UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON B.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
September 24, 2009
EPA-SAB-09-018
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: Science Advisory Board (SAB) Advisory on EPA's draft Guidelines for
Preparing Economic Analyses (2008)
Dear Administrator Jackson:
EPA's National Center for Environmental Economics requested that the SAB
review EPA's draft Guidelines for Preparing Economic Analyses (2008). The
Guidelines, originally issued in 2000 and recently updated, represent Agency policy on
the preparation of economic analysis required by legislative and administrative mandates
and are intended to provide technical guidance to analysts on the economic analysis of
environmental policy. The SAB Environmental Economics Advisory was impressed
with many facets of the updated Guidelines. We applaud EPA for a number of carefully
revised chapters and substantively improved coverage of many topics. In the enclosed
report, we provide responses to EPA's charge questions and recommendations for
additional improvements. In this letter, we provide highlights of our overarching
comments.
The current draft of the Guidelines could be improved by clearly identifying
EPA's role and discretion in setting environmental policy. Specifically, policy options
are described in the Guidelines in a manner that could allow the reader to infer that EPA
has the discretion to choose from a variety of policy instruments (e.g., regulations or
taxes) to achieve environmental targets. In reality, of course, only the legislative branch
has the power to tax, subsidize or assign liability, and both the Clean Water Act and the
Clean Air Act very clearly specify what kinds of regulations EPA may promulgate. The
Guidelines should make clear that while economic analysis can identify superior policy
options, EPA's legal authority defines its menu of choices. This might be done quite
effectively by using examples from legislation to make the limitations concrete.
In addition to clarifying EPA's role in policy, the Guidelines should be grounded
in the realities of information and political constraints, as well as market distortions,
-------
either market induced or created by government interventions — so named "second best"
conditions. For example, the theory section focuses on first-best policy choices but this
framework is nearly irrelevant to contemporary water pollution problems. The section on
subsidies does not acknowledge that many subsidies in agriculture, car manufacturing, oil
and gas, etc. are not designed to correct externalities but may in fact worsen them.
Examples, such as the perverse incentives created by federal subsidies for corn-derived
ethanol, are needed to illustrate these issues.
The Guidelines provide scant coverage of the long-standing issue in benefit-cost
analysis of valuing the benefits of protecting ecological systems and services.
Monetizing ecosystem services remains an area of significant challenge. The Guidelines
should discuss situations where "non-monetized" benefits are expected to be a significant
portion of the regulatory outcome including advice for practitioners in this case and
noting that adherence to formal dollar-based benefit-cost analysis can lead to incorrect
efficiency signals. We note that the recent SAB report on Valuing the Protection of
Ecological Systems and Services (2009) provides useful insights on this topic.
The literature cited in the Guidelines needs to be updated. In its coverage of
economic valuation methods for benefits analysis, the Guidelines did not incorporate
numerous new studies using revealed preference methods, stated preference, approaches
combining the two, and experiments in the lab and field. Other areas in which the
literature needs updating include mortality benefits valuation, empirical work on the
limited effectiveness of voluntary approaches (without financial incentives) and water
quality trading.
The Guidelines could also be strengthened with case studies. For example, the
Guidelines identify the basic steps involved in "benefits transfer" (using values of
environmental quality estimates for one location to value changes at another), but readers
would benefit substantively from a concrete real world example. The Guidelines
enumerate a step-by-step approach for economic impact analysis but again, readers would
benefit from a specific example from EPA's own experience. The Guidelines' discussion
of environmental equity impacts would be greatly enriched with a case study.
Finally, the Guidelines are focused on economic analysis needed for "traditional"
environmental problems, e.g., chemical releases from point sources to air and water.
Given the emergence of climate change as the preeminent environmental threat, EPA will
need more complex, interdisciplinary analysis to address greenhouse gas mitigation
including information from the bio-physical sciences, economics and atmospheric
sciences. Computable general equilibrium (CGE) models, a subject covered well in the
Guidelines, will be of critical importance but CGE models will likely be wedged in a
portfolio of models tracking complex processes. EPA's greenhouse gas lifecycle analysis
of various fuels (required by Congress in the Energy independence and Security Act of
2007) is an early example of the daunting analytic challenges associated with forecasting
greenhouse gas emissions under various policies. The Guidelines should anticipate a
changing role for economics amidst the extraordinary complexity posed by climate
change and other global processes.
11
-------
By providing thorough and consistent technical advice regarding the application
of benefit cost analysis to environmental problems, the Guidelines significantly elevate
the quality and transparency of the information upon which environmental decisions are
made. We again applaud EPA for developing these Guidelines and the Agency's
commitment to continually revise and improve them. Indeed, we believe these
Guidelines could serve as a successful model for all state and federal agencies who
undertake benefit-cost analysis in support of environmental decision making. We greatly
appreciate the opportunity to provide advice on this draft of the Guidelines and look
forward to the Agency's response.
Sincerely,
/Signed/
Dr. Deborah L. Swackhamer, Chair
EPA Science Advisory Board
/Signed/
Dr. Catherine Kling, Chair
SAB Environmental Economics Advisory
Committee
Enclosure
in
-------
U.S. Environmental Protection Agency
Science Advisory Board
Environmental Economics Advisory Committee
CHAIR
Dr. Catherine Kling, Professor, Department of Economics, Iowa State University, Ames,
IA
MEMBERS
Dr. Anna Alberini, Associate Professor, Department of Agricultural and Resource
Economics, University of Maryland, College Park, MD
Dr. Michael Greenstone,* 3M Professor of Environmental Economics, Massachusetts
Institute of Technology, Boston, MA. On leave to the Council of Economic Advisors
(2009).
Dr. James K. Hammitt, Professor, Center for Risk Analysis, Harvard University, Boston,
MA
Dr. Madhu Khanna, Professor, Department of Agricultural and Consumer Economics,
University of Illinois at Urbana-Champaign, Urbana, IL
Dr. John List, Professor, Department of Economics, University of Chicago, Chicago, IL
Dr. James Opaluch, Professor, Department of Environmental and Natural Resource
Economics, College of the Environment and Life Sciences, University of Rhode Island,
Kingston, RI
Dr. George Parsons, Professor, Department of Economics, College of Marine and Earth
Studies, University of Delaware, Newark, DE
Dr. James Shortle, Professor, Agricultural Economics and Rural Sociology, Pennsylvania
State University, University Park, PA
Dr. Laura Taylor, Director, Center for Environmental and Resource Economics and
Department of Agriculture and Resource Economics, North Carolina State University,
Raleigh, NC
Dr. Peter J. Wilcoxen, Associate Professor, Economics and Public Administration,
Syracuse University, Syracuse, NY
* In order to serve on the Council of Economic Advisors, Dr. Greenstone resigned from the Environmental
Economics Advisory Committee in February of 2009 and hence was not involved in writing this Advisory
thereafter.
IV
-------
Dr. David Zilberman, Professor, Agriculture Resource Economics Department, College
of Natural Resources, University of California, Berkeley, CA
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, Environmental Protection Agency,
Washington, D.C.
-------
U.S. Environmental Protection Agency
Science Advisory Board
BOARD
CHAIR
Dr. Deborah L. Swackhamer, Professor and Co-Director, Environmental Health
Sciences and Water Resources Center, University of Minnesota, St. Paul, MN
SAB MEMBERS
Dr. David T. Allen, Professor, Department of Chemical Engineering, University of
Texas, Austin, TX
Dr. John Balbus, Chief Health Scientist, Environmental Health Program, Environmental
Defense Fund, Washington , DC
Dr. Gregory Biddinger, Coordinator, Natural Land Management Programs, Toxicology
and Environmental Sciences, ExxonMobil Biomedical Sciences, Inc., Houston, TX
Dr. Timothy Buckley, Associate Professor and Chair, Division of Environmental Health
Sciences, College of Public Health, The Ohio State University, Columbus, OH
Dr. Thomas Burke, Professor, Department of Health Policy and Management, Johns
Hopkins Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD
Dr. James Bus, Director of External Technology, Toxicology and Environmental
Research and Consulting, The Dow Chemical Company, Midland, MI
Dr. Deborah Cory-Slechta, Professor, Department of Environmental Medicine, School
of Medicine and Dentistry, University of Rochester, Rochester, NY
Dr. Terry Daniel, Professor of Psychology and Natural Resources, Department of
Psychology, School of Natural Resources, University of Arizona, Tucson, AZ
Dr. Otto C. Doering III, Professor, Department of Agricultural Economics, Purdue
University, W. Lafayette, IN
Dr. David A. Dzombak, Walter J. Blenko Sr. Professor of Environmental Engineering,
Department of Civil and Environmental Engineering, College of Engineering, Carnegie
Mellon University, Pittsburgh, PA
Dr. T. Taylor Eighmy, Vice President for Research, Office of the Vice President for
Research, Texas Tech University, Lubbock, TX
VI
-------
Dr. Baruch Fischhoff, Howard Heinz University Professor, Department of Social and
Decision Sciences, Department of Engineering and Public Policy, Carnegie Mellon
University, Pittsburgh, PA
Dr. James Galloway, Sidman P. Poole Professor of Environmental Sciences
Associate Dean for the Sciences, College and Graduate School of Arts and Sciences,
University of Virginia, Charlottesville, VA
Dr. John P. Giesy, Professor and Canada Research Chair, Veterinary Biomedical
Sciences and Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan,
Canada
Dr. James K. Hammitt, Professor, Center for Risk Analysis, Harvard University,
Boston, MA
Dr. Rogene Henderson, Senior Scientist Emeritus, Lovelace Respiratory Research
Institute, Albuquerque, NM
Dr. James H. Johnson, Professor and Dean, College of Engineering, Architecture &
Computer Sciences, Howard University, Washington, DC
Dr. Bernd Kahn, Professor Emeritus of the Nuclear and Radiological Engineering
Program and Director, Environmental Radiation Center, School of Mechanical
Engineering, Georgia Institute of Technology, Atlanta, GA
Dr. Agnes Kane, Professor and Chair, Department of Pathology and Laboratory
Medicine, Brown University, Providence, RI
Dr. Meryl Karol, Professor Emerita, Graduate School of Public Health, University of
Pittsburgh, Pittsburgh, PA
Dr. Catherine Kling, Professor, Department of Economics, Iowa State University,
Ames, IA
Dr. George Lambert, Associate Professor of Pediatrics, Director, Center for Childhood
Neurotoxicology, Robert Wood Johnson Medical School-UMDNJ, Belle Mead, NJ
Dr. Jill Lipoti, Director, Division of Environmental Safety and Health, New Jersey
Department of Environmental Protection, Trenton, NJ
Dr. Lee D. McMullen, Water Resources Practice Leader, Snyder & Associates, Inc.,
Ankeny, IA
Dr. Judith L. Meyer, Distinguished Research Professor Emeritus, University of
Georgia, Lopez Island, WA
vn
-------
Dr. Jana Milford, Professor, Department of Mechanical Engineering, University of
Colorado, Boulder, CO
Dr. Christine Moe, Eugene J. Gangarosa Professor, Hubert Department of Global
Health, Rollins School of Public Health, Emory University, Atlanta, GA
Dr. M. Granger Morgan, Lord Chair Professor in Engineering, Department of
Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA
Dr. Duncan Patten, Research Professor, Department of Land Resources and
Environmental Sciences, Montana State University, Bozeman, MT
Mr. David Rejeski, Director, Foresight and Governance Project, Woodrow Wilson
International Center for Scholars, Washington, DC
Dr. Stephen M. Roberts, Professor, Department of Physiological Sciences, Director,
Center for Environmental and Human Toxicology, University of Florida, Gainesville, FL
Dr. Joan B. Rose, Professor and Homer Nowlin Chair for Water Research, Department
of Fisheries and Wildlife, Michigan State University, East Lansing, MI
Dr. Jonathan M. Samet, Professor and Chair, Department of Preventive Medicine,
University of Southern California, Los Angeles, CA
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography,
Savannah, GA
Dr. Jerald Schnoor, Allen S. Henry Chair Professor, Department of Civil and
Environmental Engineering, Co-Director, Center for Global and Regional Environmental
Research, University of Iowa, Iowa City, IA
Dr. Kathleen Segerson, Professor, Department of Economics, University of
Connecticut, Storrs, CT
Dr. Kristin Shrader-Frechette, O'Neil Professor of Philosophy, Department of
Biological Sciences and Philosophy Department, University of Notre Dame, Notre Dame,
IN
Dr. V. Kerry Smith, W.P. Carey Professor of Economics , Department of Economics ,
W.P Carey School of Business , Arizona State University, Tempe, AZ
Dr. Thomas L. Theis, Director, Institute for Environmental Science and Policy,
University of Illinois at Chicago, Chicago, IL
Dr. Valerie Thomas, Anderson Interface Associate Professor, School of Industrial and
Vlll
-------
Systems Engineering, Georgia Institute of Technology, Atlanta, GA
Dr. Barton H. (Buzz) Thompson, Jr., Robert E. Paradise Professor of Natural
Resources Law at the Stanford Law School and Perry L. McCarty Director, Woods
Institute for the Environment, Stanford University, Stanford, CA
Dr. Robert Twiss, Professor Emeritus, University of California-Berkeley, Ross, C A
Dr. Thomas S. Wallsten, Professor and Chair, Department of Psychology, University of
Maryland, College Park, MD
Dr. Lauren Zeise, Chief, Reproductive and Cancer Hazard Assessment Branch, Office
of Environmental Health Hazard Assessment, California Environmental Protection
Agency, Oakland, CA
SCIENCE ADVISORY BOARD STAFF
Mr. Thomas Miller, Designated Federal Officer, Environmental Protection Agency,
Washington, D.C.
IX
-------
NOTICE
This report has been written as part of the activities of the EPA Science Advisory Board
(SAB), a public advisory committee providing extramural scientific information and
advice to the EPA Administrator and other officials of the Environmental Protection
Agency. The SAB is structured to provide balanced, expert assessment of scientific
matters related to problems facing the Agency. This report does not reflect the policies
and views of the EPA nor of other agencies in the Executive Branch of the federal
government. Mention of trade names or commercial products do not constitute a
recommendation for use. Reports of EPA SAB are posted at http://www.epa.gov/sab.
-------
Enclosure
Advisory on EPA's Guidelines for Preparing Economic Analyses
by the
Science Advisory Board
Environmental Economics Advisory Committee
Executive Summary ES-1
Crosscutting Issues 1
Question 1: Policy Options 5
Question!: Baselines 9
Question 3: Discounting 11
Question 4: Benefits 15
Question 5: Mortality Risk Valuation 20
Question 6: Social Costs 22
Question 7: CGE Models 24
Question 8: Distributional Analyses 26
Question 10: Economic Literature 28
Question 11: Omissions 29
Appendix: Compilation of Line by Line Comments from Individual Members 30
References 41
-------
Executive Summary
The goal of EPA's Guidelines for Preparing Economic Analyses is to establish a
sound scientific framework for performing economic analyses of environmental
regulations and policies. Originally issued in 2000, the Guidelines are intended to reflect
Agency policy and guide Agency practice on the preparation of economic analyses.
EPA's National Center for Environmental Economics recently updated the Guidelines
(September 2008) and asked the Science Advisory Board for its input. The SAB
Environmental Economics Committee (EEAC) met on October 23-24, 2008 to address
the Agency's charge questions on the Guidelines. The Background section of this report
discusses some crosscutting issues. Below we provide specific responses to each charge
question.
1. Do the published economic theory and empirical literature support the statements in
the guidance document on the merits and limitations of the different regulatory and
non-regulatory approaches discussed in Chapter 4: Regulatory and Non-Regulatory
Approaches to Pollution Control?
Yes, the chapter is supported by economic literature; however it too closely
mimics textbook expositions of environmental economics. In order to make it
more useful for EPA analysts, we recommend that EPA clarify and discuss the
specific role that EPA has in policy design and implementation and provide
guidance for economic analysis done specifically within that context. We also
recommend a better distinction between efficiency and cost-effectiveness,
improvements to the discussion of "cap and trade," a better definition of design
standards and technology based performance standards and the inclusion of recent
literature on voluntary approaches and the observability of information.
2. Do the published economic theory and empirical literature support the statements in
the guidance document on the consideration of the baseline discussed in Chapter 5:
Establishing a Baseline?
Yes, this chapter provides very comprehensive guiding principles for specifying
the baseline scenario to identify the incremental benefits and costs associated with
a policy. We recommend that EPA consider the key dimensions of the economic
analysis and any phenomena in the baseline about which there is uncertainty and
to construct two or three (rather than more) scenarios that can provide
benchmarks for policy analysis.
3. Do the published economic theory and empirical literature support the statements in
the guidance document on the treatment of discounting benefits and costs discussed in
Chapter 6: Discounting Future Benefits and Costs in the following circumstances:
a. Are the descriptions of fundamental social discounting approaches,
conceptual conclusions and recommendations consistent with the
appropriate economic literature on social discounting? Are the correct
ES-1
-------
conclusions drawn from the respective literatures on discounting for
public projects (government spending) and discounting for regulations
(government-mandated spending) ?
Yes, the descriptions of fundamental discounting approaches,
conceptual conclusions and recommendations are consistent with
the appropriate economics literature. We do not believe there are
significant differences in the approach to discounting for public
projects (government spending) and discounting for regulations
(government mandated spending). We suggest that the chapter
should begin with acknowledgment of the many controversies and
complications associated with discounting and orient readers to
where in the chapter these are discussed.
b. The Guidelines do not draw a firm conclusion on the extent to which
shadow price of capital adjustments are likely to be necessary for most
EPA policy analyses. The issue depends greatly on the elasticity of capital
supply and EPA plans to pur sue additional research on this issue, as
noted in the draft Guidelines. Does EPA 's conclusion reflect the sense of
the literature or can a firmer conclusion be drawn? Does the Committee
have suggestions regarding situations where these adjustments would be
necessary or unnecessary?
Yes, the EPA conclusion does reflect the sense of the literature. As
noted in the chapter, the shadow-price of capital approach is
theoretically correct, but the quantitative significance of adjusting
for this shadow price is critically dependent on the extent to which
EPA regulations displace other investment. In an economy that is
open to foreign investment, there may be minimal displacement
and so adjustment for the shadow price of investment is negligible.
We are pleased to learn that EPA is investigating the elasticity of
investment to environmental regulation.
c. While EPA concludes that a rate of 3% is generally consistent with
estimates from low-risk government securities, the Agency would like to
more firmly establish a rigorous basis for a consumption-based rate.
What data and methods would the committee suggest EPA pursue ?
We are unable to suggest better data or methods for estimating a
consumption-based discount rate, and doubt that alternative
credible estimates would differ dramatically from the 3% real rate
specified by OMB in Circular A-4. Given the benefits of
harmonization of parameter values among federal agencies, we do
not encourage EPA to move away from this rate.
ES-2
-------
d. Chapter 6 recommends adopting an approach to long term discounting
based on the work of Newell & Pizer (2003). While EPA recognizes that
data may not clearly support a particular statistical model over other
alternatives (e.g., random walk vs. mean-reverting), the Chapter
concludes that the recommended approach is an improvement over
constant discounting. Does the committee believe this is a reasonable
conclusion from the economics literature? More specifically, is the
recommendation to use a random walk model as a default reasonable
given the state of the literature?
No, we believe that calculating present values using the Newell &
Pizer approach can be used as one of several alternatives
(complemented by appropriate caveats and discussion of the
theoretical and empirical issues), but we do not believe it should be
relied upon exclusively for reasons provided in the Advisory.
e. EPA has struggled with the question of the length of time an analysis
should capture and has arrived at some practical recommendations (see
Section 6.1.6.3 and 6.4). Are these recommendations consistent with good
economic practices? Does the committee have additional
recommendations or insights on this subject?
Yes, the recommendations are generally consistent with good
practice. In considering the time horizon an analysis should cover,
there is no general answer beyond the answer to the question of
what consequences to include: those that may have a quantitatively
significant effect on the conclusions of analysis.
4. Do the published economic theory and empirical literature support the statements in
Chapter 7: Analyzing Benefits on the merits and limitations of different valuation
approaches for the measurement of social benefits from reductions in human health
risks and improvements in ecological conditions attributable to environmental
policies?
This chapter provides good coverage of the main categories of benefits and the
methods used for their estimation. However, it fails to capture a significant
amount of recent literature on recreation demand models, combining revealed and
stated preference, validity and reliability, valuing mortality and morbidity and
ecosystem services. In particular, we urge the Agency to vastly expand its
guidance on characterizing and valuing non-monetized ecosystem systems and
services. We also recommend expanding the discussion of evaluating studies and
data.
5. Chapter 7 includes a brief discussion of the Agency's current approach to mortality
risk valuation with more details provided in Appendix B. These sections will be
updated when the Agency concludes its efforts to update its mortality risk valuation
ES-3
-------
approach. In the interim, are the discussions provided in Chapter 7 and Appendix B
clear and balanced?
We will refrain from detailed comments on EPA's approach to valuing mortality
risk until the Agency's update is complete. In the interim, we recommend EPA
consider expanding its literature review and discontinue use of old, discredited
wage-risk studies.
6. Does Chapter 8: Analyzing Costs contain an objective and reasonable presentation of
the published economic theory, empirical literature, and analytic tools associated
with estimating social costs?
Yes. As a suggestion for improvement, we recommend covering non-competitive
markets where results can be significantly different. Our detailed comments offer
suggestions for other revisions, such as examining three cases: single market
analyses, multiple market analyses and general equilibrium analyses.
7. Does Chapter 8 contain an objective, balanced and reasonable presentation of the
published economic theory, empirical literature, and analytic tools associated with
computable general equilibrium (CGE) models? Is the description of the relevance of
these models for economic analyses performed by the EPA reasonable?
Yes. We recommend discussion of the parameterization of CGE models as well
as a few minor revisions as discussed in our detailed comments.
8. Does Chapter 9: Distributional Analyses: Economic Impact Analyses and Equity
Assessment contain an objective and reasonable presentation of the measurement of
economic impacts, including approaches suitable to estimate impacts of
environmental regulations on the private sector, public sector and households? This
discussion includes, for example, the measurement of changes in market prices,
profits, facility closure and bankruptcy rates, employment, market structure,
innovation and economic growth, regional economies, and foreign trade.
Yes, this chapter contains an objective presentation of many aspects of economic
impact analyses. In our detailed comments, we offer suggestions for minor
improvements.
9. Does Chapter 9 contain a reasonable presentation and set of recommendations on the
selection of economic variables and data sources used to measure the equity
dimensions identified as potentially relevant to environmental policy analysis?
Yes, the main items in assessment equity issues are correctly identified. One
limitation is that the main distributional issues discussed in the chapter relate to
costs, not benefits. Although it may be more difficult to identify the distribution
of benefits, it would seem appropriate to at least represent the ideal case as one in
which the equity and impacts associated with both benefits and costs are
ES-4
-------
considered. We suggest that EPA consider creation of a website to catalog all the
data sources.
10. Appendix A: Economic Theory was prepared for those readers who wished to have a
better understanding of the economic foundations underlying benefit cost analyses.
Does Appendix A summarize the relevant literature in an objective and meaningful
way? Are there topics that warrant (more) discussion in this appendix that were
otherwise missed?
The Appendix provides a good discussion of core economic principles. As
suggestions for improvements, we recommend distinguishing between stock and
flow pollutants and inclusion of the concept of "user costs."
11. Please identify and enumerate any inconsistencies you may find across chapters and
other issues/topics on which we should provide further elaboration. Also, please
identify any definitions provided in the new glossary that are inaccurate or that
otherwise need revision.
In general, we would like to see broader discussion of a number of cross-cutting
issues. For example, we'd like to see a discussion of the need for transparency in
making assumptions and judgments and a discussion of the ways in which biases
and errors will matter the most. The Guidelines should more frankly
acknowledge the "second best" world of most environmental policy problems due
to information constraints, political constraints, imperfect competition and market
distortions created by taxes and other government policies. The Guidelines sorely
need case studies and examples to illustrate and make concepts concrete. We
underscore our recommendation to provide guidance to analysts for exploring a
range of ecological indicators and conceptual models of ecosystems and services.
ES-5
-------
Background
EPA's National Center for Environmental Economics (NCEE) oversees the
Agency's economic analysis of environmental issues. NCEE guides research and
development on economic methods, produces EPA's major economic reports and issues
guidance for performing economic analysis at the Agency. In this later role, NCEE
issued Guidelines for Performing Economic Analysis in 2000. These Guidelines are
meant to provide guidance on economic analysis for those performing or using such
analysis, including policy makers, analysts and contractors providing economic reports to
the EPA. In 2008, NCEE updated the Guidelines to incorporate the most recent advances
in environmental economics and asked that the SAB EEAC review the revised document.
The SAB EEAC met on October 23 - 24, 2008 to review draft Guidelines (September
2008) and respond to NCEE's specific charge questions. This face-to-face meeting was
followed by a public teleconference on March 4, 2009 to discuss and amend a
preliminary draft Advisory. On August 6, 2009, the SAB provided a quality review in a
public teleconference.
In addition to offering specific responses to charge questions, this Advisory also
presents some general advice and cross-cutting recommendations. We first identify and
discuss these cross-cutting issues below and then proceed to the specific charge
questions.
Crosscutting Issues
In general, EEAC would like to see more upfront discussion of the "whys"
associated with the material in the Guidelines, perhaps in the form of a conceptual
overview chapter. It would be useful to explain why a benefit-cost analysis is a valuable
undertaking (other than to satisfy a regulation) and why economic impact and equity
analyses can be important supplements to benefit-cost analysis, etc. Readers need to see
a definition of economic efficiency (allocative and technical, which will help in
explaining the relationship between cost effectiveness and benefit-cost analysis) and an
explanation of the relationship between economic efficiency, benefit-cost analysis, and
cost-effectiveness. We also suggest defining social costs and benefits (as distinct from
private costs/benefits), with examples.
In addition to introducing these basic concepts, EEAC would like to see
discussion of a number of cross-cutting issues. Environmental policy analysis is
inherently an integrated assessment process in which results from different sciences are
combined to predict environmental outcomes and their economic consequences.
Realistically, a great deal of judgment will have to be exercised by analysts. Given this,
the Guidelines should discuss the need for transparency in making assumptions and
judgments and the ways in which biases and errors will matter the most. For example, if
all benefits and costs are understated by about the same amount, the "answer" of whether
the benefits exceed the costs will not likely change, however if the costs are biased up
and the benefits biased down, the wrong efficiency message could be sent. Care should
also be taken to avoid multiple counting of benefits and costs when there are overlapping
-------
regulatory initiatives. Clarity and transparency in the specification of the baseline,
including the regulatory initiatives already in place, will enable more accurate
identification of the incremental effects of a regulatory initiative.
The Guidelines should not underemphasize the value of economic analysis for
deregulatory and/or non regulatory purposes. The statement in the Guidelines that
"formal economic analysis is not required for the selection and implementation of a non-
regulatory approach," is true and probably an important point to make. However, the
statement could also suggest that economic analysis is less valuable or informative in this
case. It is not. Non-regulatory approaches can bring both sizeable costs and benefits and
the same can be said for deregulatory decisions. The Guidelines should indicate that
decisions to deregulate or adopt a non-regulatory approach can be as much informed by
economic analysis as those that are purely regulatory in nature.
We recommend that the Guidelines incorporate the concept of ecosystem services
and its various components, as outlined in the Millennium Ecosystem Assessment (MA)
Synthesis Report (2005) and highlight the issue of properly valuing and characterizing
ecological systems and services in a benefit-cost analysis. Although OMB Circular A-4
does not require that all economic benefits of a policy be monetized, it does require some
scientific characterization of those contributions. Users of the Guidelines should be
warned that an inappropriate focus only on impacts that can be monetized can provide
misleading policy guidance (as with other cases of unbalanced information). In addition,
a strong recommendation should be made to provide quantitative measures of ecological
impacts and a qualitative characterization of ecological effects. We urge EPA to consider
the SAB's recent recommendation to begin with a conceptual model of the relevant
ecosystem and ecosystem services and map those effects to services or attributes that the
public values (Valuing the Protection of Ecological Systems and Services, 2009). The
SAB report covers a wide range of alternative methods for characterizing, valuing and
gauging ecological impacts. We urge EPA to evaluate and determine the appropriate use
of these alternative methods and provide much more guidance on characterizing non-
monetized effects.
We understand that EPA is developing a separate chapter dedicated entirely to
uncertainty (which we applaud), but the topic is important enough to merit some
discussion in the conceptual overview or as a cross cutting issue throughout. It would be
useful to point out that uncertainty extends not only to economic information, but to
environmental data and modeling. Uncertainty in environmental modeling can be as
much or more a source of errors than imperfections in economic assumptions and data.
In addition, analysts will be confronted with heterogeneity of data for various reasons
(geographic, economic, cultural etc). Recognizing the sources of heterogeneity and
deciding how to address them are major analytic decisions. Given that analysts usually
face asymmetric information (e.g. on costs vis-a-vis benefits), advice is needed on how to
address these information deficiencies. The Guidelines might discuss the possibility of
ensemble modeling (e.g. hydrology and ecology) and the use of a "weight of evidence"
approach, especially for the case of non-monetized ecological effects.
-------
The literature on environmental policy increasingly emphasizes that realizable
outcomes will be "second best" due to information constraints, political constraints,
imperfect competition, and market distortions created by tax and other government
interventions. This emphasis is not adequately reflected in the Guidelines. For example,
the theory section focuses on first-best optimal regulation, a framework that provides a
baseline but is of limited relevance to regulation of contemporary environmental
problems. For example, agricultural nonpoint pollution is now the leading cause of the
nation's water quality problems, yet it is less observable than emissions from points
sources and far more stochastic. Consequently, the emissions based policies emphasized
in the Guidelines are irrelevant inasmuch as they are targeted to conventional point
sources. Agriculture is subject to multiple non-environmental policy distortions that must
be considered in the measurement of the social benefits and costs of regulating
agriculture. Further, agriculture is a source of multiple externalities, some that are
positive, some that are negative, that are regulated to varying degree (including not at all)
by multiple authorities. The Guidelines should be more adapted to the complexities of
contemporary environmental problems.
Given that economics needs to apply to economic analysis, we believe some
discussion should be devoted to the allocation of EPA resources in undertaking economic
analyses. Where possible, the Agency should consider tailoring the resources spent on
the analysis with the size of the proposed regulation's impact. Analysis of the costs of a
small project may emphasize simple partial equilibrium analysis, while a larger one may
employ both partial equilibrium and CGE models. Similarly, the selection of the number
and identity, of say, products or markets to be included in an analysis should consider the
benefits and costs from the adding each individual market or product. The evaluation of
a large project may justify conducting a new study on willingness to pay (WTP) for the
amenities it provides if such information has the potential to change the regulatory
decision while a study on the impacts of a smaller project may rely on a benefit transfer
analysis if the regulatory impacts are expected to be small. For large regulations with
significant impacts, the costs of analysis may be trivial in comparison to potential
increases in net benefits if the information results in a change to the final regulation or
policy decision.
The Guidelines should discuss the analytic challenges posed by emerging
environmental problems, particularly climate change. The Guidelines are implicitly
focused on conventional point source pollutants. Attention to emerging challenges from
nonpoint pollutants, changes in carbon and other biogeochemical processes, invasive
species, etc. would give the Guidelines a more contemporary and forward looking view.
The Guidelines should point out (and perhaps demonstrate) that the analysis needed for
policy decisions to address greenhouse gas emissions will necessarily draw from a
complex interdisciplinary suite of studies, data and models. A case in point is the
Congressional requirement in the Energy Independence and Security Act of 2007 for
EPA to issue a Renewable Fuel Standard based on its calculation of lifecycle greenhouse
gas emissions for various fuels. This lifecycle analysis covers the full fuel lifecycle from
production to consumption and hence requires an extraordinary synthesis of tools and
information from the bio-physical sciences, economics and atmospheric modeling. The
-------
Guidelines could reference this specific example as well as discuss generally the
interdisciplinary challenges posed by climate change and other global issues. Given the
specter of climate change and other global processes, we expect that future revisions of
the Guidelines will need to be broader, describing a kind of regulatory analysis that draws
from all the sciences. We note that revisions will be made easier with NCEE's adoption
of a loose leaf format to update chapters as appropriate.
-------
Question 1: Policy Options
Do the published economic theory and empirical literature support the statements in the
guidance document on the merits and limitations of the different regulatory and non-
regulatory approaches discussed in Chapter 4: Regulatory and Non-Regulatory
Approaches to Pollution Control?
With a few exceptions, the presentation in Chapter 4 is supported by economic
theory and empirical evidence in the published literature. We commend the Agency on a
good chapter and offer the suggestions below for improvement.
This chapter provides a good description of policy instruments, much in the same
way that a good environmental economics text might. In this regard, we suggest that
EPA clarify and discuss the specific role that EPA has in policy design and
implementation. For example, the Guidelines point out the efficiencies that can accrue
with performance-based standards, but do not tell the reader that EPA often does not have
the discretion to choose this option, e.g. when "best available technology" standards are
required by law. Similarly, after demonstrating that social welfare is maximized by
choosing the level of pollution that equates marginal costs with marginal benefits, the
Guidelines do not inform the reader that EPA is often explicitly prevented from setting
standards by this criterion.
We suggest that EPA clarify and discuss the Agency's specific role in policy
design and implementation. This would involve stating where EPA has discretion to
make decisions in the context of the larger policy arena and then discussing the rationale
for the different approaches actually used. That is, EPA primarily issues regulations and
those regulations are typically highly prescribed by Congressional mandates as well as
the courts' interpretation of environmental laws. On the other hand, EPA analysis can
inform the design of future environmental programs by Congress so it is important to
retain the discussion of the full suite of policy options described here, but clarifying
EPA's discretion in setting policy (even if it varies from statute to statute) would be
beneficial. To this end, we recommend that the chapter be organized in two parts: (1)
standard treatment of policy options similar to what is currently covered in the Guidelines
and (2) discussion of EPA's actual discretion highlighting how economic analysis is used
in this narrower context. This would take the chapter beyond the usual treatment of an
environmental economics text and make it directly relevant to the agency. One way to
explain EPA's role is to show some examples of specific actions (e.g. how design or
performance standards are set and/or EPA's role in voluntary programs).
Another topic that would be well suited for discussion in this chapter is the issue
of asymmetric information between the regulator (EPA) and the regulated (consumers
and industry). This issue has ramifications for the design and efficiency of many kinds of
environmental regulations and its importance should be mentioned.
Along these same lines, the chapter discusses maximizing welfare without noting
-------
that in most cases EPA is charged with implementing laws that specify criteria EPA must
use. EPA's regulations are not typically based on a calculation of an "optimal" E*, but
are more tied to legally-defined criteria. Again, the chapter would be improved by
clarifying how EPA's actual authority fits into this paradigm.
This latter point could be nicely integrated into Section 4-1 which could be
improved by distinguishing efficiency from cost effectiveness. By dividing the section
into two parts: (i) Efficient Level of Pollution (MD = MACaggregate )*and (ii) Cost-
Effective Allocation of Pollution - equalize marginal cost across sources (MACi=
MAC2= ... = MACn), the distinction between optimality of pollution levels and least cost
approaches for implementation can be distinguished. Each section will need a supporting
graph and should be integrated by showing how aggregate MAC is derived from
individual MACs. See Field and Field (2008) and Baumol and Gates (1988) for dividing
the discussion this way. Also, in the discussion of the efficient level of pollution, it is
important to define social welfare using the underpinnings of Pareto Optimality.
The suggestion in the previous paragraph will allow the use a cost-effectiveness
graph to tell the cap and trade story demonstrating its property as a least-cost instrument.
Likewise, instead of telling the cap and trade story assuming an optimal level of
aggregate pollution for permit allocations, tell the story from the standpoint that a cap and
trade can achieve any aggregate level of emissions at least cost. Then, point out that the
efficient solution is a special case where the permit allocation is efficient (MD=MAC).
Use the same approach for the tax in Section 4.3.2. Again, these points can be illustrated
with the point that EPA typically does not set the optimal level of emissions, but can help
design and implement instruments to achieve least cost solutions. This discussion can be
linked to the asymmetric information issues mentioned earlier by noting that the least
cost solution can be achieved, even if the regulator knows nothing about the individual
firm costs.
The discussion in Section 4-2 should be clear about the difference between design
standards (technology forcing) and technology based performance standards. This could
be accomplished by dividing Section 4-2 by (i) design standards and (ii) performance
standards and discussing uniform and technology-based performance standards using the
cost-effectiveness graph introduced in the previous section. Uniform standards are
generally not cost effective but have a low information burden (since one need not know
MACs). Technology-based performance standards can be cost effective in principle but
have a high information burden (need to know MACs). Regardless of the form of
standard, it is always important to recognize that different options may yield different
levels of environmental improvement which need to be adequately accounted for in
analysis that compares design with performance standards.
We recommend the section on taxes include some discussion of what is taxed: the
pollutant, an input, a process, or something elsewhere? In principle it should be placed
on damages (a true Pigouvian tax), but administrative and monitoring costs may suggest
targeting the tax elsewhere. Taxing gas may be much easier and probably as effective as
' MD = marginal damages. MAC = marginal abatement cost
-------
taxing damages from auto emissions. Similar principles apply to permit trading, so
although EPA lacks the legal authority to levy a tax, a discussion of targeting would have
broad application.
On this point more generally, note that policy design is dependent on observable
and available information. For example if policy makers can not observe actual pollution
levels, but inputs and practices that cause pollution can be observed and there are
reasonable estimates on the relationship between the input, technology use and the
pollution level, then taxes or regulations may be based on the imputed pollution levels.
Policies should be designed to best take advantage of the available knowledge. In
particular identifying sources of heterogeneity among users and knowing how they affect
pollution levels can be central to policy design. The proliferation of geographic
information systems and remote systems to obtain data provide new opportunities for
policy design. Studies suggest that there is a significant efficiency gains from policies
that adjust to observed heterogeneity relative to uniform policies (Xabadia et. al., 2008).
Availability of new sources of or means to obtain information may lead to redesign of
policies - for example, availability of a technology that allows cheap monitoring of
pollution may lead to regulation on taxation based on annual pollution rather than
imputed pollution. Policy makers should reassess policy design and implementation as
technology progresses. (See Xabadia et. al., 2008) as one example. There are many other
papers that demonstrate the efficiency gains from increased targeting.
As noted in the cross cutting issues discussion earlier, we suggest adding
discussion of second best solutions covering imperfect markets, pre-existing policies,
asymmetric information, and so forth. In this chapter, EPA could incorporate the
implications of imperfect markets and other second best solutions in the relevant sections
and refer readers to the material in the cross cutting issues discussion. Examples include
monopoly, price supports in agriculture, pre-existing environmental policy, recent
biofuels legislation, etc.
When the section on market based regulations is introduced, it would be helpful to
explain that these controls tend to be least costly, have a low information burden on
regulators, provide incentives for technological advancement, and so forth. Monitoring
and enforcement costs and other administrative costs, of course, can favor direct
regulation.
It would be helpful to mention that information disclosure strategies can allow the
market to create incentives for pollution control (following Coase) with the victims
directly signaling their preferences to firms. But these are most likely to work when there
are contractual obligations between polluting firms and affected parties (e.g.
consumers/workers) and more difficult to work when they affect third parties (see
Tietenberg, 1998). Also it should be mentioned that credibility of information is
important. Information disclosure can lead to inefficient outcomes when information is
not credible (see Brouhle, K. and M. Khanna, 2007).
The section on subsidies should mention that many subsidies in existence, such as
-------
those in agriculture, car manufacturing, oil and gas, forestry, and so forth, are not
corrective subsidies designed to correct externalities and may in fact worsen externalities.
This could also be discussed in the second best discussion.
An area of omission is the relatively new literature regarding the effectiveness (or
lack there of) of voluntary approaches ( Morgenstern and Pizer, 2007 and National
Research Council, 2002, as well as a number of journal articles). These should be noted
and made a part of the section on voluntary controls. Assessment of the effectiveness of
a voluntary program in the literature has been based on estimates of participation rates
and the reduction in pollution achieved by the program relative to that in the absence of
the program. A comparison of the costs of pollution control under a voluntary program
relative to that under alternative policy options to achieve the same level of pollution
control would be valuable for assessing the cost-effectiveness of voluntary approaches.
-------
Question 2: Baselines
Do the published economic theory and empirical literature support statements in the
guidance document on the consideration of the baseline discussed in Chapter 5:
Establishing a Baseline?
Yes, the approach described in the Guidelines to establish a baseline is supported
by economic theory and empirical literature in this area. This chapter provides very
comprehensive guiding principles for specifying the baseline scenario to identify the
incremental benefits and costs associated with a policy. It describes the methods for
defining a scenario that does not include the policy (baseline scenario) and one that does
include the policy; that is, a 'with' and 'without' policy comparison. It suggests that in
some cases it may be appropriate to specify multiple baseline scenarios to describe the
state of the world in the absence of a regulation. While we agree with the need to
consider more than one baseline when it is difficult to define a unique state of the world
in the absence of the policy with a high degree of certainty, the number of baselines
constructed should be limited to as few as possible that cover the reasonable baseline
alternatives. In some cases it may also be appropriate to use probabilistic analysis with a
continuum of baselines to provide the benchmark for policy analysis.
Additionally, in defining the baseline scenario, analysts need to consider which
sectors should be included. Although the direct effects of the policy may be focused on a
few sectors, indirect impacts can be significant and should be measured. It is therefore
important to establish which other sectors of the economy may be affected, directly or
indirectly, by a policy and should be included in the baseline. Some policies can have
pecuniary effects that will affect the opportunity costs of implementing that policy. This
is particularly relevant when the pecuniary costs occur in inefficient markets; in such
cases the opportunity costs of a policy can differ from the monetary costs of the policy
(see Boardman et al., 2006).
The assumption of full compliance with the existing and newly enacted
regulations does not appear realistic. Instead, compliance rates in the baseline should be
based on available factual evidence. Assumptions about compliance rates in the policy
scenario should also be based on a realistic assessment. These rates are likely to depend
on how stringently the policy is implemented and enforced.
We also suggest including a text box distinguishing the induced innovation effects
of regulation from the Porter Hypothesis. It would be useful to clarify what the Porter
hypothesis is (i.e., define it) and to distinguish between its strong form and weak form.
The strong form of the Porter Hypothesis states that regulations can induce innovations
that can lead to cost savings that are larger than the costs of the innovation and
compliance. The weak form of the Porter Hypothesis simply states that environmental
regulations lead to innovation. We agree with the statement that there is only limited
evidence of the strong version of the Porter Hypothesis; some references to the literature
providing situations in which it might hold (such as in the presence of imperfect
information, high search costs, etc.) could be added. There is much more evidence to
-------
support the weak form of the Porter Hypothesis (which is similar to the induced
innovation hypothesis) and this should be made clear.
10
-------
Question 3: Discounting
Do the published economic theory and empirical literature support the statements in the
guidance document on the treatment of discounting benefits and costs discussed in
Chapter 6: Discounting Future Benefits and Costs? (sub parts for this charge are copied
below)
Overall, the chapter provides a clear and comprehensive discussion of
discounting. This chapter has been updated to reflect much of the current thinking on
discounting in the context of environmental decision making and we applaud EPA for
doing so. We begin with some general comments then respond to each of the subparts of
the charge question below.
Discounting is an important, complicated, and controversial topic. The results of
an economic evaluation can be extremely sensitive to the discounting approach that is
used, especially for projects where significant benefits and costs are incurred at widely
disparate times (e.g., climate change mitigation; nuclear-and hazardous-waste storage).
We suggest that the chapter begin with an acknowledgement of these issues and orient
readers to the parts of the chapter in which they are discussed. (At present, readers must
wade through the necessary but less interesting section on mechanics of discounting
before grappling with these topics). Some of the issues that could be highlighted include:
(a) differences between relatively short run ("intra-generational") and long run ("inter-
generational") discounting that arise in part because inter-generational contexts
necessarily involve a greater distributional aspect and future generations are not
represented in markets; (b) sensitivity of results to choice of discounting approach and
discount rate; (c) distinction (and frequent confounding) of efficiency and distributional
issues; (d) distinction between utility vs. consumption discount rates; (e) "ethical" or
prescriptive vs. descriptive approaches to selecting a discounting approach; (f)
uncertainty about future economic growth and other conditions; (g) constant vs. non-
constant (e.g., hyperbolic) discounting approaches; and (perhaps) (g) future changes in
relative prices that imply different consumption discount rates for different goods.
Although it is implicit in the text, the distinction between discounting to reflect
differences in timing of consequences and discounting to adjust for inflation should be
emphasized. It is conventional (and recommended) to measure effects in real dollars and
use a real rather than nominal discount rate to account for differences in timing. (Note
that if inflation rates differ across goods, the real discount rate depends on the inflation
adjuster that is used.)
There is much confusion in the literature (and in policy discussion) about the
determinants of the discount rate, e.g., whether it is a result of preferences for
consumption sooner rather than later or of the productivity of capital investment. At least
for "intra-generational" discounting, the (consumption) discount rate (or rates) is best
understood as being determined by a price (or prices) - the interest rate(s) at which
consumption can be shifted through time (e.g., consumers may shift consumption to the
future by consuming less and investing more, and may shift consumption toward the
11
-------
present by saving less or borrowing). The interest rate is determined, like other prices, as
a market equilibrium between agents who wish to shift resources through time.
Consumers adjust current and future consumption so that their indifference curves for
current and future consumption are tangent to the market opportunities for shifting
consumption through time (determined by the interest rate). As noted in the chapter, a
variety of interest rates exist (associated with riskiness of investment and time horizon)
and there are wedges between private returns associated with taxation of investment
returns and other factors. For longer-term "inter-generational" issues, the interpretation of
the discount rate as a price is less natural, because market interest rates for periods longer
than about 30 years rarely exist and future generations' do not directly participate in
current markets, so their preferences may not be adequately represented.
In the discounting chapter or elsewhere, it may be useful to highlight the
distinction between positive (i.e., descriptive) and normative perspectives (i.e.,
prescriptive) justifications for economic evaluation and the tension between these.
Economic evaluation is normative in that it is conducted in order to compare alternative
policies, yet it is positive in that it attempts to identify the policy that maximizes the
perceived welfare of the affected population. Behavior that differs from Economic
evaluation is sometimes justified as identifying policies to maximize (or at least increase)
social welfare, which is normative because it is based on an assumed social welfare
function. In contrast, economic evaluation is also sometimes justified as identifying
policies that produce potential Pareto improvements, i.e., policies such that all members
of the affected population would prefer to the alternative policy (if combined with
appropriate compensation payments). Policies that increase social welfare are not
necessarily potential Pareto improvements, that those that are potential Pareto
improvements need not increase social welfare. Moreover, individual behavior that is
inconsistent with maximization of individual utility (e.g., behavior that is dynamically
inconsistent, perhaps because it accords with hyperbolic discounting) creates a tension
between these perspectives and raises questions about how to conduct the analysis. See
Hammitt (2009, 2002) and Portney (1992).
a. Are the descriptions of fundamental social discounting approaches, conceptual
conclusions and recommendations consistent with the appropriate economic literature on
social discounting? Are the correct conclusions drawn from the respective literatures on
discounting for public projects (government spending) and discounting for regulations
(government-mandated spending) ?
The descriptions of fundamental social discounting approaches, conceptual
conclusions and recommendations are consistent with the appropriate economics
literature. We do not believe there are significant differences in the approach to
discounting for public projects (government spending) and discounting for regulations
(government-mandated spending) - in both cases, there can be a need to account for
differences in timing of benefits and costs and the relevant conceptual basis is social
valuation of consequences at different points in time.
12
-------
b. The Guidelines do not draw a firm conclusion on the extent to which shadow
price of capital adjustments are likely to be necessary for most EPA policy analyses. The
issue depends greatly on the elasticity of capital supply and EPA plans to pursue
additional research on this issue, as noted in the draft Guidelines. Does EPA 's
conclusion reflect the sense of the literature or can a firmer conclusion be drawn? Does
the Committee have suggestions regarding situations where these adjustments would be
necessary or unnecessary?
As noted in the chapter, the shadow-price of capital approach is viewed as
theoretically correct, but the quantitative significance of adjusting for this shadow price is
critically dependent on the extent to which EPA regulations displace other investment. In
an economy that is open to foreign investment, there may be minimal displacement and
so adjustment for the shadow price of investment is negligible. We are pleased to learn
that EPA is investigating the elasticity of investment to environmental regulation.
c. While EPA concludes that a rate of 3% is generally consistent with estimates from
low-risk government securities, the Agency would like to more firmly establish a rigorous
basis for a consumption-based rate. What data and methods would the committee
suggest EPA pursue ?
We are unable to suggest better data or methods for estimating a consumption-
based discount rate, and doubt that alternative credible estimates would differ
dramatically from the 3% rate specified by OMB in Circular A-4. Given the benefits of
harmonization of parameter values among federal agencies, we do not encourage EPA to
move away from this rate.
d. Chapter 6 recommends adopting an approach to long term discounting based on
the work of Newell & Pizer (2003). While EPA recognizes that data may not clearly
support a particular statistical model over other alternatives (e.g., random walk vs.
mean-reverting), the Chapter concludes that the recommended approach is an
improvement over constant discounting. Does the committee believe this is a reasonable
conclusion from the economics literature? More specifically, is the recommendation to
use a random walk model as a default reasonable given the state of the literature?
We do not recommend using the Newell and Pizer approach as a default but as
one of the alternatives for inter-generational discounting. Since the declining discount
rates under this approach are sensitive to modeling assumptions, transparency in the
assumptions underlying the determination of the discount rates used will be important as
will comparisons with other alternatives.
The conceptual idea, identified by Weitzman (1998, 2001), should be clearly
stated in the chapter: uncertainty about the appropriate discount rate to use has a
nonlinear effect on the discount factor that varies with the time horizon. Specifically,
because the discount factor for time t, [1/(1 + r)]1, is a nonlinear function of the discount
rate r, the expected discount factor is not equal to the discount factor obtained by
substituting the expected value of r in this formula. For small t, the difference between
13
-------
the two discount factors may be small, but as t becomes arbitrarily large the expected
discount factor approaches the discount factor corresponding to the minimum possible
value of r weighted by the probability assigned to that minimum value. Weitzman
developed a distribution for r by polling economists and derived a corresponding
schedule of certainty-equivalent discount rates (the discount rate associated with the
expected discount factor); Newell and Pizer built on his work by conducting an empirical
analysis of historical interest rates. Their results are naturally sensitive to modeling
choices about the intertemporal correlation of rates.
Our concern about this approach is the following. In a world with no uncertainty
about the discount rate, one can compare streams of consequences in terms of their
present values or their future values at any future date. The choice between these
perspectives has no effect on the ranking: because the present value of a policy is simply
the present value of the policy's future value (i.e., the future value discounted to the
present), whichever policy has the larger present value will also have the larger future
value.
In contrast, when the discount rate is uncertain, the rank ordering of policies by
present values and future value may differ. As shown by Weitzman, for large t the
present value is dominated by the small discount rates. But the future value is dominated
by the large discount rates (i.e., the expected value of the factor used to convert present
consequences to their future value, (1 + r)1, is dominated by the largest possible values of
r). This dependence of the evaluation on what appears to be an arbitrary choice of
perspective suggests a problem with the analysis that urges caution in its application and
invites further investigation (Gollier, 2004; Hepburn and Groom, 2007). Given this
concern, we urge caution in interpretation of results calculated using the Newell and Pizer
approach and encourage further investigation.
e. EPA has struggled with the question of the length of time an analysis should
capture and has arrived at some practical recommendations (see Section 6.1.6.3 and
6.4). Are these recommendations consistent with good economic practices? Does the
committee have additional recommendations or insights on this subject?
In considering the time horizon an analysis should cover, there is no general
answer beyond the answer to the question of what consequences to include: those that
may have a quantitatively significant effect on the conclusions of analysis (as noted in
Section 6.1.6.3). With positive discounting, the influence of consequences decreases with
their temporal distance; unless the probability-weighted magnitudes of consequences
grow sufficiently rapidly with time, their effect on the analysis will become negligible. In
general, there is no method for knowing whether a consequence may be sufficiently
important to merit inclusion except by including it and testing for its effect. For this
purpose, a rough estimate or upper bound is often sufficient. As noted in the text, many
exogenous factors are likely to influence the date at which the effects of the policy
become negligible (e.g., technological innovation or policy change).
14
-------
Question 4: Benefits
Do the published economic theory and empirical literature support the statements in
Chapter 7: Analyzing Benefits on the merits and limitations of different valuation
approaches for the measurement of social benefits from reductions in human health risks
and improvements in ecological conditions attributable to environmental policies?
This chapter provides good coverage of the main categories of benefits of
environmental policies and regulations, and of the methods used to estimate them. The
chapter might flow better if the discussion of the main categories of benefits was more
concise, and more details about the valuation of these impacts were offered in the
overview at the end of the chapter.
Chapter 7 falls significantly short of capturing a considerable amount of recent
literature on benefits. In the following pages, we identify a number of areas where the
literature of the past decade or so is not adequately reflected in the chapter. We have not
made an effort to be exhaustive in recommending additions, but rather to identify areas
as obvious omissions. In general, the Handbook of Environmental Economics published
by North Holland in 2005 would be an excellent starting place. More specifically, the
literature in the following areas needs to be updated:
1. Recreation Demand Models. A great deal of work published in the last decade on
random utility maximization (RUM) and Kuhn-Tucker models have taken these
approaches beyond the descriptions provided in the Guidelines. Updated approaches
to valuing the opportunity cost of time, identification of choice sets, and other aspects
of recreation demand should also be included. Potentially useful sources for journal
research that should be reflected in the Guidelines can be found in the collection of
articles in Herriges and Kling, (2008), among other works. In addition to covering
methods for valuing natural resources that have recreational use, this volume is also a
good source for articles related to hedonics and locational equilibrium models. For
more theoretical treatment cite Bockstael and McConnell, (2008). For more practical
discussion cite Champ et. al. (2003).
2. Combining Revealed Preference (RP) and Stated Preference (SP). A revised draft of
this chapter should discuss work over the past decade that has sought to combine
revealed and stated preference methods. This has been an area of significant interest
as researchers have attempted to understand how the strengths of each approach
might be combined to improve the performance of welfare estimators. On the
theoretical front, Herriges and Kling, (1999) raise the question of whether revealed
preferences can ever accurately estimate welfare for quality changes when weak
complementarity cannot be assured.
3. Stated Preferences: Validity and Reliability. A significant amount of recent work
related to stated preference approaches is not reflected in the Guidelines.
Understanding whether people over- or under-state their actual preferences for a non-
marketed good when asked a hypothetical question and whether approaches to
15
-------
mitigate these effects are successful remain important researchable questions.
Considerable work on this question has been undertaken beginning with studies such
as Bohm's (1972) experimental lab study which compared bids in hypothetical and
real experimental markets that elicited subjects' stated value to sneak preview a
Swedish television show. His results suggest that people moderately overstate their
real values when asked a hypothetical question. Other early work includes the studies
by Bishop and Heberlein, (1979), Duffield and Patterson, (1992), and others who
compared stated preference estimates to those obtained from actual transactions.
Subsequent research has included both field and laboratory experiments. For
instance, List and Gallet (2001) report the results of a meta-analysis to determine
whether important experimental parameters systematically affect the relationship
between hypothetical and real responses, concluding that certain elicitation methods
that yield less hypothetical bias than others. Others (e.g., Cummings and Taylor,
(1999), List, (2001), Lusk and Prevant, (2008), etc.) have studied hypothetical bias
and incentive compatibility focusing specifically on the dichotomous choice
elicitation format in contingent valuation. These and other references addressing this
literature should be discussed in the Guidelines. Remaining research questions and
needs should also be identified.
A final area in which SP validity and reliability research has appeared is in the arena
of choice experiments. Although many choice experiment-based studies are not
conducted in ways that would preserve incentive compatibility, some limited
evidence exists suggesting little or no hypothetical bias when estimating marginal
attribute values (see List et al., 2006, and Lusk and Norwood, 2005)
4. Valuing Mortality. There are a number of studies that need to be updated in this
area, please see charge question #5 for specifics.
5. Valuing Morbidity. The Guidelines should provide a sense as to whether the
research community is satisfied with existing estimates of morbidity benefits. Is
the usual approach—symptom days—still judged acceptable? Likewise, this
section points out that frequently used approaches, such as the cost-of-illness or
averting expenditures, do not capture the full WTP to avoid an episode of illness.
It would be useful to note that some studies (e.g., Rowe and Chestnut, 1985, and
Alberini and Krupnick, 2000) have estimated that total WTP can be two to four
times as large as the cost of illness, even for minor acute respiratory illnesses (as
in Alberini and Krupnick's case).
Ecosystem Services and Benefits Assessment. The literature review is outdated,
and there are no examples of recent studies that attempted to place a value on
ecosystems and ecosystem services. It would be useful to know which measures
of ecological system function was used in those studies, whether market or non-
market valuation approaches were used, and what the strengths or shortcomings
of these studies were.
16
-------
Ecosystem services encompass a broad array of goods and services,
ranging from standard market goods like agricultural products to far more
complex and less well identified services such as nutrient cycling. Valuing the
more complex ecosystem services is very challenging because it requires
sequential linkages from the policy of interest to the direct effects on organisms to
the associated indirect effects through changes in the functioning of ecosystems to
the resulting changes in services and finally to the associated social values. Our
understanding of each of these elements is rudimentary in some cases. Yet, we
need to develop analyses to support policy decisions in spite of the many
uncertainties.
This document should provide more extensive guidance on how to carry
out valuation of ecosystem services within the context of policy decisions that
EPA needs to make. A variety of techniques are available that can provide useful
input to policy within the context of less than complete knowledge. In addition to
standard cost-benefit methods, these approaches include cost effectiveness
analyses, choice-based methods (e.g., Opaluch et al, 1993; Unsworth and Bishop,
1994; Adamowicz et al, 1998), etc. The report should also draw from recent
reviews, including EPA SAB (2009) and the National Research Council (2004).
The report should address important challenges facing the practitioner,
including: What does valuation of effects on ecological systems share with
valuation of other benefits? In what way is it different? What are unique
difficulties when valuing ecosystems or ecosystem services? What approaches
are available in face of the many uncertainties?
As mentioned in the discussion of cross-cutting issues, we urge the
Agency to vastly expand its guidance on characterizing non-monetized benefits.
We recommend that the Guidelines incorporate the concept of ecosystem services
and its various components, as outlined in the Millennium Ecosystem Assessment
(MA) Synthesis Report (2005) and highlight treatment of ecological systems and
services in benefit-cost analysis. Users of the Guidelines should be warned that an
inappropriate focus only on impacts that can be monetized can provide misleading
policy guidance (as with other cases of unbalanced information). In addition, a
strong recommendation should be made to provide quantitative measures of
ecological impacts and a qualitative characterization of ecological effects. These
quantitative measures and qualitative descriptions are needed whether or not
benefits can be monetized. We note that EPA's Office of Research and
Development has an extensive Ecosystem Services Research Program that may be
an excellent resource for economists who need information ecosystem impacts for
economic analysis.
We urge EPA to consider the SAB's recent recommendation to begin with
a conceptual model of the relevant ecosystem and ecosystem services and map
those effects to services or attributes that the public values (Valuing the
Protection of Ecological Systems and Services, SAB, 2009). The SAB report
17
-------
covers a wide range of alternative methods for characterizing, valuing and
gauging ecological impacts. We urge EPA to evaluate and determine the
appropriate use of these alternative methods and provide much more guidance
characterizing non-monetized effects. Another recommendation from the SAB
report that could be appropriated for the Guidelines is to label aggregate
monetized benefits as "total monetized economic benefits," not "total benefits."
We believe the SAB report provides other useful examples relevant to the
Guidelines.
In addition to updating the literature in the area just enumerated, we recommend
expended treatment of the discussion about assessing studies and data. We applaud
EPA's discussion of validity concepts (page 7-41) as the basis for choosing among
studies for inclusion in a benefit-cost analysis, but the treatment of the validity and
reliability of estimation methods is uneven. Much of the discussion about validity centers
on stated preference methods. However, we feel that due to (usually untested)
assumptions that they make about individuals' perceptions of environmental quality and
identification of effects, revealed preference methods should be scrutinized for quality
and validity, as should CGE models.
The material in Sections 7.4.2.3 "Considerations in Evaluating Stated Preference
Results" and 7.4.3 "Benefits Transfer," could be used as a starting point. For example,
the validity tests (content, criterion and convergent) discussed on 7-41 - 7-42 apply to all
types of studies, not just stated preference, as do various biases associated with survey
non-response (7-42 - 7-43). Other validity concepts that should be discussed include:
• Internal Validity: is there plausibly exogenous variation in the variable of
interest (the one capturing health risks or environmental quality)?
• External Validity: Can the study's results be generalized to the overall
population of interest? Can the study's results be generalized to the time
period of interest? Is the study's treatment relevant for the program that is
under consideration?
• Theoretical Validity: Can the study's results be interpreted as a measure of
willingness-to-pay (WTP) (or a bound on WTP)?
A discussion of the trade-offs between revealed and stated preference approaches
could be very helpful to readers. EPA should note that flaws of both stated preference
(SP) and revealed preference (RP) should be considered when evaluating studies and
performing regulatory analysis. For example, most hedonic properties, compensating
wage differentials and other revealed preference studies assume, without testing, that
people know the correct risks or level of environmental quality. These studies will often
estimate the willingness to pay or accept for changes that do not match well the policy
change under consideration. In contrast, validity questions related to stated preference
studies are a limitation that should be not disguised or diminished. It would be useful to
emphasize that judgment is essential to good analysis and that the results from stated (and
revealed) preference studies should be carefully assessed with economics and common-
sense basics. For example, is WTP for a change in environmental quality a reasonable
18
-------
fraction of one's income? Does it increase in predictable ways as income increases, and
are the estimates of income elasticity of WTP reasonable and consistent with other
evidence?
Chapter 7 states that the Agency should use benefit transfer only as a last resort,
when time and budget constraints do not allow original benefit estimation. We agree, but
the reality appears to be that benefits transfer is the most common approach to
completing a benefit-cost analysis. If correct, this should be acknowledged. Further, the
exposition of the possible benefit transfer techniques reads somewhat mechanically and
does not clearly discuss the underlying assumptions. For example, unit value transfer
presumes that the original good, the characteristics and the tastes of its population of
beneficiaries are the same as the policy good/locale. When a value function transfer is
done, it is implicitly assumed that the population of beneficiaries to which we are
applying the transfer has potentially different characteristics, but similar tastes, as the
original one.
EPA should distinguish the criterion used to evaluate an original study from those
that can be used to evaluate a study for use in benefits transfer. Given the importance of
the process of "benefits transfer" in benefit-cost analysis performed for EPA, separate
guidelines for analysts on to how to evaluate studies to use in a benefits transfer is
warranted. In doing so, we note that there is an extensive literature that provides
guidance on benefits transfer that the Agency can draw upon such as the special issues in
Ecological Economics. Some issues that are likely to belong in such a set of
considerations include: similarity of environmental good valued in original study to
environmental good being valued in benefits transfer; similarity of original study sample
to population of interest in the policy/regulatory setting; and overall quality of the
original study benefits (i.e., the set of considerations in the above list "benefits - original
studies").
The document discusses meta-analysis as one way of conducting benefit transfer.
Almost two years ago, two meta-analysis experts briefed the EEAC about meta-analysis.
They reminded us that a meta-analysis seeks, at best, to establish whether certain aspects
of study design and execution influenced final values, and its results should be interpreted
with caution. They offered a number of recommendations, warning against the
"ecological fallacy" and against pooling values from studies conducted with extremely
different methods. The chapter would benefit from reviewing the main lessons from that
presentation.
19
-------
Question 5: Mortality Risk Valuation
Chapter 7 includes a brief discussion of the Agency's current approach to mortality risk
valuation with more details provided in Appendix B. These sections will be updated
when the Agency concludes its efforts to update its mortality risk valuation approach. In
the interim, are the discussions provided in Chapter 7 and Appendix B clear and
balanced?
We will refrain from extensive comments on mortality risk valuation until the
Agency's update is complete. In the meantime, we recommend the Agency consider
expanding discussion in the following manner to improve the balance in the current
version of this section.
The literature review about methods used for valuing mortality risks should be
updated. Generally, the Chapter does a good job recognizing the advantages and
limitations of using the two main metrics in mortality benefits valuation (the VSL and the
VSLY) and of the different methods used for estimating them (revealed preference,
usually in the context of occupational risk, and stated preference). It also does a good job
discussing factors—such as age and pre-existing conditions—that matter with
environmental exposures and may affect the VSL. The chapter does, however, overlook
Viscusi and Aldy (2007), who look at age and the VSL in a wage-risk context. It could
also include discussion of alternative methods for estimating VSL such as the "chained"
approach linking WTP to reduce non-fatal injury with risk tradeoffs between fatal and
non-fatal injury (Carthy et al., 1999).
Despite its careful discussion of the limitations implicit in using the VSL
estimated from compensating wage studies when valuing mortality risks associated with
environmental exposures, Appendix B and much of EPA's current practice continue to
rely on old wage-risk studies. All of these wage-risk studies use old data, i.e., risk levels
that are no longer likely to exist and obsolete preferences for risk and income; they are
based on cross sections of data, do not control for self-selection into risky jobs and for
heterogeneity in preferences for risk and income, and contain a massive measurement
error in the risk variable. Better studies to discuss that control for unobserved
heterogeneity as well as use better risk measures are Kneisner, et al. (2006), Viscusi
(2004), Kochi (2008), among others.
The Guidelines should discuss the research concerning the VSL for specific
causes of death that are associated with environmental exposures—cancer and cardio-
and cerebro-vascular illnesses (e.g., Gayer et al., 2000, 2002; Davis, 2004; Johannesson
and Jonsson, 1991; Alberini and Chiabai, 2007). The Guidelines should also discuss
research needs in this area. A number of studies using revealed preference methods infer
values for risks related to the agency's policies (e.g., Davis, 2004, Gayer et al. 2002,
Greenstone and Gallagher (2008), Ashenfelter and Greenstone (2004), Gayer et al.
(2000)) and others. Likewise, there are newer stated preference studies that infer values
20
-------
for risks related to agency's policies that should be referenced (see Krupnick (2007) for
examples).
In several places, the Agency refers to the importance of "the impacts of risk and
population characteristics" on valuation estimates (e.g., p. 7-6, line 36; p. 7-8, line 31-43;
p. B-4, lines 29-32), yet the discussion is generally focused on population characteristics.
The agency should consider adding more discussion about the impact of risk
characteristics on valuation, the newer literature valuing the events of relevance to
environmental policy, and relate these issues to its recommended default value.
21
-------
Question 6: Social Costs
Does Chapter 8: Analyzing Costs contain an objective and reasonable presentation of the
published economic theory, empirical literature, and analytic tools associated with
estimating social costs?
In general, the Chapter does contain an objective and reasonable presentation of
economic theory, literature and analytical tools. While generally quite well done, we
found several areas where improvements could be made.
At the end of Section 8.3.1.3 Discounting, the Guidelines state "In calculating
firms' private costs, e.g. the internal cost of capital used for pollution abatement, analysts
should use a discount rate that reflects the industry's cost of capital." While this quote is
correct, we fear that it could be misinterpreted to suggest that costs and benefits might be
legitimately discounted at different rates within a single analysis. Doing so could lead
practitioners to make significant errors. For example, an analysis with different discount
rates applied to public and private costs could justify a project whose private costs greatly
exceed its public cost savings when both occur at the same time period in the distant
future. Yet, as the time approaches, this decision would be reversed, implying a form of
time inconsistency. The Guidance should be written to make clear that such a practice
would not be appropriate.
The beginning of the chapter indicates that costs are usually viewed as
straightforward to estimate, but in fact estimating costs presents many challenges. For
example, estimating costs of new regulations requires forecasts many years into the
future. As pollution control technologies are implemented over time, firms can learn
from experience and the development of new technologies may reduce the cost of
achieving the standards. At the same time, ex ante cost estimates may be based on the
assumption that everything works as anticipated, but in practice deviations from expected
outcomes often mean higher than anticipated costs. These challenges can be exacerbated
since industry often has more information about costs than regulators and are likely to
have the incentive to overstate their costs. All these arguments reinforce the report's
indication that measuring costs are not at all straightforward, nor are cost estimates
"hard" numbers.
The report differentiates between partial equilibrium analyses which model a
single market or a small number of markets, versus general equilibrium analyses that
model the entire economy. The Chapter might be better organized by examining three
cases: single market analyses, multiple market analyses and general equilibrium
analyses. Significantly different challenges are faced when carrying out multi-market
partial equilibrium models than single market models, and it might make the explanation
more clear.
Also, some additional guidance would be helpful on carrying out multimarket
partial equilibrium models (e.g., Section 8.1.2, page 8-5). For example, the report could
indicate the conditions under which multimarket models are likely to be necessary, and
22
-------
how multimarket analyses should be carried out. The Just, Hueth and Schmitz textbook
provides a thorough treatment of these issues, and could be both the source of
information for the document summary and an excellent reference for further
information.
It is important to note that it is not simply the number of markets that are affected
directly by a regulation, but also their size and influence on the economy that determine
whether partial equilibrium analysis is adequate. Indeed, the first example under CGE
(Sec 8.1.2) is of a single but large market (electric utilities). Similarly, a partial
equilibrium analysis of a significant change in "the labor market" is unlikely to be
adequate, given its influence on nearly all markets in the economy. It might also be worth
noting that definition of "a market" is not always clear cut. Is "the labor market" one or a
large collection of segregated markets with somewhat permeable boundaries, by age,
education, experience, geography, etc.?
On page 8-4, immediately below the figure, the text states "While in reality at
least part of the compliance cost will likely be spent on abatement-related purchases from
other industries - and is thus not necessarily a loss to society - in this market, the
deadweight loss resulting from the regulation is lost completely." It should be made clear
that expenditures in other markets are losses to society except for the portion that is a
quasi-rent. We are concerned this passage could be interpreted as expenditure per se in
other markets are not losses to society. Again, the Just, Hueth, and Schmitz (2005)
textbook is an excellent source for appropriate accounting of multimarket effects.
In general, the treatment in the report focuses too much on perfectly competitive
markets, and provides too little discussion of non-competitive market environments. In
our introductory comments and in response to Charge Question 1, we discuss the need to
incorporate real world "second best" conditions in the Guidelines. With respect to costs,
results will differ significantly in a non-competitive market, or in a market where there
are other distortions. In general, a complex game theoretic formulation is needed to
assess effects in markets that are not perfectly competitive.
23
-------
Question?: CGE Models
Does Chapter 8 contain an objective, balanced and reasonable presentation of the
published economic theory, empirical literature, and analytic tools associated with
computable general equilibrium (CGE) models? Is the description of the relevance of
these models for economic analyses performed by the EPA reasonable?
Overall, the discussion of the structure and use of computable general equilibrium
models in Chapter 8 is concise and very good. It clearly summarizes the design and
structure of CGE models, their strengths and weaknesses, and the role of such models in
economic analysis at EPA.
The only major topic the section does not discuss is the parameterization of CGE
models. Some use behavioral parameters estimated econometrically from extensive time-
series data, while other models use parameters calibrated to a single input-output table or
taken from the literature. Estimation is clearly preferable where adequate data exist. One
of the principal benefits of general equilibrium modeling over input-output analysis is its
ability to capture substitution in production and consumption, so it is important that the
relevant elasticities be tied as closely as possible to the historical record. A paragraph on
parameterization should be added to the section. A brief discussion of parameterization
should be added to Section 8.4.4 on Input-Output analysis as well.
Secondly, the Guidelines' discussions of CGE modeling in general and the
concept of general equilibrium welfare analysis in particular should be expanded to
address the role of models that introduce pollution (or equivalently environmental
services) in non-separable specifications for consumer preferences. Such specifications
introduce the prospect for feedback effects where policies to reduce externalities lead to
changes in the demand for market goods and then in turn the amount of pollution giving
rise to the externalities. These responses "feedback" and affect the demands for market
goods. The process can be expected to continue with the models describing systems
where the market and non-market interactions affect the ultimate market equilibrium. It is
important to draw distinctions between sorting models with multiple markets and what
might be described as extended partial equilibrium analyses and CGE models. Recent
advances in both types of models allow EPA to consider using these structures to assess
when the changes associated with their policies would be large enough that conventional
practices that assume away general equilibrium effects need to be modified or at least
qualified. There is sufficient research that the Guidelines can begin to introduce candidate
procedures for addressing these issues.
The section would benefit from a few minor revisions as well. First, the
discussion at the beginning of Section 8.1.2 should emphasize that the need for general
equilibrium analysis depends on the scope of the policy's effects rather than just the
number of markets. A policy might have significant effects in a single market, but if the
market is large enough (the labor market, for example), general equilibrium analysis
would still be warranted. Similarly, a policy affecting a large number of very small
markets might be adequately addressed by partial equilibrium analysis.
24
-------
A second minor revision is that the discussion of the benefits of dynamic models
should be expanded slightly. Not only are such models useful for capturing saving and
investment effects, they can also be used to examine policies that themselves change over
time—becoming more stringent, for example. In addition, models based on intertemporal
optimization by the underlying agents can also capture anticipation effects: changes in
behavior occurring when policies are announced that don't take effect until some point in
the future.
A third minor point is that the discussion of the ability of CGE models to capture
transition costs should be expanded slightly and the conclusion that they cannot capture
such costs should be refined. It is true that many models are inappropriate for short run
analysis because they assume that capital and labor are completely mobile between
sectors, and because they typically use substitution elasticities that reflect medium to long
run behavior. However, those are characteristics of existing models rather than the
methodology itself. Models that use sector-specific capital stocks and costs of
adjustment in investment are able to capture important short-run costs due to
misallocation of capital. In principle, models with adjustment costs in labor demand
could capture additional short-run costs due to labor misallocation as well. It would be
most accurate to say that many existing CGE models are not designed for analyzing
short-run transitional costs, rather than it being a problem inherent in the methodology.
Finally, because CGE models differ considerably from one another in design and
parameterization, it is valuable to use multiple models when possible, especially for
policies expected to have very large effects on the economy. When EPA uses CGE
analysis, it often does use multiple models and it would be very useful to note that in the
text.
Finally, the issues of validity and reliability mentioned in other areas of the
Advisory are equally relevant to CGE models. We suggest EPA make this point in this
chapter and refer readers to the broader discussion of these issues elsewhere.
25
-------
Question 8: Distributional Analyses
Does Chapter 9: Distributional Analyses: Economic Impact Analyses and Equity
Assessment contain an objective and reasonable presentation of the measurement of
economic impacts, including approaches suitable to estimate impacts of environmental
regulations on the private sector, public sector and households? This discussion
includes, for example, the measurement of changes in market prices, profits, facility
closure and bankruptcy rates, employment, market structure, innovation and economic
growth, regional economies, and foreign trade.
Chapter 9 contains an objective presentation of many of the aspects of economic
impact analyses (EIA). It tackles market prices, profits, facility closures, unemployment,
market structure, innovation, and growth. Although Chapter 9 contains information on
estimating the economic impact of policies, discussing three approaches briefly: Direct
Compliance Costs, Partial Equilibrium and Computable General Equilibrium (CGE), it
points to Chapter 8 for more details. To make Chapter 9 stand alone, developing brief
examples from chapter 8 for both partial and CGE may further the reader's understanding
of why these are more effective than the direct compliance costs. Also, augmenting the
warning of the complexity of CGE modeling may be worthwhile.
While there is attention to the implementation of CGE models, the discussion of
partial equilibrium models basically directs the reader to find supply and demand curves
or elasticities, but there are multiple ways of implementing partial equilibrium models.
One which would tie this chapter to the benefits chapter is the "production function"
approach. More discussion concerning implementation would be valuable.
The Guidelines acknowledge that input-output (I-O) models have important
conceptual shortcomings as measures of economic benefits and costs. In addition, it
might be pointed out that I-O models ignore opportunity cost of resources, implicitly
assuming that inputs used in some new activity would otherwise go be idle and have no
opportunity cost. For example, I-O analyses frequently use multipliers to calculate jobs
"created" due to the direct, indirect and induced effects of an activity. A proper measure
of benefits would account for the opportunity costs by subtracting the value of labor in its
next best use.
Some other miscellaneous points to consider:
• Linear programming (LP) is more of an optimization method than an
economic concept or model.
• There should be some discussion of the marginal cost of public funds as a cost
of policies. This may belong in Chapter 8 rather than Chapter 9.
There should be a discussion of the implications of distortions other than taxes for costs,
such as imperfect competition and rent seeking public interventions.
26
-------
Question 9: Measuring Equity Effects
Does this chapter contain a reasonable presentation and set of recommendations on the
selection of economic variables and data sources used to measure the equity dimensions
identified as potentially relevant to environmental policy analysis?
Data sources for both items are reasonably well addressed in Chapter 9. For
information on the former, the U.S Census (household and economic) provides a majority
of the data while industry rating agencies provide a deeper understanding on which
industries are susceptible as described in the document (9.5.2 - 9.5.4). However, there is
some uncertainty in how to obtain information on government entities as only accessing
data through "community or state finance agencies" is mentioned (9.5.2.2). Information
on assessing how the populations of interest are being affected can be found via various
environmental sources pointed out in Section 9.8.4. Overall, the chapter contains all the
relevant economic variables necessary in an equity analysis. To support analysts, EPA
might consider creation of a website to catalog all of the data sources.
The chapter also provides a reasonable discussion of what the analyst should
consider in measuring the distributional aspects of regulations. The main items in
assessing equity issues are correctly identifying the populations of concern and
accounting for how the populations of interest are being affected. In doing so, it is
important to balance data acquisition costs against the value of accuracy.
One limitation we note is that the main distributional issues that are discussed in
this chapter relate to the cost side: i) direct compliance expenditures (p. 9-17), ii) indirect
costs (taking into account multipliers, GE effects, etc.), and iii) enforcement costs. The
bulk of the chapter relates to direct compliance expenditures, and discusses how
regulation influences prices, through changes in the composition of industry for example.
This focus on the cost side misses many important issues related to environmental justice.
EPA should note that there may be just as much interest in considering the distributional
effects of benefits of an environmental change. The costs of identifying the distribution of
benefits may be much more data intensive and costly than identifying the distribution of
costs. Nonetheless, it would seem appropriate for this document to represent the ideal
case as one in which the equity and impacts associated with both benefits and costs are
considered.
In the context of describing the benefits, costs, or net benefits distribution across
populations, it would be useful to describe how the concepts of Lorenz curves and/or
Gini coefficients could be used.
27
-------
Question 10: Economic Literature
Appendix A: Economic Theory was prepared for those readers who wished to have a
better understanding of the economic foundations underlying benefit cost analyses. Does
Appendix A summarize the relevant literature in an objective and meaningful way? Are
there topics that warrant (more) discussion in this appendix that were otherwise missed?
The Appendix provides a thorough and clean discussion of the core economic
foundations relevant to benefit-cost analysis with two exceptions. First, a discussion
explaining the distinction between stock and flow pollutants should be added. A stock
pollutant is an unwanted byproduct of production or consumption that accumulates
through time whereas a flow pollutant does not accumulate. Much of the Guidelines
deals only with a special case where the damage from pollution comes exclusively from
the one-period flow of the pollutant and does not consider the general case where the
damage comes the accumulated stock of the pollutant. This distinction can be important
when undertaking benefit-cost analysis for pollution reduction as the form of the damage
function differs. Since greenhouse gases are a prominent and potentially catastrophic
stock pollutant, this is an especially important topic for the Guidelines to address.
Second, the concept of "user cost" should be defined and explained. User cost
relates to forgone future benefits of a resource. That is, exhaustible resources used today
will not be available for future use. Benefit-cost analysis related to resource stocks will
often need to consider and estimate user costs so its inclusion is important.
28
-------
Question 11: Omissions
Please identify and enumerate any inconsistencies you may find across chapters and
other issues/topics on which we should provide further elaboration. Also, please identify
any definitions provided in the new glossary that are inaccurate or that otherwise need
revision.
Most of our advice on cross-cutting issues is provided in the Introduction to this
Advisory. Below are a few additional ideas that merit consideration.
First and foremost, the Guidelines sorely need case studies and examples to
illustrate and make concepts concrete and meaningful.
International trade in market and nonmarket goods is not adequately covered in
the Guidelines. The discussion of costs, for example, generally assumes a closed
economy without international competition. The discussion of policy instruments
focuses exclusively on the management of "internal" externalities.
Dynamics are another issue receiving inadequate attention in the Guidelines,
although dynamic models are discussed in the CGE chapter. Dynamics become relevant
to policy analysis, and to estimation of benefits and costs in several contexts. One
context is stock pollution problems, climate change being the leading example. Another
is when the costs of pollution control or the benefits of pollution reductions have dynamic
elements. The costs of pollution control have dynamic elements when there are capital
adjustment costs and when there is induced technological change - both features of the
"real world." Benefits of pollution reductions have dynamic elements when pollutants
are stock pollutants, and when those damaged by pollution have capital adjustment costs
in adapting to environmental conditions, and when environmental conditions induce
innovations among those who are damaged. Dynamics in these contexts are important in
benefit and cost estimation.
29
-------
Appendix: Compilation of Line by Line Comments from Individual Members
The following comments are suggested edits from individual panelists. As these
comments were minor and not considered for full panel deliberation, they are listed
separately here for the Agency's consideration.
Page 4-3, Line 13: Drop statement"... does not attempt to detail the relative merits of
putting them into practice ..." . Report discusses merits of various approaches.
Section 4.1: Divide this section into two parts: (i) Efficient Level of Pollution (MD =
MACaggregate) and (ii) Cost-Effective Level of Pollution - minimize cost across sources
(MAC1= MAC2= ... = MACn). Each will need a supporting graph and should be
integrated by showing aggregate MAC is derived from individual MACs. See Field and
Field textbook for dividing the discussion in this way. Also, in discussion of efficient
level of pollution be sure to define social welfare using underpinnings of Pareto
Optimality. In the discussion of efficient level of pollution note that in the presence of
uncertainly, one may prefer to think of the efficient level of pollution as a distribution
about E*.
Section 4.2: Divide this section by (i) design standards and (ii) performance standards for
clarity. They seem to bump up against each other in the discussion. Be clear about the
difference between design standards (technology forcing) and technology based
performance standards. Show uniform and technology-based performance standards
using cost-effectiveness graph introduced in previous section. Uniform standards are not
cost effective but have a low information burden (do not need to know MACs) and
technology-based performance standards are cost effective in principle but have a high
information burden (need to know MACs).
Page 4-6, Lines 2-4: Definition of Command and Control (CAC) is not quite right. CAC
sets requirements on specific firms. Please clarify. Also, consider dropping the
terminology CAC. Simply refer to these directly as design and performance standards.
Page 4-6, Lines 28-29: Drop "...firms are not responsive to price signals ..." and "...
random events and emergencies ...". Emergency argument for standards is ok but it can
be applied in the context of any regulatory approach including market-based approaches.
Page 4-6, Lines 31-32: This sentence is incorrect. Polluters may have face different
design or performance standards under CAC regulations.
Section 4.3: When the section on market based regulations is introduced, state why
economists tend to prefer these controls: they tend to be least costly, have a low
information burden on regulators, and provide incentives for technological advancement.
Monitoring and enforcement costs and other administrative costs, of course, can favor
direct regulation.
30
-------
Page 4-8, Coase Box: Even when the conditions mentioned in the opening sentence of
the third paragraph of the Coase Box are met, the Coasian solution may not be reached
due to asymmetric information and bargaining strategies (ala principle-agent theory)
followed by the parties to the transaction. Please qualify the statement accordingly. Also,
full information is not a necessary condition for a bargain, even a socially efficient
bargain, to take place.
Page 4-9, Line 31: Not sure what is meant by "...when only one permit price exists."
When would more exist? Please clarify or drop.
Section 4.3.1.1: Use cost-effectiveness graph to tell the cap & trade story demonstrating
its property as a least costly instrument. Also, instead of telling the cap and trade story
assuming an efficient level for permit allocations from the start, tell the story that for any
aggregate level of emission there exists a cap and trade will be least costly. Then point
out that the efficient solution is a special case where the permit allocation is efficient
(MD=MAC).
Page 4-11, Lines 23-24: Mention an ambient-based trading scheme directly for dealing
with non-uniform mixing (hot spots) and then explain why, due to administrative costs,
something intermediate (zones) between emissions based scheme and ambient base
scheme maybe efficient.
Section 4.3.2: Expand the discussion of the revenue raising property of the tax and how
it can be used to displace other distortionary taxes.
Section 4.3.2: Again, separate cost effectiveness for any given level of emission from
efficient (Pigouvian) for special case.
Section 4.3.2: Include some discussion of targeting the tax. Do you target the tax on the
pollutant, input, process, or elsewhere? In principle you should place it on damages, but
administrative and monitoring costs may suggest targeting elsewhere. Taxing gas is much
easier and probably as effective as taxing damages from auto emissions.
Somewhat related to the previous point, please note that policy design is dependent on
observable and available information. For example if the policy makers can not observe
actual pollution levels but inputs and practices that cause pollution can be observed and
there are reasonable estimates on relationship between the input and technology use and
the pollution level, then taxes or regulations or fees are based on the imputed pollution
levels. Policy makers should be aware of the available information and design policies to
best take advantage of the available knowledge. In particular identifying sources of
heterogeneity among users and knowing how they affect pollution levels is crucial to
policy design. The proliferation of geographic information systems and remote systems to
obtain data provide new opportunities for policy design. Studies suggest that there is a
significant efficiency gain from policies that adjust to observed heterogeneity relative to
uniform policies (Xabadia et-al 2008). Availability of new sources of or means to obtain
information may lead to redesign of policies - for example, availability of a technology
31
-------
of cheap monitoring of pollution may lead to regulation on taxation based on annual
pollution rather than imputed pollution. Policy makers should reassess policy design and
implementation as technology progresses. See Xabadia, Angels, Goetz, Renan-Ulrich and
Zilberman, David, "The Gains from Differentiated Policies to Control Stock Pollution
When Producers are Heterogeneous". American Journal of Agricultural Economics, Vol.
90, No. 4, pp. 1059-1063.
Section 4.3.3: Divide section by (i) subsidy per unit emission and (ii) other subsidies. It
more or less falls out that way now, but a clear separation would help.
Page 4-14, Lines 2-4: Drop "However, there may be cases in which a subsidy is more
feasible than an emissions tax especially when it is difficult to identify polluters, or when
research and development activities relevant to emission abatement would otherwise be
under-funded." First part of sentence is handled elsewhere and what does "Under-
funded" mean? Is there a market failure here?
Page 4-18, Lines 20-24: Drop this passage. A more meaningful distinction between
liability rules and other regulations is that they are ex post regulations usually used in
cases of accidents or episodic environmental events, not typical flow pollutant cases.
Section 4.4.2: Mention that information disclosure strategies can allow the market create
incentives for pollution control (following Coase) with the victims directly signaling their
preferences to firms. But these are most likely to work when there are contractual
obligations between polluting firms and affected parties (e.g. consumers/workers) and
more difficult to work when they affect third parties (see Tietenberg, "Disclosure
Strategies for Pollution Control," Environmental and Resource Economics, April-June
1998, v. 11, iss. 3-4, pp. 587-602. Also it should be mentioned that credibility of
information is important. Information disclosure can lead to inefficient outcomes when
information is not credible (see Brouhle, K. and M. Khanna, "Information and the
Provision of Quality-Differentiated Goods," Economic Inquiry: 45(2): 377-395, April,
2007).
Section 4.4.3: Move this section into the market incentive sections and expand the
discussion to include issues of limited assets, activity level incentives, courts costs and so
forth. See Segerson (1995) for a nice summary.
Page 4-18, Footnote 54: Another study that shows the types of firms that had incentives
to improve environmental performance following negative stock market returns is
Khanna, M., W. Quimio, and D. Bojilova, "Toxic Release Information: A Policy Tool for
Environmental Protection," Journal of Environmental Economics and Management, 36
(3): 243-266, November 1998.
Page 4-19, Line 19: ".. .information on investment options .." This speaks more to the
public goods property of information provision than it does information disclosure. It has
not really been mentioned until now and is different than information disclosure.
Consider bringing information provision into chapter. There may be a under provision of
32
-------
information on technology following a public goods argument. See Goulder and Parry
(2008).
Section 4.5.3: Weitzman argument. When MAC's are uncertain, and if MD is believed to
be constant or flat, then favor a price instrument. If MD is believed to have thresholds or
be vertical then favor a quantity instrument. This avoids making costly mistakes. This
message does not come across clearly in this section. Also, the "degree of uncertainty" is
given as a factor in choosing among policies. This is not technically correct. For
example, large uncertainty about emissions is of no policy consequence if the marginal
damage cost is constant.
Page 4-22, Line 31: It is incorrect to say that voluntary programs require firms to set
goals (also see line 38) or to say that they definitely achieve environmental
improvements. Also, it is not accurate to say that most voluntary programs set goals (line
37). They also do not make it simpler to monitor and measure if participants are meeting
the goal - most voluntary programs do not require firms to provide emissions data to the
EPA (for a review of these issues and comparison of programs see Khanna, M. and D.T.
Ramirez, "Effectiveness of Voluntary Approaches: Implications for Climate Change
Mitigation," in Voluntary Agreements in Climate Policy, edited by A. Baranzini and P.
Thalmann, Edward Elgar Publishers, pp. 31-66, 2004.)
Page 4-23, Footnote 63: The references on voluntary programs are a little outdated. A
more recent and updated citation on motivations to participate in different types of
voluntary programs, challenges in evaluating their effectiveness and evidence about their
effectiveness is Khanna, M. and K. Brouhle, "Effectiveness of Voluntary Environmental
Initiatives," Chapter 6 in Governing the Environment: Interdisciplinary Perspectives, ed.
By M. Delmas and O. Young, Cambridge University Press, Cambridge, U.K.
(forthcoming). Khanna, M., "The 33/50 program: Program Design and Effectiveness," in
Reality Check: The Nature and Performance of Voluntary Environmental Programs in
the United States, Europe and Japan edited by R. D. Morgenstern and W. Pizer, RFF
Press, Washington DC, 15-42, 2007.
Page 4-24: The box on water quality trading directs people to certain EPA guidance on
the design of trading programs. While we think there is merit to these documents, I have
reservations about fully endorsing them because there is economically flawed advice
about some design elements. This box also misses an opportunity to discuss the
importance of economic science to the design of markets. Contemporary water quality
markets fail to achieve the promise of trading because of participation and coordination
failures that occur in part because of flaws in market design and development. On
balance, I recommend dropping the box.
Page 5-2, Line 29: Recommending that the analyst provide " A clear written statement
about the current state of the economy.." is too broad and may be unnecessary. Instead it
is important to clearly specify the current and future state of economic variables that are
relevant for the analysis.
33
-------
Page 5-8, Lines 1-4. The Guidelines mention that regulations can lead to innovation
(which may not occur in the baseline) which can lead to cost savings. They also mention
that there is no statistical evidence supporting this claim of cost savings. Hence they
suggest that analysts should avoid assuming differing rates of technological innovation
based on regulatory stringency. That last statement (line 4) is incorrect and should be
deleted. Instead there should be a brief discussion about the potential for induced
innovation due to environmental regulations. Regulations can create incentives for
technological innovation that lower the cost of compliance. It should also be recognized
that regulations may also discourage innovation in some cases or crowd out other
innovations.
Page 6-2, Line 21: NEARLY "any policy"
Page 6-2, Line 37: Can do half-cycle correction, i.e., assume effects occur at mid-year.
Page 6-2, Footnote 73: refers to exponential fn that was apparently deleted from draft
Page 6-4, Line 25 : Note NPV = NFV /(l+r)T.
Page 6-7, Lines 18-20: Text is duplicative
Page 6-10, Footnote 80: Qualify that result requires rate of return > consumption interest
rate.
Page 6-15, Section 6.3. Although this section on "intergenerational discounting" is
concerned with long time horizon problems, much of the text refers to climate change
(e.g., paragraph beginning Page 6-16, Line 6). This is understandable (as that is the most
prominent long-horizon problem) but the text could be revised to avoid the impression
that it is only about climate change.
Page 7-9, Lines 33-34: The text here mentions possible approaches for valuing
morbidity—ex ante and ex post. Please note that studies have also varied in whether they
controlled for the opportunity to mitigate the illness (e.g., before or after taking
medication).
Page 7-9, Lines 9-10: Recommend not listing fetal loss as a non-fatal health effect -
some people would disagree.
Page 7-11, Line 20: Better to say that social costs include private costs (reflected in
individual WTP) and external/public costs (e.g., medical care expenses paid by insurance
or public sources). The existence of externally paid costs does not mean the individual
"understates" own WTP.
Page 7-17, Lines 2-4:1 guess this statement is true, but a tighter bound would be one
shouldn't include effect for which cost of gathering information exceeds expected
improvement in net benefits from choosing a better policy given this information. But I
34
-------
don't want to make too much of the analytic cost point - seems to me that analytic costs
are very small compared with B and C of the major regulations we're mostly considering
(e.g., >100 M/yr for many years).
Page 7-19, Lines 15-16: It would help to clarify discussion of when multiple methods are
substitutes or complements. In some cases different components will be captured by
different methods (e.g., private WTP to reduce health risk, public cost of illness of
treating illness), in which case it is appropriate to use both methods and add results
(making sure they do not double-count subcomponents). In other cases where different
methods provide alternative estimates of the same component, multiple estimates may be
useful to triangulate on most accurate value, but should not be added.
Page 7-26, Line 22: Re opportunity cost of time, what if driving to the recreational site is
itself enjoyable? After all, in some cases much of the benefit of travel is the journey, not
the destination. I'm sure this is discussed in travel cost literature, but I don't follow it.
Page 7-28, Lines 9-11: Strike this sentence. We have no compelling evidence that
instantaneous workplace deaths reflect the same tradeoffs that individuals are willing to
make over environmental risks.
Page 7-28, line 17: replace "believe" with "assume."
Page 7-28, line 21: the work by Black and Kniesner (2003) uses only risk measures
known be fraught with measurement errors (even the "best" data was based on
inconsistent reporting of deaths (see Drudi, 1997) and aggregated in a manner that creates
serious endogeneity problems (see Leigh, 1995 (JEEM) and Mrozek and Taylor, 2002
(JPAM)). The discussion here could place Black and Kniesner in this context and then
look forward to the newer literature.
Page 7-28: The discussion of wage-risk studies needs to be updated with recent results by
Kniesner et al. (2007) and Viscusi and Aldy (2007).
Page 7-28: Text on hedonics. What's the point of placing the discussion of the source of
workplace risk data where they are now? It's a non-sequitur, and the text does not
elaborate on the implications of these sources and of the level of resolution of the
workplace risk data.
Page 7-28: Disagree with "Further, while estimates from the hedonic literature have been
relatively consistent over the years, questions persist about..." Au contraire, Costa and
Kahn (2004) find that in the US the compensating wage differentials required by workers
to accept riskier jobs have grown, while workplace risks have declined, resulting in VSLs
that have grown over time. Liu and Hammitt (1997) have likewise found that the
compensating wage differentials have grown in Taiwan over 16 years, resulting in
progressively larger VSLs.
35
-------
Page 7-29, line 18 - 19: The Ashenfelter and Greenstone (2004) study is not a hedonics
type of study! It's a revealed preference study of agency choices from which a VSL is
inferred.
Page 7-30, lines 1-10: Would this section be more appropriate in the valuation of
mortality/morbidity section?
Page 7-30: Sources of Risk, first paragraph. It would be useful to cite Eeckhoudt and
Hammitt (2001) and also Evans and Smith (2006) who expanded on the notion of
competing v. specific risks.
Page 7-31, Lines 36-38: Please explain the following sentence: "For example, if
perceived risks are found to be lower than expert risk estimates, then WTP can be
estimated with the lower, perceived risk (Blomquist, 2004)."
Page 7-35, Line 44: strike "with minimal additional assumptions".
Page 7-39, Line 29: change "experience, especially..." to read "experience in posted-
price markets, especially..."
Page 7-39, Line 12: John Quiggin prefers to spell his last name "Quiggin" and not
"Quiggen."
Page 7-39, Lines 2-5: Better to say statistical precision than efficiency. As I understand it,
efficiency refers to making most use of information available in the data; it is not for
comparing different datasets.
Page 7-39, Lines 38-42: Can you please provide some more recent applications of stated-
preference studies using "multi-attribute choice questions?"
Page 7-40, Lines 37-41: Suggesting that choice questions allow someone to "express
support for a program" is counter to the goal of using this method to estimate an actual
WTP (not some general notion of "support").
Pages 7-40 to 7-43: Much of the material feels old and outdated... Assertions about the
properties of SP data and the influence of survey design on SP responses are introduced
throughout the stated preference section without proper citations to supporting evidence
(e.g., page 7-39, lines 27-36, lines 31-33, and lines 34-36).
Page 7-42, Lines 2-10: The section on Criterion validity over-simplifies the issues. First,
for public goods, what is meant by market data? (voluntary contributions markets?
political markets?) One should note that theoretically demand revealing mechanisms for
public goods are difficult to implement (e.g., Groves-Ledyard), which makes testing the
validity of SP surveys for public goods that much more difficult. There have been some
studies which compare hypothetical voting on a public good to later actual referenda on
36
-------
such goods (e.g. Johnston, 2006,) - but this type of criterion test will not generally be
available for the policy outcomes being considered by the EPA.
Page 7-42, lines 12-28: For studies that focus on policy-relevant outcomes for the EPA,
convergent validity tests will generally not be available (since the goods being considered
are not actually "deliverable" by the researcher). Even if similar goods could be
delivered, if they are public goods it will be difficult to assure the RP (actual transaction)
data are free from biases associated with public good provision. (This comment
essentially reiterates the comment above for lines 2-10 on the same page.)
Page 7-43: Survey non-response bias is "created by those who refuse to take the survey"
only if their WTP is systematically different from that of those persons who did take the
survey. Please make this point clear.
Page 7-44, Line 14, etc.: I don't like the emphasis on doing a new study for each
endpoint, since I think that RP & SP studies have so many validity/reliability concerns.
On the contrary, I think we are on much stronger grounds having several studies that are
relevant from which we can transfer estimates. This may depend on endpoint - for VSL, I
want to know what multiple studies say; for some unique ecosystem, I would not care so
much about estimated values for effects on other ecosystems.
Page 8-5, Line 44: "In reality, deadweight losses already exist in many if not most
markets as a result of taxes, regulations, and other distortions" But presumably, in many
cases the regulation will have the goal of correcting existing distortions, such as external
costs of pollution. The term "deadweight loss" here is not really a loss if you are
considering policies designed to correct externalities. Although the report recognizes that
benefits of pollution reduction need to be considered, the term "loss" is not really
appropriate here. This is more an issue of terminology, rather than substance.
Section 8.1.1: One of the down sides of partial equilibrium approaches is the possibility
of "double counting" impacts. To this end, a clear warning or explanation of doubling
counting should be included in the discussion in chapter 9 or chapter 8. Double counting
occurs if the outputs from firms operating upstream and downstream are both impacted
by the new policy and the impacts are considered separately. The impact on the
downstream firm is typically passed on to the upstream firm. For a simple example,
consider a fuel policy affecting both a delivery business and a local production business.
If one of the local business inputs comes from the delivery business and the policies
impact on the local business through this input is including in the partial equilibrium
adjustment for the local business, any partial equilibrium analysis for the delivery
business should account for this. This warning is especially of interest because the
apparent interest in CGE analysis may lead an analyst to estimate multiple partial
equilibrium models in place of a CGE due to their relative difficulty.
Section 8.1.3: I wouldn't use the term "economic impacts" since that term generally
refers to Economic Impact Analysis, so the term might be incorrectly interpreted. Why
not "economic effects".
37
-------
Section 8.2.1: Alternative Concepts of Cost. The discussion in this section is confusing.
If I understand it correctly, these are not alternative concepts of costs, but rather are
categories of cost. If properly defined, these categories are simply decompositions of
social cost, not "alternative concepts". Does this Section intend to indicate that, for
efficiency purposes, only total social costs matter, but decompositions of social cost
could provide information on the incidence of costs or the distributional consequences?
EPA either needs to explain how these categories are useful, or if they are not useful the
discussion could simply be dropped.
Section 8.3.2.1: The report should be more consistent in differentiating between benefits
and costs. For example, this section refers to things like "irreversible environmental
impacts" as a cost. While it is true that there is no true distinction between costs and
negative benefits (e.g., irreversible environmental impacts), the report has separate
sections on costs and benefits, and the discussion should be kept consistent by including
environmental values in the benefits section.
Section 8.4.4: Input-Output analysis doesn't really belong in this section, since it does
not provide a well defined measure of economic costs. We recommend that the
discussion be moved to Chapter 9 given its close connection to Economic Impact
Analysis and distribution across sectors. Note that Input-Output Analysis does not
provide a measure of economic costs and benefits, but output from an I-O analysis could
provide some information on how economic effects are distributed across sectors of the
economy.
Section 8.2.3. Shouldn't this Section refer to "Distribution of Costs" rather than
"Distributional Costs"
Page. 9-1. Footnote 179—avoiding double-counting likely needs more discussion and an
example to illuminate the issue.
Page 9-7 and Table 2: Data sources for profiles I would like to add The Thomas Registry
is another data source. The Thomas Register, which dates back to 1906, is used primarily
by purchasing agents. Lavin [1992] states that the Thomas Register is the best example
of a directory which provides information on manufacturers by focusing on products.
According to Lavin, "The Thomas Register is a comprehensive, detailed guide to the full
range of products manufactured in the United States. Covering only manufacturing
companies, it strives for a complete representation within that scope." The EPA should
also see the many other types of sources of business information discussed in Lavin, M.
R., 1992, Business Information: How to Find It, How to Use It, 2nd ed. (Oryx Press,
Phoenix).
Page 9-9, Line 15. They basically punt on pass-through. Perhaps more direction on what
to do when basic elasticities are not available.
38
-------
Page 9-9, Lines 41-42. Is this a cost or benefit? I see it as a net benefit that should be
estimated somehow.
Page 9-9, Footnote 192: This then leads to the producers of equipment to dig out the coal
being hurt. Be clear about how far down the chain we go.
Page 9-12, Lines 20-22: Regional analysis. Lacks a thorough discussion of regional
economies and trade. In this way the document can update FN 204 by including work of
Copeland and Taylor and Greenstone.
Page 9-16, Lines 10-12: Another indicator that could be considered at the community
level is the foreclosure rate.
Page 9-18, Lines 1-8: This gets us back to the proper counterfactual. In this example
they merely discuss the direct cost of the regulation without recognizing that these
expenditures have other benefits and costs. For example, they confer tax breaks
(complying with regulations is a deductible expense) and that the new capital is more
productive than old capital. But a key consideration is whether, and to what extent, the
displacement of investment leads this new capital to be less productive than innovation
that it displaced.
Page 9-18 - 9-20: Some mention should concern temporal aspects of benefits and costs.
For example, the entirety of Ch. 9 contains sections on equity issues for an analysis to
consider. In addition to this, a discussion of household movement (Tiebout sorting) may
be of interest to account for the long-term equitable distribution. That is, although there
may be short run benefits for socially or economically disadvantaged populations, they
may not hold in the long run. If these households are not home owners, they may be left
out of the gain in benefits if market forces result in disadvantage populations moving
because the gains remain attached to home or land and therefore the owner.
Page 9-23: Textbox "2" should be "9.2".
Page 9-26: Extra period in box 9.3.
Page 9-29: What is the definition of poor? What about gender? This seems to be a
relatively ad hoc list of equity factors, are there other identifying characteristics that
might be relevant? What about intergenerational equity? .
Page B-4, Footnote 293: Sunstein (1997) and Hammitt and Liu (2004) are more recent
citations.
Page B-7, Footnote 309: Adjustments in the VSL for population characteristics "does
imply" (not "may imply") support for variation in protection across the population.
Page A-2, line 4: change "y" to "P" and "x" to "Qd" to be consistent with Figure A.I.
39
-------
Page A-2, line 19: insert language so sentence reads "The total WTP is equal to the SUM
OF THE marginal WTP for each unit up to Q4"
Figure A.3: "P" on the vertical axis needs a "m" subscript to be consistent with the text.
Page A-8, line 7: Change sentence to read: "Benefit-cost analysis can also be SEEN as a
type of..."
Page A-8, line 8: We suggest striking "that economists strive to avoid" from the sentence.
Figure A.6: The demand curve could be labeled in the figure directly (especially given
there is no title for the figure).
Page A-12, line 32: Change "correct monetary measures of utility change" to read "exact
monetary measures of utility change" to be consistent with standard language found in
the literature.
Page A-15, line 12: Strike "However," — it is out of place given the preceding sentence.
Figure A. 10: Consider changing the title on the horizontal axis from "Regulation" to
"Pollution Abatement" or something similar.
40
-------
References
Adamowicz, W., P. Boxall, M. Williams, and J. Louviere. (1998) "Stated preference
approaches for measuring passive use values: Choice experiments and contingent
valuation". American Journal of Agricultural Economics 80:64-75.
Alberini, Anna and Alan Krupnick. (2000) "Cost-of-Illness and WTP Estimates of the
Benefits of Improved Air Quality: Evidence from Taiwan," Land Economics 76(1), 37-
53.
Alberini, Anna and Aline Chiabai. (2007) "Urban Environmental Health and Sensitive
Populations: How Much Are the Italians Willing to Pay to Reduce their Risks?" Regional
Science and Urban Economics 37(2), 239-258.
Ashenfelter, O. and M. Greenstone. (2004) "Estimating the Value of a Statistical Life:
The Importance of Omitted Variables and Publication Bias," American Economic Review
Papers and Proceedings 94(2).
Baumol, William J. and Wallace E. Gates. (1975) The Theory of Environmental Policy.
Englewood Cliffs, NJ: Prentice Hall.
Bayer, Patrick, Nathaniel Keohane, and Christopher Timmins. 2009. "Migration and
Hedonic Valuation: The Case of Air Quality." Journal of Environmental Economics and
Management., 58(1).
Bayer, Patrick, Robert McMillan, and Kim Reuben. 2004. "An Equilibrium Model of
Sorting in an Urban Housing Market." NBER Working Paper.
Bayer, Patrick and Christopher Timmins. 2005. "On the Equilibrium Properties of
Locational Sorting Models." Journal of Urban Economics, 57(3): 462-77.
Bayer, Patrick and Christopher Timmins. 2007. "Estimating Equilibrium Models of
Sorting across Locations." The Economic Journal, 117(518): 353-74.
Bishop, R. and T. Heberlein. (1979) "Measuring Values of Extra-Market Goods: Are
Indirect Measures Biased?" American Journal of Agricultural Economics 61: 926-930.
Bishop, Richard C., Thomas A. Heberlein and Mary Jo Kealy. (1983) "Contingent
Valuation of Environmental Assets: Comparison with a Simulated Market, "Natural
Resources Journal 23 (2), 619-63 3.
Black, Dan A. and Thomas J. Kniesner. (2003) "On the Measurement of Job Risk in
Hedonic Wage Models." Journal of Risk and Uncertaint, 27(3): 205-220.
41
-------
Boardman, A., D. Greenberg, A. Vining and D. Weimer. (2006) Cost Benefit Analysis:
Concepts and Practice. Prentice Hall.
Bockstael, N.E. and K. E. McConnell. (2008) Environmental and Resource Valuation
with Revealed Preferences: A Theoretical Guide to Empirical Models. Springer.
Bohm, P. (1972) "Estimating Demand for Public Goods: An experiment" European
Economics Review Vol. 3.
Brouhle, K. and M. Khanna. (2007) "Information and the Provision of Quality-
Differentiated Goods," Economic Inquiry 45(2): 377-395.
Carbone, Tared C. and V. Kerry Smith "Evaluating Policy Interventions with General
Equilibrium Externalities" Journal of Public Economics, June 2008 .
Carlsson, F. and P. Martinsson. (2001) "Do Hypothetical and Actual Marginal
Willingness to Pay Differ in Choice Experiments?" Journal of Environmental Economics
and Management 41: 179-92.
Carson, R. T. (2007) The Stated Preference Approach to Environmental Valuation Vol. I-
III. Ashgate: the International Library of Environmental Economics and Policy.
Carthy, Trevor, Susan Chilton, Judith Covey, Lorraine Hopkins, Michael Jones-Lee,
Graham Loonies, Nick Pidgeon, and Anne Spencer. (1999) "On the Contingent Valuation
of Safety and the Safety of Contingent Valuation: Part 2— The CV/ SG "chained"
Approach", Journal of Risk & Uncertainty 17: 187-213.
Chestnut. L. G.. S. Colome. L. R. Keller, W. Lampert. R. Rowe, and S. L.
Wojciechowski. (1988) "Heart Disease Patients' Averting Behavior. Costs of illness, and
Willingness to Pay to Avoid Angina Episodes." EPA Report 230-10-88-042.
Washington: EPA. Office of Policy Analysis.
Champ, P.A., K.J. Boyle, and T.C. Brown, Eds. (2003) A Primer on Non-Market
Valuation. Kluwer Academic Press.
Costa, Dora L., and Matthew E. Kahn. (2004) "Changes in the value of life, 1940-1980",
Journal of Risk & Uncertainty 29: 159-180.
Cummings, R. and L. Taylor. (1999) "Unbiased Value Estimates for Environmental
Goods: A Cheap Talk Design for the Contingent Valuation Method," American
Economic Review 89:649-665.
Cummings, R., S. Elliott, G. Harrison, and J. Murphy. (1997) "Are Hypothetical
Referenda Incentive Compatible?" Journal of Political Economy 105: 609-621.
Davis, L. (2004) "The effect of health risk on housing values: evidence from a cancer
cluster," American Economic Review 94:1693-1704.
42
-------
Drudi, D. (1997) "A Century-Long Quest for Meaningful and Accurate Occupational
Injury and Illness Statistics". BLS Compensation and Working Conditions Winter 19-27.
Duffield, J. and J. Patterson. (1992) "Field Testing Existence Values: An Instream Flow
Trust Fund for Montana Rivers" manuscript.
Duffield, John W. and David A. Patterson. (1991) "Inference and Optimal Design for a
Welfare Measure in Dichotomous Choice Contingent Valuation," Land Economics 67(2),
255-239.
Eeckhoudt, L.R., and J.K. Hammitt. (2001) Background Risks and the Value of a
Statistical Life, Journal of Risk and Uncertainty 23: 261-279.
Ellickson, Bryan. 1971. "Jurisdictional Fragmentation and Residential Choice." American
Economic Review, 61(2): 334-39.
Environmental Protection Agency Science Advisory Board. (2009) Valuing the
Protection of Ecological Systems and Services.
Epple, Dennis and Holger Sieg. 1999. "Estimating Equilibrium Models of Local
Jurisdiction." Journal of Political Economy, 107(4): 645-81
Espinosa, J. Andres and V. Kerry Smith "Measuring the Environmental Consequences of
Trade Policy: A Non-Market CGE Analysis," American Journal of Agricultural
Economics., August 1995.
Evans, Mary F., and V. Kerry Smith. (2006) "Do we really understand age-VSL
relationship?" Resource and Energy Economics 28: 242-261.
Field, Barry C. and Martha K. Field. (2009) Environmental Economics: An Introduction.
New York: McGraw-Hill Publishing.
Gayer, Ted, James T. Hamilton, and W. Kip Viscusi. (2000) "Private Values of Risk
Tradeoffs at Superfund Sites: Housing Market Evidence on Learning About Risks,"
Review of Economics and Statistics 82(3), 439-451.
Gayer, Ted, James T. Hamilton, and W. Kip Viscusi. (2002) "The Market Value of
Reducing Cancer Risk: Hedonic Housing Prices with Changing Information," Southern
Economic Journal 69(2), 266-289.
Gollier, Christian. (2004). "Maximising the Expected Net Future Value as an Alternative
Strategy to Gamma Discounting," Finance Research Letters 1: 85-89.
Goulder, Lawrence H. and Ian W. H. Parry. (2008) "Instrument Choice in Environmental
Policy", Review of Environmental Economics and Policy 2(2): 154 - 174.
43
-------
Greenstone, Michael and Justin Gallagher. (2008) "Does Hazardous Waste Matter?
Evidence from the Housing Market and the Superfund Program", Quarterly Journal of
Economics 123(3):951-1003.
Hammitt, J.K., and J.-T. Liu. (2004) "Effects of Disease Type and Latency on the Value
of Mortality Risk," Journal of Risk and Uncertainty 28: 73-95,.
Hammitt, J.K., J.-T. Liu, and J.-L. Liu. (2000) "Survival is a Luxury Good: The
Increasing Value of a Statistical Life," manuscript, August 2000.
Hammitt, James K. (2002) "QALYs versus WTP," Risk Analysis 22(5): 985-1001.
Hammitt, James K. (2009) "Saving Lives: Benefit-Cost Analysis and Distribution,"
University of Pennsylvania Law Review Pennumbra 157: 189-199.
Hepburn, Cameron, and Ben Groom. (2007). "Gamma Discounting and Expected Net
Future Value," Journal of Environmental Economics and Management 53: 99-109.
Herriges, Joseph and Catherine Kling (Eds.). (1999) Valuing Recreation and the
Environment: Revealed Preference Methods in Theory and Practice, Darmstadt,
Germany: Springer.
Herriges, Joseph and Gathering Kling (Eds.) (2008) Revealed Preferences Approaches to
Environmental Valuation, Ashgate,
Hintermann, Beat, Anna Alberini, and Anil Markandya. (2008) "Estimating the Value of
Safety with Labor Market Data: Are the Results Trustworthy?" forthcoming in Applied
Economics, available at
http://www.informaworld.com/smpp/ti tle~content=g769588680~db=all~tab=forthcoming
?stem=3#messages.
Johannesson, M., and B. Jonsson. (1991). "Willingness to Pay for Antihypertensive
Therapy—Results of a Swedish Pilot Study," Journal of Health Economics 10: 461-474.
Johannesson, M., et al. (1993) "Willingness to Pay for Antihypertensive Therapy--
Further Results," Journal of Health Economics 12: 95-108.
Johannesson, Magnus, Bengt Jonsson and Lars Borquist. (1991), "Willingness to Pay for
Antihypertensive Therapy—Results for a Swedish Pilot Study" Journal of Health
Economics 10:461-474.
Johnston, Robert J. (2006) "Is hypothetical bias universal? Validating contingent
valuation responses using a binding public referendum" Journal of Environmental
Economics and Management 52(1): 469-48.
Just, Richard E., Andrew Schmitz, and Darrell L. Hueth. (2005) Welfare Economics of
Public Policy: A Practical Approach to Project and Policy Evaluation Edward Elgar
Publishing.
44
-------
Khanna, M. and D.T. Ramirez. (2004) "Effectiveness of Voluntary Approaches:
Implications for Climate Change Mitigation," in Voluntary Agreements in Climate
Policy, edited by A. Baranzini and P. Thalmann, Edward Elgar Publishers, pp. 31-66.
Khanna, M. and K. Brouhle. (forthcoming) "Effectiveness of Voluntary Environmental
Initiatives," Chapter 6 in Governing the Environment: Interdisciplinary Perspectives, ed.
By M. Delmas and O. Young, Cambridge University Press, Cambridge, U.K.
Khanna, M. (2007) "The 33/50 program: Program Design and Effectiveness," in Reality
Check: The Nature and Performance of Voluntary Environmental Programs in the United
States, Europe and Japan edited by R. D. Morgenstern and W. Pizer, RFF Press,
Washington DC, 15-42.
Khanna, M., W. Quimio, and D. Bojilova. (1998) "Toxic Release Information: A Policy
Tool for Environmental Protection," Journal of Environmental Economics and
Management 36 (3): 243-266,.
Kniesner, T.J., W.K. Viscusi, and J.P. Ziliak. (2006) "Life-cycle Consumption and the
Age-Adjusted Value of Life, Contributions to Economic Analysis and Policy 5(1) article
4.
Kochi, I. (2008) "Endogeneity and Estimates of the Value of a Statistical Life," Colorado
State University working paper.
Kokoski, Mary and V. Kerry Smith "A General Equilibrium Analysis of Partial
Equilibrium Welfare Measures: The Case of Climate Change," American Economic
Review, June 1987; reprinted in H. Kunreuther and A.Rose, editors, The Economics of
Natural Hazards (Cheltenham, U.K.: Edward Elgar, 2004).
Krupnick, Alan. "Mortality-Risk Valuation and Age: Stated Preference Evidence,"
Review of Environmental Economics and Policy (2007): 216-282.
Lavin, M. R. (1992) Business Information: How to Find It How to Use It 2nd ed. Oryx
Press, Phoenix.
Leigh, J.P. (1995) "Compensating Wages, Value of a Statistical Life, and Inter-industry
Differentials", Journal of Environmental Economics and Management 28: 83-97.
List, John A. (2006) "Using Choice Experiments to Value Non-Market Goods and
Services: Evidence from Field Experiments" The B.E. Journal of Economic Analysis and
Policy 6 (2).
Lusk, Jay son L. and F. Bailey Norwood. (2005) "Effect of Experimental Design on
choice-Based Conjoint Valuation Estimates" American Journal of Agricultural
Economics 87(3):771-785.
45
-------
List, John A. and Craig A. Gallet. (2001) "What Experimental Protocol Influence
Disparities between Actual and Hypothetical Stated Values? Environmental and
Resource Economics 20(3):241-254
List, John. (2001) "Do Explicit Warnings Eliminate the Hypothetical Bias in Elicitation
Procedures? Evidence from Field Auctions for Sportscards," American Economic Review
91 (5):1498-1507.
Lusk, J., D. Fields, and W. Prevatt. (2008) "An Incentive Compatible Conjoint Ranking
Mechanism" American Journal of Agricultural Economics 90: 487-98.
Mariah D. Ehmke, Jay son L. Lusk, and John A. List. (2008) "Is Hypothetical Bias a
Universal Phenomenon? A Multinational Investigation," Land Economics, 84(3), 489-
500
Morgenstern, Richard and William Pizer. (2007) Reality Check: The Nature and
Performance of Voluntary Environmental Programs in the United States, Europe, and
Japan. Resources for the Future, RFF press.
Mrozek, Janus R., and Laura O. Taylor. (2002) "What determines the value of life? A
meta-analysis", Journal of Policy Analysis and Management 21: 253-270.
National Research Council. (2004) Valuing ecosystem services; toward better
environmental decision-making. Washington, DC: National Academies Press.
National Research Council. (2002) New Tools for Environmental Protection, National
Academy Press.
Neill, H., R. Cummings, P. Garderton, G. Harrison, and T. McGuckin. (1994)
"Hypothetical Surveys and Real Economic Commitments," Land Economics 70(145-154.
Newell, R.G. and William Pizer. (2003) Discounting the Distant Future: How Much do
Uncertain Rates Increase Valuations? Journal of Environmental Economics and
Management 46(1): 43-59.
Opaluch, James J., Stephen K. Swallow, Thomas Weaver, Christopher Wessells and
Dennis Wichelns. (1993) "Evaluating Impacts from Noxious Facilities: Including Public
Preferences in Current Siting Mechanisms" Journal of Environmental Economics and
Management 24(1).
Portney, P. (1992) Trouble in Happy ville, Journal of Policy Analysis and Management
11: 131-132.
Herriges, J.A. and Cathy L. Kling, Eds. (2008) Revealed Preferences Approaches to
Environmental Valuation. The International Library of Environmental Economics and
Policy, Volumes I and II, Ashgate Publishing.
46
-------
Rowe, R. D.. and L. G. Chestnut. (1985) "Oxidants and Asthma in Los Angeles: A
Benefit Analysis. EPA Report 230-07-85-010. Washington, DC . Office of Policy
Analysis.
Segerson, Kathleen and Thomas Miceli. (1998) "Voluntary Environmental Agreements:
Good or Bad News for Environmental Protection?" Journal of Environmental Economics
andManagement 36(2): 109-30.
Segerson, Kathleen. (1994) "Property Transfers and Environmental Pollution: Incentive
Effects of Alternative Policies." Land Economics 70(3): 261-72.
Sieg, Holger, V. Kerry Smith, H. Spencer Banzhaf, and Randy Walsh. 2002.
"Interjurisdictional Housing Prices in Location Equilibrium." Journal of Urban
Economics, 52(1): 131-53.
Sieg, Holger, V. Kerry Smith, H. Spencer Banzhaf, and Randy Walsh. 2004. "Estimating
the General Equilibrium Benefits of Large Changes in Spatially Delineated Public
Goods." International Economic Review, 45(4): 1047-77.
Smith, V. Kerry, Holger Sieg, H. Spencer Banzhaf, and Randy Walsh. 2004. "General
Equilibrium Benefits for Environmental Improvements: Projected Ozone Reductions
under EPA's Prospective Analysis for the Los Angeles Air Basin." Journal of
Environmental Economics and Management, 47(3): 559-84.
Smith, V. Kerry and J Andres Espinosa"Environmental and Trade Policies: Some
Methodological Lessons," Environment and Development Economics, February 1996;
reprinted in A. A. Batabyal and H. Beladi, editors, The Economics of International Trade
and the Environment (New York: Lewis Publishers, 2001).
Smith, V. Kerry and Tared C. Carbone "Environmental Economics and the "Curse" of the
Circular Flow" in Jun lie Wu, P. Barkley, and B. Weber, editors, Frontiers of Rural and
Resource Economics, (Washington, D.C.: Resources for the Future, 2008)
Sunstein, C.R. & R.H. Thaler. (2003) Libertarian Paternalism, AER Papers &
Proceedings 93: 175-179.
Sunstein, Cass. (1997) "Bad Deaths" Journal of Risk and Uncertainty 14: 259-282,.
Tietenberg, Tom. (2008) "Disclosure Strategies for Pollution Control," Environmental
and Resource Economicsll(3-4): 587-602.
Unsworth, R.E., and R.C. Bishop. (1994) "Assessing Natural Resource Damages Using
Environmental Annuities ." Ecological Economics 11:35-41.
Viscusi, W. Kip, and Joseph E. Aldy. (2007) Labor Market Estimates of the Senior
Discount for the Value of Statistical Life, Journal of Environmental Economics and
Management 53: 377-392.
47
-------
Viscusi, W.Kip. (2004) "The Value of life: Estimates with Risks by Occupation and
Industry." Economic Inquiry 42: 29-48.
Walsh, Randall L. 2007. "Endogenous Open Space Amenities in a Locational
Equilibrium." Journal of Urban Economics, 61(2): 319-44.
Weitzman, Martin L. (2001) Gamma discounting, American Economic Review 91: 261-
271.
Weitzman, Martin L. (1998) "Why the Far-Distant Future Should be Discounted at its
Lowest Possible Rate" Journal of Environmental Economics and Management 36: 201-
208.
Whitehead, J.C., S.K. Pattanayak, and B.R.Gelso. (2008) "Combining Revealed and
Stated Preference Data to Estimate the Nonmarket Value of Ecological Services: An
Assessment of the State of the Science" Journal of Economic Surveys 22(5):872-908.
Xabadia, Angels, Renan-Ulrich Goetz, and David Zilberman. (2008) "The Gains from
Differentiated Policies to Control Stock Pollution When Producers are Heterogeneous".
American Journal of Agricultural Economics 90(4): 1059-1063.
48
------- |