calculations, overall conclusions
about the magnitude of net benefits and the likelihood they will be
positive or negative for any of our evaluated scenarios cannot be drawn
with any degree of confidence.  As such, we cannot conclude that
strategies for attainment of a tighter ozone NAAQS would either pass or
fail a cost-benefit test.  In other words, we cannot make an estimate of
whether costs will outweigh benefits (or vice versa).  As we improve our
databases of control technologies and refine our understanding of the
magnitude of the relationships between air pollution and premature
mortality, our confidence in estimates of costs and benefits will likely
improve.

Challenges to Modeling Full Attainment in All Areas

Because of the poor air quality conditions in several large urban areas
(Southern California, Chicago, Houston, and the Northeastern urban
corridor, including New York and Philadelphia) and because of
limitations on the available database of currently known emissions
control technologies, EPA recognized from the outset that known and
reasonably anticipated emissions controls would likely be insufficient
to bring many areas into attainment with either the current or
alternative, more stringent ozone standards.  Therefore, we designed
this analysis in two stages:  the first stage focused on analyzing the
air quality improvements that could be achieved through application of
documented, well-characterized emissions controls, and the costs and
benefits associated with those controls.  The second stage utilized
extrapolation methods to estimate the costs and benefits of additional
emissions reductions needed to bring all areas into full attainment with
the standards.  Clearly, the second stage analysis is a highly
speculative exercise, as it is based on estimating emission reductions
and air quality improvements without any information about the specific
controls that would be available to do so.  

The structure of the RIA reflects this 2-stage analytical approach. 
Separate chapters are provided for the cost, emissions and air quality
impacts of modeled controls and for extrapolated costs and air quality
impacts.  We have used the information currently available to develop
reasonable approximations of the costs and benefits of the extrapolated
portion of the emissions reductions necessary to reach attainment. 
However, due to the high level of uncertainty in all aspects of the
extrapolation, we judged it appropriate to provide separate estimates of
the costs and benefits for the modeled stage and the extrapolated stage,
as well as an overall estimate for reaching full attainment.  There is a
single chapter on benefits, because the methodology for estimating
benefits does not change between stages.  However, in that chapter, we
again provide separate estimates of the benefits associated with the
modeled and extrapolated portions of the analysis.

In both stages of the analysis, it should be recognized that all
estimates of future costs and benefits are not intended to be forecasts
of the actual costs and benefits of implementing revised standards. 
Ultimately, states and urban areas will be responsible for developing
and implementing emissions control programs to reach attainment with the
ozone NAAQS, with the timing of attainment being determined by future
decisions by states and EPA.  Our estimates are intended to provide
information on the general magnitude of the costs and benefits of
alternative standards, rather than precise predictions of control
measures, costs, or benefits.  With these caveats, we expect that this
analysis can provide a reasonable picture of the types of emissions
controls that are currently available, the direct costs of those
controls, the levels of emissions reductions that may be achieved with
these controls, the air quality impact that can be expected to result
from reducing emissions, and the public health benefits of reductions in
ambinent ozone levels.  This analysis identifies those areas of the U.S.
where our existing knowledge of control strategies is not sufficient to
allow us to model attainment, and where additional data or research may
be needed to develop strategies for attainment.  EPA plans to address
some of these areas in the RIA analysis for the final rule through
additional research on control technologies, sensitivity analyses using
air quality models, and refinement of methods for extrapolating the
costs and benefits of reaching full attainment.

In many ways, regulatory impact analyses for proposed actions are a
learning process that can yield valuable information about the technical
and policy issues that are associated with a particular regulatory
action.  This is especially true for RIAs for proposed NAAQS, where we
are required to stretch our understanding of both science and technology
to develop scenarios that illustrate how certain we are about how
economically feasible the attainment of these standards might be
regionally.  The proposed ozone NAAQS RIA provided great challenges when
compared to previous RIAs.  Why was this so?  Primarily because as we
tighten standards across multiple pollutants with overlapping precursors
(e.g. the recent tightening of the PM2.5 standards), we move further
down the list of cost-effective known and available controls.  As we
deplete our database of available choices of known controls, we are left
with background emissions and remaining anthropogenic emissions for
which we do not have enough knowledge to determine how and at what cost
reductions can be achieved in the future when attainment would be
required.  This is not necessarily a weakness of the analysis as much as
the result of a natural process.  With the more stringent NAAQS, more
areas will need to find ways of reducing emissions, and as existing
technologies are either inadequate to achieve desired reductions, or as
the stock of low-cost existing technologies is depleted (causing the
cost per ton of pollution reduced to increase), new technologies will be
developed to fill these needs.  While we can speculate on what some of
these technologies might look like based on current research and
development and model programs being evaluated by states and localities,
the actual technological path is highly uncertain.  

Because of the lack of knowledge regarding the development of future
emissions control technologies, a significant portion of our analysis is
based on extrapolating from available data to generate the emissions
reductions necessary to reach full attainment of an alternative ozone
NAAQS and the resulting costs and benefits.  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.  Over longer time horizons, such as the time allowed for areas
with high levels of ozone pollution to meet the ozone NAAQS, the
opportunity for technical advances is greater (See Chapter 5 for
detail).  Also, due to the nature of the extrapolation method for
benefits (which focuses on reductions in ozone only at monitors that
exceed the NAAQS), we generally understate the total benefits that would
result from implementing additional emissions controls to fully attain
the ozone NAAQS (i.e., assuming that the application of control
strategies would result in ozone reductions both at nonattaining and
attaining monitors).

These extrapolated benefits are uncertain, but the relative uncertainty
compared to the modeled benefits is similar, once the underestimation
bias has been taken into account.  The emissions and cost extrapolations
do not have a clear directional bias, however, they are much more
uncertain relative to the modeled emissions and cost estimates, because
of the lack of refined information about the relationship between
emissions reductions and ozone changes in specific locations, and
because of the difficulties in extrapolating costs along a marginal cost
curve well beyond the observed data without accounting for shifts in the
cost-curve due to improvements in technology or use of technologies over
time.

8.3  	What did we learn through this analysis?  

As in our analysis for the PM NAAQS RIA, in selecting controls, we
focused more on the ozone cost-effectiveness (measured as $/ppb) than on
the NOx or VOC cost-effectiveness (measured as $/ton).  When compared on
a $/ton basis, many VOC controls appear cost-effective relative to NOx
reductions.  However, when compared on a $/ppb basis, NOx reductions are
almost always more cost-effective than VOC controls because of the much
lower conversion of VOC to ozone.  The one exception to this is in urban
areas which are VOC limited.  In those locations, NOx reductions can
actually result in increases in ozone, and as such, VOC reductions can
be cost-effective relative to NOx on a $/ppb basis.

Our knowledge of technologies that might achieve NOx and VOC reductions
to attain alternative ozone NAAQS is insufficient.  In some areas of the
U.S., our existing controls database was insufficient to meet even the
current ozone standard.  After applying existing rules and the
hypothetical controls applied in the PM NAAQS RIA across the nation
(excluding California), we were able to identify controls for 35 states
and DC that reduced overall NOx emissions by 17 percent and VOC by 4
percent.  For California, the percentages were 8 percent for NOx and 10
percent for VOC.  After these reductions, remaining emissions were still
substantial, with over 7 million tons of NOx and 9 million tons of VOCs
remaining.  The large remaining inventories of NOx and VOC emissions
suggests that additional control measures need to be developed, with
appropriate consideration of the relative effectiveness of NOx and VOC
in achieving ozone reductions. 

Most of the overall reductions in NOx achieved in our illustrative
control strategy were from non-EGU point sources.  This was due to the
fact that:  1) EGUs have been heavily controlled under the recent NOx
SIP call and Clean Air Interstate Rules.  The EGU program we included in
our strategy for meeting the alternative ozone standards was not
intended to achieve overall reductions in NOx beyond the CAIR caps, but
instead to obtain NOx emission reductions in areas where they would more
effectively reduce ozone concentrations in downwind nonattainment areas;
and 2) mobile sources are already subject to ongoing emission reduction
programs through the Tier 2 highway, onroad diesel and nonroad diesel
rules.  Thus, the opportunities for controlling NOx emissions were much
greater in the non-EGU sector than in the mobile or EGU sectors. 
However, the remaining NOx emissions from EGU and mobile sectors are
still greater than non-EGU sources, and additional reductions from these
sectors may need to be considered in developing strategies to achieve
full attainment.  We are evaluating technologies and programs that might
be applied in these sectors in the future.  Exploratory analyses
indicate that there are opportunities to achieve emission reductions
from EGU peaking units on High Energy Demand Days (HEDD) with targeted
strategies. Another area under analysis is the energy efficiency/clean
distributed generation based emission reductions. Potential changes in
the generation mix as a result of increase in the use of renewables and
Renewable Portfolio Standards (RPS) are also likely to create changes in
emission behavior. However, overall regional or national emissions
levels stay constant under a given cap. 

EPA’s existing mobile source programs will help areas reach
attainment. These programs promise to continue to help areas reduce
ozone concentrations between 2020 and 2030.  In California, continued
implementation of mobile source rules including the onroad and nonroad
diesel rules and the locomotive and marine engines rule are projected to
reduce NOx emissions by an additional 25 percent and VOC emissions by an
additional 11 percent during this time period.  These additional
reductions will significantly reduce the overall cost of attainment
relative to what California might have needed to reduce from other
sectors if attainment were to be required in 2020.  However, delaying
attainment by 10 years will result in delayed health benefits as well. 
Based on a simple scaling exercise, we estimate that between $0.3 and $1
billion in benefits could have been realized from full attainment with
the 0.070 ppm alternative each year between 2020 and 2030.  However, the
potential for extra costs of up to between $0.3 and $4 billion per year
suggests that allowing for delayed attainment until 2030 for these
severe nonattainment areas may make economic sense.  We are unable at
this time to identify controls that would achieve the full attainment in
California by 2020.

Tightening the ozone standards can provide significant, but not uniform,
 health benefits.  The magnitude of the benefits is highly uncertain,
and is not expected to be uniform throughout the nation. While our
illustrative analyses showed that the benefits of implementing a tighter
standard will likely result in reduced health impacts for the nation as
a whole, the particular scenarios that we modeled show that some areas
of the U.S. will see ozone (and PM2.5) levels increase.  This is due to
two reasons.  The first reason is that the complexities involved in the
atmospheric processes which govern the transformation of emissions into
ozone result in some locations and times when reducing NOx emissions can
actually increase ozone levels on some days (see Chapter 2 for more
discussion).  For most locations, these days are few relative to the
days when ozone levels are decreased.  However, in some urban areas the
net effect of implementing NOx controls is to increase overall ozone
levels and increase the health effects associated with ozone. This same
phenomenon results in some areas also seeing increases in PM2.5
formation.  The second reason is that the particular control strategy
that we modeled for EGU sources is a modification to controls on sources
within the overall cap and trade program in the Eastern U.S, established
under the CAIR.  As with any cap and trade program, changes in
requirements at particular sources will result in shifts in power
generation and emissions at other sources.  Because under our chosen EGU
control scenario the overall emissions cap for the CAIR region remains
the same, some areas of the country will see a decrease in emissions,
while others will see an increase.  This is not unexpected, and is an
essential element of the cap and trade program.  Our goal in selecting
the EGU control strategy was to focus the emissions reductions in areas
likely to benefit the most from EGU NOx emissions reductions, with
emissions increases largely occurring in areas in attainment with the
ozone NAAQS.  However, this necessarily means that in those areas where
emissions increases occurred, ozone levels would also be expected to
increase, with commensurate increases in health impacts.  On a national
level, however, we expected overall health benefits of the modeled EGU
strategy to be positive.  In addition, our air quality modeling analysis
showed that while ozone levels did increase in some areas, none of these
increases resulted in an attaining area moving into nonattainment. 
Adjustments to our control scenario might achieve a pattern of
reductions that achieves further air quality improvement.  

Tightening the ozone standards can incur significant, but uncertain,
costs 

An engineering cost comparison demonstrates that the cost of the 0.070
ppm Ozone NAAQS control strategy ($3.9 billion per year) is only
slightly higher than the Clean Air Interstate Rule ($3.6 billion per
year) and roughly one and half to just over four times higher than the
PM NAAQS 15/35 control strategy with annual engineering costs of $850
million. It should be noted that for the Ozone NAAQS $3.9 billion
represent the cost of partial attainment.  Full attainment using
extrapolation methods are expected to increase total costs
significantly.  Yet, the magnitude and distribution of costs across
sectors and areas is highly uncertain.  Our estimates of costs for a set
of modeled NOx and VOC controls comprise only a small part of the
estimated costs of full attainment.  These estimated costs for the
modeled set of controls are still uncertain, but they are based on the
best available information on control technologies, and have their basis
in real, tested technologies.  Estimating costs of full attainment
required significant extrapolation of the cost curve for known
technologies, and was based on generalized relationships between
emissions and ozone levels.  Based on air quality modeling sensitivity
analyses, there is clearly significant spatial variability in the
relationship between local and regional NOx emission reductions and
ozone levels across urban areas.  However, because we were unable to
analyze all of the urban areas that are expected to need reductions, we
used the same ratio of ozone to emissions throughout the U.S.  This
introduces significant uncertainty into the calculation of the emissions
reductions that might be needed to reach full attainment.  In addition,
because VOCs are generally much less effective than NOx in achieving
ozone reductions at key monitors (with the exception of California), we
did not use any VOC control data in the extrapolation to full
attainment.  This meant that in some areas, we assumed the need for more
expensive NOx controls than might be required if a specific area chose
to use a combination of NOx and VOC controls.  However, VOC controls
would have to be very inexpensive relative to NOx controls on a per ton
basis in order for VOC controls to be a cost-effective substitute for
NOx reductions.  Extrapolating costs by applying a cost-curve based on
known technologies also introduces uncertainties.  For some locations,
the extrapolation requires only a modest reduction beyond known
controls.  In these cases, the extrapolation is likely reasonable and
not as prone to uncertainties.  However, for areas where the bulk of air
quality improvements were derived from extrapolated emissions reductions
that go well beyond the area of the known controls, the increasing
marginal costs can suggest a cost per ton which stretches credibility. 
For example, in California, extrapolation to full attainment results in
a marginal cost for the last ton of NOx of $89,645 in Los Angeles and
$74,495 in Kern County, which are five to six times larger than the
marginal cost at the last known cost effective control.  Economic theory
would suggest that as marginal costs rise, research and development to
produce new, more cost effective technologies will also increase,
leading to a downward shift in the overall cost curve.  We did not
assume any shift in the cost curve to reflect technological innovation,
instead we provide a sensitivity analysis by showing estimates assuming
a high and low fixed cost per ton.  We are likely overstating costs in
the future when using the marginal cost and high fixed estimates.

Non-EGU point source controls dominate the estimated costs. These costs
account for about 70 percent of modeled costs.  The average cost per ton
for these reductions is approximately $3,400, and the highest marginal
cost for the last cost effective control applied is $15,267.  Mobile
source controls were also significant contributors to overall costs,
accounting for over 25 percent of total modeled costs.

California costs and benefits are highly uncertain..  California faces
large challenges in meeting any alternative standard, but their largest
challenges may be in attaining the existing standard.  Because our
analysis suggested that all available controls would be exhausted in
attempting (unsuccessfully) to meet the current 0.08 ppm standard
(effectively 0.084 ppm) all of the benefits and costs in California are
based on extrapolation.  Both the benefits and the costs associated with
the assumed NOx and VOC reductions in California are particularly
uncertain.  The costs are uncertain to the point where we have little
confidence that they represent a meaningful characterization of possible
future costs of implementation in California.  As such, we recommend
comparison of costs and benefits for the rest of the U.S. as a basis for
judging the relative merits of implementation.  Costs for full
attainment in California will clearly be substantial, but the level of
uncertainty about those costs is simply too great to provide any useful
conclusions.  This is also true for many other areas of the U.S., but
the uncertainties are magnified in the case of California.

Costs and benefits will depend on implementation timeframes. States will
ultimately select the specific timelines for implementation as part of
their State Implementation Plans.  To the extent that states seek
classification as extreme nonattainment areas, the timeline for
implementation may be extended beyond 2020, meaning that the amount of
emissions reductions that will be required in 2020 will be less, and
costs and benefits in 2020 will also be lowered.

Both costs and benefits should be viewed in absolute and relative terms.
Costs should be viewed relative to the overall size of the economy. 
While total costs for the 0.070 ppm standard are significant at $13 to
$26 billion, this amounts to roughly 0.1% of an estimated GDP of $17
trillion in 2020.  These costs are also relatively small on a per
household basis.  Based on an estimated 130 million households in 2020,
the annual cost per household ranges from $100 to $200.  Likewise,
annual benefits per household range from $11 to $240.  Both costs and
benefits of the rule will thus be a relatively small proportion of total
economic production, as well as a small percentage of total annual
household income.      

 Because these two areas adopted 8-hour ozone implementation plans
calling for post-2020 attainment after we had  completed much of our
analysis, these areas are assumed to meet the current standard in 2020
and 2021 respectively, somewhat earlier than the date in their plans,
which results in a steeper glide path, and higher costs and benefits,
for all the standards analyzed.  See chapter 4.

 See footnote 1.

 PAGE   

8- PAGE   12 

