IMiad WM«« 01(10* *4 tilt MUnMMntM Mft>CAttC-*T-W
i, 4>C
Report of the
dean Air Scientific Advisory
Committee (CASAC)
Review of the
Office of Policy, Planning
3nd Ev3iU3iiofi s
Material Damage
Assessment
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
- WASHINGTON. D C. Z04SO
June 30, 1987
The Honorable Lee M. Thomas
Administrator
U.S, Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
Dear Mr, Thomas:
The Material Damage Review Subcommittee of the Clean Air Scientific
Advisory Committee (CASAC) has completed its review of several documents
pertaining to an analysis of the effects of idd deposition on materials.
This review, requested by the Office of Policy, Planning, and Evaluation,
focused on the contractor prepared-report entitled "A Damage Function Assess-
ment of Building Materials: Ths Impact of Acid Deposition" (Mathtech, Inc.,
1986). The Subcommittee assessed four components of the analysis, namely;
degree that the materials inventory is representative of urban areas,
physical damage functions relating acid deposition to material damage, eco-
nomic damage calculations for estimating incremental acid deposition costs,
and extrapolation from the case study cities to other major urban areas of
the United States.
Generally, the Subcommittee concludes that the 1986 Mathtech report
is well done, given the limitations in the available data and the scope of
the study, and it represents an improvement over earlier efforts. The
report identifies the assumptions and many of the potential omissions, errors,
and biases inherent in the work, and tries to account for a range of possible
alternatives by furnishing lower and upper damage estimates. Although the
researchers have performed competent work in view of the limited resources
and research direction, the work reflects continued limitations in knowledge
and data bases available to the researchers.
In view of the uncertainties involved, especially in paint damage costs,
the Subcommittee believes that the total costs from add deposition should
not be used f» the Sulfur Oxides National Ambient A1r Quality Standards
(NAAQS) rulefflaking process. Nevertheless, the conceptual framework and
procedures that are used in this report do provide useful information which
should be considered. The analyses contained in this report should be
considered as complementary to the supply/demand model approach that is now
incorporated 1n the draft Regulatory Impact Analysis (RIA) for Sulfur Oxides.
-------
-2-
Thank you for the opportunity to present our views on this important wel-
fare issue. We request that tht Agency officially respond to the scientific
advice contained in the attached report.
Sincerely,
cc; A. James Barnes
Bruce Jordan
Dick Livingston
Richard Morgensttrn
Bill O'Netl
Craig Potter
Janet Schefd
Terry Yosie
Morton lippmann
Chairman
Clean Air Scientific
Advisory Committee
Warren B. Johnson
Chai rmin
Material Damage Review
Subcommittee
U^<^^
-------
SAB-CASAC -₯? -~Q3l
Report of the Clean Air Scientific Advisory Committee
Material Damage Review Subcommittee
ECONOMIC DAMAGE TO BUILDING MATERIALS
EXPOSED TO ACID DEPOSITION
June 1987
Scltnct Advisory Board
U.S. Environmental Protection Agency
Washington, DC
-------
NOTICE
This report has been written as part of the activities of the Science
Advisory Board, a public advisory group providing extramural scientific
information and advice to the Administrator and other officials of the
Environmental Protection Agency. The Board is structured to provide a
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 docs mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
-------
U.S. Environmental Protection Agency
Science Advisory Board
Clean A1r Scientific Advisory Committee
Material Damage Review Subcommittee
Chairman
Dr. Warren B, Johnson, Manager, Research Aviation Facility, National
Center for Atmospheric Research, Boulder, Colorado
Members
Dr. Robert Baboian, Head of the Corrosion Laboratory* Texas Instruments,
Inc., Attleboro, Massachusetts
Or. A. Myriek Freeman III, Department of Economics, Bowdoin College,
Brunswick> Maine
Or. Wesley Magtt, Fuqua School of Business, Duke University, Durham,
North Carolina
Dr. Robert D. Rowe, Vice President, Environmental and Resource
Economics, Energy and Resource Consultants, Inc., Boulder, Colorado
Or. Kip Viscusi, Department of Economics, Northwestern University*
Evanston, Illinois
Mr. John Vocom, Vice President and Chief Consulting Engineer, TRC
Environmental Consultants, Inc., East Hartford, Connecticut
Executj ve Secretajry
Mr. A. Robert Flaak, Environmental Scientist, Science Advisory Board
(A-IQ1F), U.S. Environmental Protection Apncy, Washington, DC
20460
-------
U.S. Environmental Protection Agency
Science Advisory Board
Clean A1r Scientific Advisory Committee
Chairman
DP. Morton lippmann, Professor, Institute of Environmental
New York University Medical Center, Tuxedo, New York i0987
Members
Or. Robert Frank, Professor of. Environmental Health Sciences* Johns
Hopkins School of Hygiene and Public Health* 61S N. Wolfe Street,
Baltimore, Maryland 21205
Dr. Warren B. Johnson, Manager, Research Aviation Facility, National
Center for Atmospheric Research, P.O. Box 3000» Boulder,
Colorado 30307
Or. Timothy Larson, Environmental Engineering and Science Program,
Department of Civil Engineering FX-10, University of Washington,
Seattle, Washington 98195
Dr. Gilbert S. Omenn, Professor and Dean, School of Public Health and
Community Medicine, SC-3Q, University of Washington, Seattle,
Washington 98191
Or, James H. Ware, Associate Professor, Harvard School of Public Health,
Department of B1ostat1sties, 677 Huntlngton Avenue, Boston,
Massachusetts 02115
Dr. Jerry Wtsfllowskl, A1r and Industrial Hygiene laboratory, California
Department of Health, 2151 Berkeley Way, Berkeley, California 94704
Executlve Secretary
Mr. A. Robert Flaak, Environmental Scientist, Seitnce Advisory loard
(A~101F), U.S. Environmental Prottctlon Agency, Washington, DC 20460
-------
TABLE OF CONTENTS
1, EXECUTIVE SUMMARY 1
2. INTRODUCTION 3
3. ASSESSMENT OF THE COMPONENTS OF THE ANALYSIS 5
A. Materials Inventory 5
8. Physical Damage Calculations 6
1) Paint 6
2) Mortar 6
3) Econometric Considerations 7
C. Economic Damage Calculations 7
1) General 7
2) Reasons for Repainting 8
3} Behavior 9
4} Costs of Repainting 9
5) Silicate-Base Paints 10
6) Stone Buildings 10
0. Extrapolation Procedures 11
1) General ' 11
2) The Mix of Cities 11
3) Materials Use by Region 11
4. ADDITIONAL ISSUES 12
A. Use of Results in the Rulemaking Process 12
B. A More Credible Approach 12
APPENDICES
A, Additional Specific Comments A-l
B. Citations of Review Documents B-l
C. Charge to the CASAC Material Damage
Review Subcommittee C-l
-------
This is the report of the Material Damage Review Subcommittee of the
Clean Air Scientific Advisory Committee (CASAC). The Subcommittee was
formed at the request of the Office of Policy, Planning and Evaluation
(OPPE) to review several documents pertaining to an analysis of the
effects of acid deposition on materials. These documents are expected to
be used in revising the Regulatory Impact Analysis (RIA) associated with
the National Ambient Air Quality Standards (NAAQS) for Sulfur Oxides,
The primary review document was the report "A Damage Function
Assessment of Building Materials: The Impact of Acid Deposition", pre-
pared in May 1936 by the Mathtech Corporation for OPPE, This report
attempts to develop quantitative estimates of the economic damages
associated with the exposure of common construction materials to current
levels of acid deposition. -It also relates to the broader question of
uncertainties concerning the magnitude and distributional impacts of the
benefits and costs associated with stricter emission control programs.
Generally, the CASAC Material Damage Review Subcommittee concludes
that the 1986 Mathtech report is well-executed and documented, given the
limitations in the available data and the scope of the study, and it
represents an improvement over earlier efforts. The report Identifies
the assumptions and many of the potential omissions, errors, and biases
inherent in the work, and tries to account for a range of possible altern-
atives by furnishing lower and upper damage estimates. The researchers
have performed competent work in view of the limited resources and research
direction. Nevertheless, the work reflects continued limitations in
knowledge and data bases available to the researchers.
We believe that the reported economic damage estimates dye to acid
deposition in the study region can only be considered suggestions. This
is primarily due to the potentially limited reliability and applicability
of the paint and mortar damage functions, although other components of
the analysis further diminish accuracy.
The Subcommittee's review suggests the final economic estimates are
likely (although not certain) to be overstated by the assumptions used
in the absence of desired ddtd. As a result, the lower bound estimates
appear to have more credence than the best or upper bound estimates.
The zinc estimates do ippeir to be of acceptable reliability, while the
more important paint ind mortar damage estimates ire of very limited
reliability.
-------
While the damages are potentially overstated for the materials and
regions considered, the omission of several potentially important regions
and materials may be an important counterbalancing consideration.
The Subcommittee has the following major findings and recommenda-
tions;
t The materials inventory conducted for the four study cities
addresses the major structural uses and susceptible materials.
The inventory represents a substantial improvement in the data
base over previous work. However, it is difficult to judge how
representative these four cities are of the 17-state study area.
Therefore, the extrapolation from the four study cities to other
cities must be considered tentative and uncertain.
The physical damage- functions selected for paint, stone, and
mortar are perhaps the most accurate functions available, but
are of highly uncertain reliability and may be of limited
applicability in this analysis. The damage function for zinc
has a higher reliability and lower uncertainty than paint, stone
or mortar. While this damage function for zinc appears to relate
realistically the damage to remediation scenarios, the damage
function for paint, based on erosion, is not necessarily related
to remediation measures (e.g., repainting). This is an especially
serious shortcoming since paint damage is potentially the most
important component of total damage costs from acidic deposition.
As a result, the physical damage functions severely limit the
confidence one may place in the overall results for use in policy
analysis. Due to these uncertainties, the current paint, stone
and mortar damage functions should not be used in the RIA.
Given the data available, the methodology for making economic
damage calculations appears reasonable, but given the problems
with the damage functions and the many assumptions necessary,
there may be significant biases in the total damage costs.
More could be done to analyze actual behavioral responses to
paint erosion as well as the nature of the costs consumers incur
with repainting activities. In addition, important improvements
can be made concerning perceptions, behavioral responses, and
valuation of damages.
Overall, the lower end of the cost range is more credible than
the upper end of the range. It may be, however* because of the
substantial uncertainties involved, that the low end of the
range is too high. Thus, the Subcommittee does not endorse any
particular estimate as being correct. Rather, we regard the un-
certainties as being sufficiently fundamental to warrant further
study.
-2-
-------
The 1986 Mathtech analyses represent an alternative approach to
materials damage estimation compared to the supply/demand model
approach (Mathtech 1982) currently incorporated into the draft
RIA, The approaches taken in the two reports are quite different
and the results should be considered as complementary to each
other.
The 1986 Mathtech report should be cited in the draft RIA, but
should not replace the current economic analysis. Rather, a
more thorough comparison of the results of the two efforts should
be included as a basis for indicating "plausible" levels of
economic damages under alternative sets of restrictive assumptions
and available data. The limitations of the analyses should also
be acknowledged.
A substantially improved understanding of the types and levels
of damages induced by acid deposition and an improved response
by the public is needed.
The Subcommittee also identified a number of other issues and made
additional specific comments in Appendix A.
2. INTRODUCTION
On April 21, 1986, the Clean Air Scientific Advisory Committee
(CASAC) was briefed by Dr. Thomas Lareau of the Office of Policy,
Planning and Evaluation (OPPE) concerning Agency sponsored materials
damage analysis. This analysis is intended to be used by the Office of
Air Quality Planning and Standards (OAQPS) in preparing the Regulatory
Impact Anaylsis (RIA) on the National Ambient Air Quality Standards
(NAAQS) for Sulfur Oxides. Or. Lareau stated that it was the intention
of OPPE to request formal CASAC review at a later date.
In July 1986, OPPE formally requested CASAC review of several
documents concerning material damage. These documents were the final
Mathtech report, "A Damage Function Assessment of Building Materials:
The Impact of Acid Deposition" (March 1986), and supporting documents
(see Appendix B for full citations) including:
t "Economic Benefits of Reduced Acidic Deposition on Common
Building Materials: Methods Assessment"
"Material Effects Assessment"
"Economic Damages to Building Materials Exposed to Acidic
Deposition"
"Derivation of Metallic Corrosion Damage Functions for Use in
Environmental Assessments"
-3-
-------
"Atmospheric Acid Deposition Damage to Paints"
t "Benefit Analysis of Alternate Secondary NAAQS for Sulfur Dioxide
and TSP"
CASAC formed a Subcommittee in July 1986, chaired by Dr. Warren
Johnson of the National Center for Atmospheric Research, with the charge
(See Appendix C) to review the 1986 Mathtech report and the supporting
documents to determine if the methods discussed were scientifically
credible and whether the data were appropriate for estimating materials
damage from acid deposition in a 17-state area of the United States. In
particular, the Agency asked for an assessment of four components of the
analysis*.
Materials Inventory - Does the inventory provide a representative
sample of the distribution of materials in urban areas that can be
used to extrapolate to other urban areas?
Damage FujKjtions^ - Do the physical damage functions accurately de-
scribe tfieTeTation between acid deposition and materials damage?
Ecgnomlc Damage Galcy1 ations - Are the assumptions about baseline
maintenance practices appropriate for estimating incremental acid
deposition costs?
Extrapolation - Is a credible method used for extrapolating from
the four case study cities to other major urban areas in the
Northeast and North Central United States?
The Subcommittee was also asked to provide its judgment on the following
questions;
t Are the analyses useful input for the Sulfur Oxides KAAQS
rulemaking process?
Are these analyses a more credible approach to materials damage
estimation than the supply/demand model approach that is currently
incorporated into the draft RIA?
The Subcommittee conducted its initial review via the mail* On
December 4, 1986, tht Subcommittee held a public meeting in Washington,
DC to formally discuss the documents with Agency staff, and to obtain
comments from the interested public.
-4-
-------
3. ASSE_SSMEBT..OF. IHE_ CQWQfOrrS OF THE ANALYSIS
A- Materials Inventory
Does the inventory provide a representative sample of the
distribution of materials in urban areas that can be used
to extrapolate to other urban areas?
The method used for the inventory in the four urban areas (New Haven,
Portland, Pittsburgh* and Cincinnati) Is a reasonable approach to take
without spending inordinate amounts of time and funds to expand the
field survey or baseline portion of the study. However, it is difficult
to judge how representative these four communities are of the entire area
of the study.
Obviously* many assumptions were made to develop the inventory
and it is difficult to determine the degree of bias each may introduce
into the results. An In-depth critical analysis of all assumptions
used, to see if they introduce bias or only uncertainty, would be useful.
Three major questions are addressed under this heading. The first
is the coverage of the inventory with resptct to materials and uses.
Although the choice of materials is limited, this limit is necessary to
fit within the scope of the project. Nevertheless, the inventory covers
the major materials used in structures, as well as other susceptible
materials. In this respect, the work is a substantial improvement over
previous studies.
The second question is whether the sample provides an adequate
basis for extrapolating damages for the city in which the sample is
taken; that is, does the sample provide an adequte basis for estimating
total materials exposed by building group ind material for tach city?
It is difficult to answer this question without examining the details of
the sampling design, which are apparently described in Rosenfield (1984)*l
The Mathtech report does include sensitivity analysis of damage estimates
based on the standard deviation in each sample city. However, sensitivity
analyses would not bi expected to uncover fundamental biases in assump-
tions. The errors in extrapolating to other cities may be considerable.
1 Rosenfield, G.H. (1984). "Spatial Sample Design for Building Materials
for Use with an Acid Rain Damage Survey." U.S. Geological Survey,
Reston, Va. In; R.S. Schmitt and H.J. Smollrt, eds. The Changing Role
of Computers in Public Agencies. Prestnted at the annual conference of
Urban and Regional Information Systems Assoc. August 1984.
-5-
-------
The question of extrapolation outside of the sample cities, that
is, to other urban areas, is discussed in Section 3,0. Extrapolation
Procedures.
B- Physical Damage Calculations
Do the physical damage functions accurately describe the
relation between acid deposition and materials damage?
1) Paint
As identified in the report, paint damage is by far the largest
component of total material damage costs. The damage function for paint
is based on paint film failure due to erosion measured by weight loss.
Peeling, flaking, blistering, soiling and fading are undoubtedly more
important forms of paint damage that triggers repainting.
The authors faced a serious lack of theory or data on the relation-
ship between acid deposition and paint physical damage measures other
than erosion. As a result, the erosion function was used to proxy all
paint damage functions. This is a highly uncertain and debatable
assumption that limits the reliability of the entire analysis. However,
if increased acid deposition also increases the rate of peeling and
cracking at a rate similar to the erosion of film thickness, and the time
chosen to repaint is a function of peeling and cracking (and that time
is about the same as for film erosion), the error may be small. If
the time to repaint is primarily based upon cracking or peeling rather
than film thickness, and if the increased rate of cracking and peeling
caused by acidic deposition is lower (or higher) than for film thickness,
then the damages will be lower (or higher) than reported. Research is
sorely needed to develop relationships between all types of paint damage,
acidic deposition and other environmental factors.
2} Mortar
The mortar damages estimated 1n the report are quite significant.
The authors identify the limitations of damage functions and composition
information upon which they base their estimates. The stone and mortar
damage functions are based upon information that is extremely limited
and, at present, of limited reliability. Further, the procedures used
suggest these damages may be overstated. From the assumptions used, it
takes about 50 years of exposure to current levels of deposition to
reduce the repointing service interval by ten years.
tt appears that the calculations used have assumed continuous past
exposure to current levels, which is probably Incorrect and substantially
overstates damages. Taking this into consideration would result in
reduced current and near-term future damages and/or require discounting
procedures as used in the stone analysis. If total past exposure and
-6-
-------
the current and near-term mortar damages were reduced by halft the total
damage numbers for all categories would be reduced by about 20 to 25
percent, which is substantial,
3} E cp n pme tjn c_C p n s j d e r at ions
There is little discussion in the report of possible econometric
problems with the damage functions. For example4 could there be an
omitted variable problem; i.e., could additional pollutants besides sulfur
dioxide (S02) and hydrogen ion (H } affect damages? Synergistic effects
also can occur, e.g., the combined effect of acid rain and road salts.
If so, the coefficients on SO^ and H could be inflated.
Of major concern is the fact that damage functions are all apparently
extrapolated from current ambient conditions to some background concentra-
tions (apparently zero for S02 and neutral pH for H*) and are generally
linear functions. Is this realistic, since many damage functions for
other adverse environmental effects are not linear? In fact, they
sometimes show thresholds below which damage is imperceptible. If this
is also the case here, the estimated damages may be too high.
C. Economic Damage Calculations
Are the assumptions about baseline maintenance practices appro-
priate for estimating incremental acid deposition costs?
1) Seneraj
Given the data available to the authors and the apparent current
lack of understanding of how maintenance decisions are actually made,
the economic damage calculations are reasonable. Although the use of
critical values (Table 4-5) is an acceptable means of calculating changes
in maintenance intervals, it nay be a strong simplifying assumption.
Overall, the calculations of economic damages art carefully done and rely
upon assumptions which, although stringent, do not appear to substantially
bias the estimates in either direction.
The benefit estimation approach that Is used in this analysis
involves the maintenance costs induced by acid deposition. However,
optimal maintenance strategies as defined may not be the optimal welfare
damage mitigation strategy. For example, If people choose to suffer
rather than undertake thi maintenance, then benefits are likely to be
overestimated. To whit extent will individuals and firms suffer unre-
paired damage from pollution?
Missing from this analysis is a good discussion of individual be-
havior as it relates to maintenance. In particular, when will individuals
perceive damage and act on repairs, and what repairs will they undertake?
-7-
-------
Do these assumed critical loss levels make any sense in terms of actual
behavior? How does the type of damage that will occur relate to that
found in the earlier studies of critical losses by Haynie and Martin
Marietta Environmental Systems?
A field study should be conducted that interviews consumers and
firms regarding repair actions. Since the behavioral response is critical
to the economic damage calculations, these values should be based on
solid information, not assumptions. The most important limitation (and
therefore research need) in the economic component of the analysis* is a
field study to improve the estimation procedure by determining how build-
ing owners perceive relevant damages; how they respond to damage through
changes in maintenance intervals, use of different materials, and so on-,
and how they value damage and response activities.
While the economic analysis can be substantially improved, it is not
currently among the major sources of error or bias in the overall
assessment.
2) Reasons for Repainting
Painted surfaces are the largest man-made exposed surfaces subject
to environmental damage. Therefore, perception, behavioral response,
and valuation are highly dependent upon the damage function for paint
and the method of calculating costs for damage and/or remediation. The
fundamental problem with using the Mathtech damage function as a basis
for computing damage costs 1s that paint erosion may seldom be the basis
for decisions to repaint. Mathtech hints on Page 4-13 of their report
that erosion is not an appropriate basis for repainting. On Page 2-37
of the 1984 draft Mathtech report "Economic Benefits of Reduced Acidic
Deposition on Common Building Materials; Methods Assessment", this is
stated more strongly;
The present damage functions for paint are not adequate
to characterize damage definitively. Peeling and cracking
of the painted surface are the primary forms of damage that
are experienced in the real world. These types of damage
are not adequately treated in the existing calculation.
This point is further emphasized in a survey conducted in 1%8
by BetterHomes and iardens on residential painting practice.* In the
section of the report on this survey dealing with the respondents' most
2 "Residential Paint Markets, 1968." A study of consumers by Setter
Homes and Gardens, conducted in cooperation with the National Paint,
Varnish* and Lacquer Association, Washington, DC,
-8-
-------
recent exterior painting activity, a table is presented giving the
reasons for the .most recent repainting. Of 1,106 total replies, 41
percent listed the predominant reason to repaint was that the old paint
was blistering and peeling. The second and third reasons, at 20 percent
each, were to protect the undersurface, and because the previous paint
was flat and drab. Erosion might be involved to a limited degree in
only two of the six specific reasons given.
From the above discussion, it appears that costs for acidic deposi-
tion to paint are probably overstated. Further, it is likely that damage
costs for other construction materials are also incorrect because of
faulty assumptions about the nature of the damage and response in the
form of maintenance practice to remedy the damage. For example, some
households repaint prior to the required need to change color or to
prepare a house for resale. In these cases, the incremental damage for
air pollution would be zero.
3) Behavior
If additional resoures were available, the Subcommittee would urge
the following changes: First, a better understanding is needed of the
behavioral response to paint erosion. When will consumers undertake
repainting due to erosion and what is the incremental effect of acid
deposition on these decisions? A second set of concerns pertain to
unrepaired damage. For consumers who eventually do repaint, what is the
value of the aesthetic loss from erosion before repainting? In addition,
some individuals may choose not to repaint at all and simply incur the
unrepaired damage, which will be less costly than the price of repainting.
A third class of concerns pertain to the economic cost of repainting.
Even with only modest additional effort, we could get a better estimate
of this cost. What fraction of households hire painters for exterior
work, and what is the rate they will pay? For consumers who do their own
repainting the calculation is more uncertain. The commercial cost sets
an upper bound on the painting cost, but how far below this amount one
should calculate is not clear. Consider a worker who earns $lO/hour
who faces commercial rates of $20/hour. He will choose to repaint even
if he greatly dislikes doing so, provided that the disutility per hour
of painting has a value less than $20. If he likes to paint, this value
may be below his market wage of $lO/hour. A "neutral" but probably
incorrect assumption would use the worker's wage rate as the painting
labor cost if it were below the cost charged by professional painters*
and would use the professional painters' wage otherwise.
4) Cosits_o_f Repai ntJM
Once the decision has been made to repaint a house, the question of
cost arises. Various figures are cited in the report, such as $1
per square foot {Page 5-6), the unit maintenance cost (Table 5-2), and
-9-
-------
data from various studies (Table 5-15). Does this information pertain to
interior painting, exterior painting, or both? Ue believe that more work
can be done in this area of the report.
For overall unit painting costs, the assumption is that these costs
should be reduced by 20 percent to account for painting done by homeowners.
This assumption may lead to an underestimation of damages for two reasons.
First, while the report cites evidence that SO percent of all architectural
paint is sold to non-professionals who do the painting themselves, do-it-
yourself activity seems much higher for interior painting. Do-it-yourself
painting of exteriors may be less than the 50 percent assumed in the re-
port. Second, the report bases its estimate of the labor cost of do-it-
yourself painting on the assumption that home owners' wages are the same
as those of professional painters. The relevant market wage is the
average wage of that class of workers owning their own homes, which may
be higher. Substantial literature exists on estimating the opportunity
cost of household leisure .time. Some references to this literature
might be used to justify the assumption that the opportunity cost of
leisure time is 50 percent of the relevant wage.
5) Silicate-Base Paint
The analysis of damages to painted surfaces is based on the assump-
tion that carbonate-base paint will be replaced with silicate-base paint.
But based on the discussion of the advantages of silicate-base paints, a
substantial fraction of surfaces initially painted with carbonate-base
paint would be repainted with silicate-base paint, in any case, indepen-
dently of acid deposition. Thus» the added cost of silicarte-base paints
estimated on page 5-10 of the Mathtech report ($42 million) is not rele-
vant if silicate-base paint would be the paint of choice in any case,
that it, independent of resistance to pollution. If this is valid, then
damages to painted surfaces are overestimated in this report. The report
examines the sensitivity of damage estimated to assumptions about the
percentage split between the two types of paint. It dots not examine
the sensitivity to changes in the split as carbonate-base paint surfaces
are repainted with silicate-base paints.
The report indicates that silicate-base paints are more expensive,
which is misleading. They are more expensive per gallon but not per
year, unless one only wants paint to last three years.
6) Stone Buildings
The damage estimate to stone by11 dings is bastd on an estimate of
replacement costs of about $20 per square foot. The basts of this esti-
mate is uncertain. Dots this refer to the cost of replacing stone facings
only? If so, it may be inappropriate to apply this number to many of the
stone buildings in the Inventory. The stone in most residences and many
-10-
-------
smaller commercial and public buildings is an integral part of the load
bearing external walls. Replacement of that stone may involve tearing
clown and replacing the structure,
D. Extrapo1 at i on P rocedu res
Is a credible method used for extrapolating from the four
case study cities to other major urban areas in the Northeast
and North Central United States?
1) General
Given the data available, the section on Extrapolation Procedures
in the report seems to provide plausible procedures for other cities in
the 17-state area. The reasonableness of the results are conditioned
upon the acceptances of the calculated physical and economic damages for
the case-study cities. Because it is an extrapolation, rather thin a
repetition of the earlier approach to all 113 cities, the estimates are
necessarily imperfect.
To improve the readers' confidence in the extrapolations, it would
be helpful if the authors could take three case-study cities to extrapolate
to the fourth city. For example, could the experiences of tht three other
cities be used to project the Portland experience, and how well would
this projection compare with what was actually found?
2) The Mix of Cities
This analysis extrapolates the experiences in New Haven, Portland,
Pittsburgh* and Cincinnati. How were these cities chosen? Is there any
indication that they provide a reliable basis for extrapolating to other
cities? What about large cities, such as New York, Boston, and Philadel-
phia? Having a reasonable range is important given the fact that economic
damages to materials vary by a factor of almost 100 from the estimate
for Portland to the estimate for Pittsburgh. Even on a per-tapita basis,
the discrepancy is almost a factor of seven, which suggests that a good
mix of cities is very important to provide a reliable basis for comparison.
Thus, more justification should be given for the four cities chosen.
3) Materials Use by Region
The dita from the sample cities varies substantially in materials
use by region. The four-city sample does not provide an adequate basis
for understanding this variation across the whole geographic arta and
range of city sues covered by the extrapolation. Given the importance
in the overall estimates of large urban areas, this is a shortcoming of
the report.
-11-
-------
A key part of the sample is the extrapolation of wall area by ma-
terial and building type to five census divisions. The Mathteth report
states on page 6-4 that this was based on sample data from the case study
inventory and from Department of Energy materials distribution informa-
tion. However, without having reviewed Lipfert (1985),3 -jt is difficult
for the Subcommittee to fully evaluate the data of the extrapolation
procedures adequately.
4, ADDITIONAL ISSUES
A. Use of Results in the Rulemaking Process
Are the analyses useful input for the S02 NAAQS rulemaking
process?
The calculations for the four sample cities show the potential for
significant damages to exterior materials. The extrapolation estimates
show the possibility that area-wide damages in this category, when com-
bined with estimates of damage in other categories, may be comparable in
magnitude to control costs. This comparison might suggest that this work
is useful input to the secondary National Ambient Air Quality Standards
for S02*. however, in view of the great uncertainty in paint damage costs,
the total costs from acidic deposition should not be used as part of any
rulemaking.
A better understanding of the types of damage induced by acidic de-
position, and the response of the public to this damage, is needed before
the economic estimates derived from such studies can be considered free
of extensive uncertainties and biases. Given the substantial uncertain-
ties involved, the Subcommittee urges that EPA consider additional work
in areas where it is feasible to improve our knowledge {e.g., economic
damages).
B. A More Credible Approach
Do these analyses represent a more credible approach to ma-
terials damage estimation than the supply/demand model ap-
proach (Mathtech 1982) that is currently incorporated into
the draft RIA?
Lipfert, F.V. (1985). Report in preparation by Brookhaven National
Laboratory.
-12-
-------
There is no simple answer to this question due to the substantial
differences in method, data, geographic coverage, and coverage of eco-
nomic sectors and activities between this report and the earlier document
(Mathtech 1982). Also the two reports analyze the effects of air pollu-
tion that is measured differently. Thus* these reports are not alterna-
tive measures of the same economic phenomenon, and are therefore not
directly comparable.
The principles that underlie the present report are more easily
grasped than the method of indirect estimation developed for the 1982
Mathtech document. Yet the methods used in this latest report are con-
ceptually correct only ynder restrictive assumptions concerning the
absence of material substitution, direct utility losses, etc. The 1982
report derives estimating techniques from conceptually correct economic
models of behavior. The magnitude of the biases introduced by the re-
strictive assumptions In the 1986 report are unknown*
The geographical coverage of the two reports is also quite different.
The 1982 report is national in scope while the 1986 report covers only
urban areas within the Northeastern part of the United States. In the
1982 report, separate estimates were developed for the household sector
and for the commercial/industrial sectors, for the household sector,
the 1982 report covered all cleaning and maintenance activities and
related losses in utility and welfare, including those involving interior
soiling, etc. However the estimates were derived implicitly from data
an differences in household expenditure patterns across areas that dif-
fered according to air pollution levels. The present report focuses
only on exterior damages to residences and related structures, ignoring
losses in utility, damages to residential interiors, and increased costs
due to mitigating activities. On the other hand, the present report Is
based on quantitative data on the effects of pollution on matt rials.
As for the commercial/industrial sector, the 1982 report covered
only a limited number of industrial sectors, but it included all forms
of costs-increasing damages due to pollution, rather than just damages
to exterior structures. The present report includes all commercial/
industrial structures but 1s limited to exterior structural damages in
urban areas.
The two reports should be considered complementary to tach other.
The present report emphasizes the physical awchanisms by which pollution
causts damages, while thi 1982 report focusts on the behavioral responses
of economic agents to these effects. The behavioral approach therefore
encompasses a wider range of effects and responses than those explicitly
modeled in the physical mechanisms method. However, the approach focusing
on behavioral responses yields results that are linked more directly to
the underlying changes In welfare or utility which the studies art at-
tempting to measure.
-13-
-------
The present approach identifies real physical mechanisms by which
pollution affects individuals. The 1982 report shows that individuals
appear to perceive these effects and respond in economically meaningful
ways which can be interpreted as reflecting losses in economic well-being.
The Subcommittee believes that the results and limitations of both studies
should be included in any assessment of the economic implications of
physical damages due to air pollution.
-14-
-------
APPENDIX A
-------
APPENDIXA
ADDITIONAL SPECIFIC COMMENTS
1. Although the report lists many of the omissions, errors, and passible
biases that have not been addressed, it does little to indicate which
are more important than others* This may give the perhaps false
impression that these limitations approximately net out. Any evidence
on the relative magnitude of the errors* omissions, and biases would
increase the usefulness of the overall data,
2. A subsection in either section 1 or section 5 should list all the
biases used in the methodology and make an informed guess as to
whether the net effect of all the biases is to underestimate or to
overestimate the total economic damages from acid deposition. Since
some damages are purposely omitted, e.g. infrastructure systems, it
would help to provide some guidance as to whether these omitted
damages are of the same order of magnitude as those which art esti-
mated.
3. Section 1 provides a useful overview and summary of the entire report.
The limitations subsection in this section (and in Sections 4 and 6}
help put the accuracy and coverage of the estimates in perspective.
Page 1-6 does not explain why national estimates of damages are not
provided, given that the NAAQ5 are national. If it is true that
damages are negligible outside tht 17-state region, the report should
say so clearly and document this finding.
4. Page 1-8 does not explain why tht damages to infrastructure systems
were not estimated. Does this introduce a serious bias? Why was
this class of damages omitted?
5. The distinction between regional and local $02 sources on Page 1-11
is not clearly explained and should be for exposition*! purposes,
6, The last suggestion for future research on Page 1-19 (calculate
confidence intervals) is the logical next task and would provide a
useful extension to the estimates in this report* especially the
ranges of estimates reported. The ranges are useful, byt they are
not a substitute for confidence intervals, as the authors are careful
to explain.
7. The recommendations for future research could include the effects on
art objects, since it 1s difficult to assess costs of this type of
damage. One potentially significant underestimate: could values for
works of art that cannot be repaired or replaced to be e^lvalent to
the original pieces? Aside from irreplaceable works of art, the
procedures may be resulting in a "mild" upward bias in the estimates.
A-l
-------
8. Section 2 provides a competent summary of benefit estimation and a
description of the methodology used in the report. Figure 2-8 is
quite useful. Although the decision to allow only one reaction to
acid deposition, for example, additional maintenance provides a
reasonable approximation, two additional implications of this as-
sumption might usefully be explained. First, this behavioral re-
sponse assumes that there is no trade-off involved in selecting
the maintenance interval, i.e., consumers chose the same level of
building quality no matter how expensive the maintenance becomes to
keep up that level of quality. Obviouslyt this is a strong assump-
tion. Second, the approach assumes that maintenance is sufficiently
frequent to avoid any damage to the underlying materials, e.g., wood
siding, bricks,
9. The reasons for the zero entries in the materials inventory probably
should be documented, perhaps in an appendix,
10. Page 4-36 shows substantial differences between the estimates of
Haynie and Martin Marietta Environmental Systems concerning initial
paint thickness and critical loss. However, the sensitivity analysis
conducted at the end of Section 5 does not include sensitivity to
variations in these key assumptions.
11. There is uncertainty in how the upper and lower bound estimates were
derived. The report lists, for example, several factors that could
either increase or decrease the paint estimates. These possible
errors appear to be multiplicative. This is important to the deter-
mination of the spread of reasonable numbers and should be more
completely documented.
12. The source of the data in Section 5 used for regional adjustments
to cost and price indices is not described.
13. Without access to Lipfert's report, it is not possible to evaluate
the regressions used on Pages 6-8 and 6-16 for the commercial/
industrial category.
14. Tables in Section 6 refer to the Northeastern Census Division. Is
it not the New England Census Division?
15. Why was Michigan, especially Detroit, excluded from the extrapola-
tion?
16. Because the Pittsburgh extrapolation contained the poorest perfor-
mance of the four test cities, and 50 percent of the damages occur
in the three largest metropolitan areas, more data should be gath-
ered in these three areas. Approximating the building counts
through regression equations is the assumption which we question
most. This component of the extrapolation is likely to be a major
source of error for large cities.
A-2
-------
APPENDIX B
-------
APPENDIX B
CITATIONS OF REVIEW DOCUMENTS
1. "A Damage Function Assessment of Building Materials; The Impact
of Acid Deposition." Mathtech, Inc. Hay 1986* Final Report.
2. "Economic Benefits of Reduced Acidic Deposition on Common Build-
ing Materials: Methods Assessment." Mathtech, Inc. June
1984. Draft Final Report,
3. "Material Effects Assessment." Draft of the materials section of
the 1985 NAPAP Assessment (unpublished) (National Acid Preci-
pitation Assessment Program).
4. "Economic Damages to Building Materials Exposed to Add Deposi-
tion." T.J. Lareau et al.
I, "Derivation of Metallic Corrosion Damage Functions for Use in
Environmental Assessments." Brookhaven National Laboratory.
April 1985.
6. "Atmospheric Acid Deposition Damage to Paints," Fred Haynfe.
U.S. EPA Report EPA/600/H-85/019. January 1986.
7. "Benefit Analysis of Alternative Secondary NAAQS for Sulfur
Dioxide and TSP." Volume 1. U.S. EPA Report EPA-4SQ/S-83-
OOla. August 1982.
B-l
-------
APPENDIX C
-------
APPENDIXC
CHARGE TO THE CASAC MATERIAL DAMAGE REVIEW SUBCOMMITTEE
The Mathtech report (1986), together with the supporting documents
are to be reviewed to determine whether the methods ire credible and
whether the data are appropriate for estimating materials damage from
acid deposition In a l?-state arei of the United States.
The following components of the analysis are to be assessed:
t Mater1 a1s Inventpry - Does the inventory provide i re-
pristntative sample of the distribution of materials in
urban areas that can be used to extrapolate to other urban
areas?
* Pamaje Functions r Do the physical damage functions accu-
ratefy describe the relation between acid deposition and
materials damage?
t E conomlc Damage Cat cu1 at ions - Are the assumptions about
base11ne ma1ntenaneipractices appropriate for estimating
incremental acid deposition costs?
* Extrapolation - Is i credible method used for extrapolating
the four case study cities to other major urban areas in
the Northeast and North Central United States?
Based on these assessments:
Are the analysts useful input for tht SO? NAAQS rulemtking
proctss?
Oo these analyses represent a more credible approach to
materials damage estimation than the supply/demand model
approach (Mathtech 1982} that is currently incorporated into
the draft Regulatory Impact Analysis (RIA)?
C-l
------- |