UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
September 26, 1996 OFHCEOFTWADM**™™
SCIENCE ADVISORY BOARD
EPA-SAB-ACCACA-LTR-96-010
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M. Street, SW
Washington, DC 20460
RE: Review of "The Benefits and Costs of the Clean Air Act, 1970 to 1990" by the
Physical Effects Review Subcommittee (PERS) of the Advisory Council on Clean
Air Compliance Analysis (ACCACA)
Dear Ms. Browner:
This letter report summarizes the major consensus conclusions drawn by the
Physical Effects Review Subcommittee (PERS) of the Advisory Council on Clean Air
Compliance Analysis (ACCACA, or "the Council") as a result of its analyses of the May
3, 1996 draft of the EPA document entitled "The Benefits and Costs of the Clean Air
Act, 1970 to 1990", which is referred to hereafter in this letter as the "Document". The
substance of this letter report was formulated at the PERS meeting in Alexandria, VA on
June 4, 1996 and reported orally to the Council at its meeting in Alexandria on June 5
and 6, 1996. This report was reviewed and endorsed by PERS at a technical editing
teleconference held by the PERS on July 30, 1996. Many more specific technical and
editorial issues have been noted by members of the PERS, and these have been
transmitted to EPA Staff. We trust that they will be given due consideration in the
preparation of the final revisions of the Document.
Although the PERS members have reviewed and provided comments on all
chapters in the Document, the consensus conclusions summarized herein are limited to
our analyses of the chapters on health and welfare effects and the corresponding
exposure-response relationships for which the PERS has specific expertise.
We are pleased to report that the authors have been very responsive to the
critiques and suggestions provided by PERS following its earlier reviews of preliminary
drafts. There are now much clearer statements of: a) what has been analyzed; b) the
bases for the selection of data and models used in the analyses and of those effects
either not analyzed or not monetized; c) the nature and extent of the uncertainties
associated with the results reported; and d) what is known about the nature and extent
of those effects of air pollution which cannot be adequately analyzed and/or monetized.
Recycled/Recyclable
Printed with Soy/Canda Ink on paper that
contains at toast 50% recycled fiber
-------
The only significant exception is the area of welfare effects, especially in terms of
agricultural and forest productivity and ecological impacts. In these areas the literature
reviews in the Document are largely limited to review papers and are not up-to-date. In
particular, we suggest that the Staff utilize the up-to-date reviews and analyses in the
recently completed Criteria Document and Staff Paper on Photochemical Oxidants in
preparing the final draft. It provides sufficient data for the valuation of some damages to
crops and forests. Inclusion of such analyses would help to balance the overall report.
The PERS's major concerns are focussed on the analyses and interpretations of
the epidemiology data derived from the time-series (daily mortality) studies and the
cross-sectional (annual mortality) studies for particulate matter (PM) and its closely
associated co-pollutants such as sulfur dioxide (SO^, nitrogen oxides (NO,,), and ozone
(03). It has become clear in recent years that these pollutants, whether individually or
collectively, are significantly associated with excess mortality, and that the associations
generally are most closely and consistently related to the ambient air concentrations of
thoracic particles (PM10) and/or fine particles (PM 25). These pollutants have some
common sources (e.g., combustion), and their concentrations tend to rise and fall
together over large geographic regions. If PM exposure is not an independent causal
factor for excess mortality, then it is at least a good surrogate measure of the exposure
to the pollutant mixture. Therefore, for the kinds of aggregate analyses used in this
Document, it is appropriate to use the PM effects coefficients from the epidemiological
studies as surrogates for the exposure to the mixtures. It follows that it is inappropriate
to also add separate SO2 or O3 effects estimates in addition to those for PM, since they
would produce "double-counting" (unless they are included in the model simulta-
neously). In multiple-regression studies where separate coefficients for each pollutant
were reported from separate models, the PM-only coefficient is the most practical one to
use in this aggregate effects analysis, as it is the most consistently available. To the
extent that other air pollutants have effects on mortality that do not correlate well with
those of PM, then the reliance on PM as a surrogate for the effects of the mixture on
mortality may understate the total impact of air pollution on mortality.
The majority of the PERS members concluded that the rationale for excluding
annual mortality studies from the benefits summation was weak. One major reason for
their omission may have been that the authors considered only the two recent prospec-
tive cohort studies [Harvard 6-cities (Dockery efa/., 1993) and the 151 communities in
the American Cancer Society study (Pope et a/., 1995)]. By neglecting to cite earlier
"ecological" studies (e.g., Ozkaynak and Thurston, 1987; Lipfert, 1984; Lave and
Seskin, 1977) the Document suggested that only the two recent prospective cohort
studies provided useful data on annual rates of mortality in relation to PM exposures.
What the two cohort studies did show is that the association between ambient air
pollution and mortality cannot be explained by confounding influences of smoking and
other personal risk factors. Thus, the similar coefficients of response among the cohort
and ecological studies indicate a consistent body of data implicating air pollution as a
causal factor for variations in annual mortality rates.
-------
The fact that excess mortality is significantly associated with ambient PM
concentrations for both short-term and long-term exposures is important in terms of the
credibility of each kind of exposure as a significant cause for adverse responses. If
there were no evidence for excess annual mortality, then it would imply that the excess
daily mortality was either due to very short-term advancement of mortality, or was an
artifact of additional influential variables not considered in the analyses. Conversely, if
there were no evidence for excess daily mortality on highly polluted days, the likelihood
of chronic air pollution exposure being a causal factor for an excess of annual mortality
would be reduced.
The draft Document analyzed exposure-response relationships for both acute
and chronic mortality, but recommended that only the acute mortality coefficients be
used for the summary benefits calculations. The PERS recommends that both the
acute and chronic mortality benefits of the pollution reductions attributable to the Clean
Air Act (CAA) (1970-1990) be fully presented and discussed in the revised Document.
The discussion should include a summary of the major uncertainties of both types of
studies and provide a balanced discussion of the advantages and limitations of each
kind of analysis for an unbiased cost accounting.
The daily mortality analyses have some advantages as evidence of PM-associ-
ated excess mortality. They are not complicated by unknown intercommunity differ-
ences in ethnicity, lifestyle, housing density, socio-economic factors, climate, etc. that
may confound the interpretation of the annual mortality. Also, more individual studies
are available and more comprehensive analyses and reviews have been performed for
this data set, including those in the recently completed PM Criteria Document. Thus,
daily mortality should continue to be a significant element in the analysis of benefits of
the CAA.
With respect to the treatment of the daily mortality data in the Document, the
PERS offers the following comments and recommendations:
a) Expand listing of time-series analyses to other studies having the statisti-
cal power to detect small relative risks. In other words, do not omit good
quality studies that failed to report statistically significant associations
between daily mortality and PM concentrations;
b) List only the effects associated with PM concentration, on the basis that
PM is representing the pollution mixture;
c) Do not characterize the range of mortality coefficients based on the single
highest and single lowest reported positive coefficients. If high and low
estimates are to be reported, they should be based on the confidence
-------
intervals of the distribution of the PM coefficients from the individual
studies, such as, for example, those at 2 standard deviations from the
mean.
The daily mortality analyses have the disadvantage that the extent of prematurity
of mortality cannot be accurately determined or inferred from the available literature.
Thus, as discussed in the June 1996 Letter Report of the Council, it is difficult to assign
monetary values to the reductions in daily mortality attributable to the reduced pollution.
By contrast, the annual mortality can more readily be expressed in terms of
reduced lifespan, providing a basis for the monetary valuation of the estimated life
extension. On the other hand, the annual mortality studies have some limitations that
do not affect the daily mortality studies to the same degree. These include:
a) Assigning the appropriate period of exposure to the coefficient of re-
sponse. If the relevant exposure extends over a decade or more, then a
coefficient determined on the basis of current or recent years' exposure
could be inflated when pollution levels have decreased markedly.
b) Insufficient or inappropriate adjustment for confounding factors affecting
long-term mortality rates. There may be differences among communities
in diet, exercise, ethnicity, climate, industrial effluents, etc. which could
affect longevity. Such intercommunity differences are much less likely to
affect the results of time-series studies of daily changes in the same
community.
Despite these concerns, cross-sectional (annual) mortality rate studies have
consistently shown evidence for excess mortality in relation to indices of fine airborne
particles (sulfates and PM25), as discussed previously. In completing the analyses of
annual mortality, we recommend that EPA Staff use a somewhat different approach
than that used for the daily mortality studies, where we recommended using a mid-value
and statistically based upper and lower bounds based on the results of the numerous
valid studies. For annual mortality, we recommend that the benefits analysis should be
based primarily on the Pope et al. (1995) study results for the following reasons:
a) It is based on a much larger population (> 500,000) and many more
communities (151 vs. 6) than the other prospective fine particle cohort
study (Dockery et al., 1993).
b) Its results are consistent with the results of Dockery et al. (1993), which
reported even larger mortality responses, but only in six cities.
c) Its results are consistent with ecological population studies of annual
mortality based on 1960 and 1970 census data (e.g., Lave and Seskin,
1977), 1970 census data (Lipfert, 1984) and 1980 census data (Ozkaynak
-------
and Thurston, 1987). These ecological studies could not account for
individual risk factors, such as smoking and occupational exposures,
which are now known to have only modest influences based on the recent
analyses from the prospective cohort studies of Dockery et al. (1993) and
Popeefa/. (1995).
A specific issue associated with the annual mortality study of Pope et al. (1995)
is that the mortality rates were determined by relating mortality associated with long-
term pollutant exposures using ambient air data for approximately 15 years prior to
death. If mortality is also related to exposures earlier in life, when fine particle pollution
levels could have been 50-100% higher, then the calculated coefficient may overesti-
mate the effect. On the other hand, Pope et al. (1995) noted that their population was
largely composed of middle class Caucasians who have lower mortality rates and
possibly lower exposures to pollution than poorer minority populations. Their study may
therefore have yielded lower coefficients of response than those appropriate for minority
populations, or for the population as a whole. In any case, as noted previously, the
assumption is made that the fine particle associations with annual mortality represent
the overall responses to community air pollution as a complex mixture rather than to the
particulate mass components alone.
In summation, the annual mortality studies are important for several reasons.
These include:
a) The excess mortality estimates are greater than that of the annual sum of
daily mortality from the more numerous time-series mortality studies,
suggesting that excess daily mortality during pollution episodes is not
largely balanced by deficits in daily mortality during cleaner days that
follow.
b) The annual mortality estimates can be expressed in terms of duration of
life-shortening as well as number of excess deaths. This provides an
alternate means for determining the benefits of pollution controls that may
be more easily interpreted by risk assessors, risk managers, and the
public.
c) Chronic cardiopulmonary damage produced by long-term pollution expo-
sure can account for both reduced lifespan and predisposing individuals to
be at special risk for excess daily mortality during the stress associated
with peak pollution episodes.
Thus, the document should not have omitted the estimates of mortality based on
the annual studies from its summary tabulations. These estimates differ from those
based on the time-series studies, both in terms of their nature (number and quality of
life-years saved) and their implications for assigning monetary values to them. Thus,
they should be presented as separate ledger entries from those for daily mortality, and
-------
no combined mortality estimate should be presented. It should be made clear that
combining them would be double-counting, since excess daily mortality is one compo-
nent of excess annual mortality. Presenting them separately provides the reader with
alternative, plausible means of judging the extent of the benefits of the estimated
reductions in mortality attributable to the control activities resulting from the Clean Air
Act between 1970 and 1990.
The quantitation of the mortality associated with lead (Pb) is based on a different
kind of analysis from either the time-series (daily) or chronic (annual) mortality studies of
populations, and it is appropriate to tabulate its impact on mortality separately. It is
based on epidemiological studies showing statistically significant associations between
blood lead (PbB) and blood pressure, combined with the well-established association
between blood pressure and mortality rates. Thus, the PERS recommends separate
discussions and tabular summation for: a) the excess daily mortality associated with the
non-Pb criteria pollutants, using daily PM concentration as the surrogate index; b) the
excess annual mortality associated with the non-Pb criteria pollutants, using annual
average PM concentration as the surrogate index; and c) the excess annual mortality
associated with Pb exposure attributable to air emission, as indexed by population
averages of PbB.
The benefits analyses for the morbidity associated with exposure to lead (Pb)
also have uncertainties. However, those for Pb have a much clearer mechanistic basis
for the underlying biological responses that supports the extensive and generally
consistent epidemiological data base. The control of exposure to Pb is clearly one of
the great success stories of the CAA as well as being the best documented success,
and this is reflected in the thorough and well described benefits analysis in the Docu-
ment.
Regarding the estimates of morbidity effects for the other criteria pollutants, the
PERS was most concerned about the treatment of chronic bronchitis, especially the
staff decision not to include the number of cases of chronic bronchitis avoided in its
summaries of quantified effects in Chapter 5 and Appendix D. The question of what
monetary value to assign to the reduction in the incidence of this endpoint is separate
from the question of the reliability of the statistical evidence linking PM to increases in
the incidence of chronic bronchitis. We understand that the former question will be
addressed by the Council. But whether or not a monetary value is assigned to this
effect, the estimated number of cases avoided should be included as a quantified health
effect.
With regard to the other morbidity effects listed and evaluated for exposure-
response relationships, the PERS concludes that the Document provides as thorough,
careful and complete a summation as the available literature permits. We hope the
Council's report will emphasize the severe limitations imposed on these analyses by the
paucity of relevant data, especially the lack of sufficient scientific information and/or
economic tools to monetize ecological benefits. We hope EPA and the Congress will
-------
mobilize the resources needed to generate the data to mount more thorough and
satisfactory analyses of costs and benefits of air pollution controls in future projects of
this kind.
Sincerely,
Dr. Morton Lippmann, Chair
Physical Effects Review Subcommittee (PERS)
Dr. Richard Schmalensee, Chair
Advisory Council on Clean Air Act
Compliance Analysis (ACCACA)
-------
NOTICE
This report has been written as part of the Science Advisory Board, a public
advisory group providing extramural scientific information and advice to the Administra-
tor and other officials of the Environmental Protection Agency. The Board is structured
to provide balanced, expert assessment of scientific matters related to problems facing
the Agency. This report has not been reviewed for approval by the Agency and, hence,
the contents of this report do not necessarily represent the views and policies of the
Environmental Protection Agency, nor of other agencies in the Executive Branch of the
federal government, nor does mention of trade names or commercial products consti-
tute a recommendation for use.
-------
ABSTRACT
The Physical Effects Review Subcommittee (PERS) of the Advisory Council on
Clean Air Compliance Analysis (ACCACA, or the Council) of the Science Advisory
Board (SAB) reviewed the Agency's May 3, 1996 draft Report to Congress entitled
"The Benefits and Costs of the Clean Air Act, 1970-1990." The Agency was responsive
to the PERS critiques of preliminary drafts, with exceptions being agricultural and forest
productivity and ecological impacts. In these areas, the Agency's draft document is still
not up-to-date.
For PM-associated excess daily mortality, the PERS noted that it is difficult to
assign monetary values to the reductions to mortality. By contrast, the annual mortality
can more readily be expressed in terms of reduced lifespan, providing a firmer basis for
the monetary valuation. The Agency should not have omitted the estimates of annual
mortality studies from its summary tabulations, and estimates of annual mortality should
be presented as a separate ledger item from the daily mortality. Combining these
estimates would be double-counting, since excess daily mortality is one component of
excess annual mortality. The PERS also recommended a separate discussion for lead
mortality. The control of exposure to lead is a success story, and this is reflected in the
thorough and well described benefits analysis in the Agency's draft document.
With the exception of the treatment of chronic bronchitis, the estimates of the
morbidity effects of the criteria pollutants evaluated for exposure-response relationships
in the Agency's draft document has provided a thorough, careful and complete an
evaluation and summation as the available literature permits.
Key Words: Air Pollutants, Clean Air Act, Cost-Benefit Analysis, Economic Valuation,
Mortality Effects, Morbidity Effects, Physical Effects of Air Pollutants, Valuation Method-
ologies
-------
REFERENCES
Dockery, D.W.; Pope, C.A. Ill; Xu, X.; Spengler, J.D.; Ware, J.H.; Fay, M.E.; Ferris,
B.G., Jr.; and Speizer, F.E. .1993. An association between air pollution and
mortality in six U.S. Cities. N. Engl. J. Med. 329:1753-1759.
Lave, LB. and Seskin, E.P. .1977. Air Pollution and Human Health, Baltimore, M.D.;
The Johns Hopkins University Press
Lipfert, F.W. .1984. Air Pollution and Mortality:Specification Searches Using SMSA-
Based Data. J. Environ. Econ. Manage. Vol. 11: 208-243
Ozkaynak, H. And Thurston, G.D. .1987. Associations Between 1980 U.S. Mortality
Rates and Alternative Measures of Airborne Particle Concentration. Risk Anal.
Vol.7:449-461
Pope, C.A. Ill; Thun, M.J.; Namboodiri.M.; Dockery, D.W.; Evans, J.S.; Spizer, F.E., and
Heath, C.W., Jr. .1995. Particulate Air Pollution is a Predictor of Mortality in a
Prospective Study of U.S. Adults. Am. J. Respir. Care Med., Vol. 151: 669-674
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
ADVISORY COUNCIL ON CLEAN AIR COMPLIANCE ANALYSIS
PHYSICAL EFFECTS REVIEW SUBCOMMITTEE
CHAIR
Dr. Morton Lippmann, Professor, New York University Medical Center, Nelson Institute of
Environmental Medicine, Tuxedo, NY
VICE-CHAIR
Dr. A. Myrick Freeman, Professor of Economics, Department of Economics, Bowdoin College,
Brunswick, ME
MEMBERS AND CONSULTANTS
Dr. David V. Bates, Professor Emeritus of Medicine, Department of Health Care and Epidemiol-
ogy, University of British Columbia, Vancouver, BC CANADA
Dr. Gardner M. Brown, Jr. Professor of Economics, Department of Economics, University of
Washington, Seattle, WA
Dr. Timothy V. Larson, Professor of Environmental Engineering and Science Program,
Department of Civil Engineering, University of Washington, Seattle, WA
Dr. Lester B. Lave, Professor, Graduate School of Industrial Administration, Carnegie-Mellon
University, Pittsburgh, PA
Dr. Joseph S. Meyer, Assistant Professor, Department of Zoology and Physiology, University of
Wyoming, Laramie, WY
Dr. Robert D. Rowe, Director, Hagler Bailly Consulting, Inc., Boulder, CO
Dr. George E. Taylor, Jr., Professor and Chairman, Department of Environmental and Resource
Sciences, University of Nevada, Reno, NV
Dr. Bernard Weiss, Professor, Department of Environmental Medicine, University of Rochester
Medical Center, School of Medicine & Dentistry, Rochester, NY
Dr. George T. Wolff, Principal Scientist, Environmental & Energy Staff, General Motors
Corporation, Detroit, Ml
SCIENCE ADVISORY BOARD STAFF
Dr. K. Jack Kooyoomjian, Designated Federal Official, Science Advisory Board (1400), U.S.
Environmental Protection Agency, 401 M Street, SW, Washington, DC 20460
Dr. Donald G. Barnes, Staff Director, Science Advisory Board
Mrs. Diana L. Pozun, Staff Secretary, Science Advisory Board (1400), U.S. Environmental
Protection Agency, 401 M Street, S.W., Washington, D.C. 20460
------- |