Science Advisory Board Workshop Summary: Science for Valuation of
EPA's Ecological Protection Decisions and Programs;
Workshop Held December 13-14, 2005, Washington D.C.
-------
Table of Contents
1. WORKSHOP BACKGROUND AND OBJECTIVES 3
2. INTRODUCTORY REMARKS AT THE WORKSHOP 5
3. KEYNOTE PRESENTATION - GLOBAL VIEW FROM THE PERSPECTIVE OF THE
MILLENNIUM ECOSYSTEM ASSESSMENT 6
4. INTRODUCTION TO C-VPESS WORK ON AN "EXPANDED AND INTEGRATED
APPROACH" FOR VALUING ECOLOGICAL PROTECTION 16
5. PANEL DISCUSSION WITH EPA SENIOR MANAGERS 21
6. OVERVIEW OF METHODS BEING CONSIDERED BY C-VPESS 27
7. ADDRESSING UNCERTAINTY IN ECOLOGICAL VALUATION AND EXPERT
ELICITATION 37
8. DISCUSSION OF SPECIFIC METHODS FEATURED IN WORKSHOP BREAKOUT
GROUPS 43
8.1. Break-out Report: Economic Analysis and Ecological Production Functions 43
8.2. Group Expressions of Value: Referenda and Citizen Juries 44
8.3. Deliberative Approaches for Modeling, Valuation, and Decision Making 48
8.4. Social/Psychological Methods for Ecosystem Values Assessments 53
8.5. Spatial Representation of Biodiversity and Conservation Values and Ecological
Services 55
9. PANEL DISCUSSION: EXPERTS' FEEDBACK ON VALUATION METHODS 59
10. SUMMARY AND NEXT STEPS 79
11. AGENDA AND LIST OF INVITED PARTICIPANTS 80
12. BACKGROUND MATERIAL AND PRESENTATIONS FOR BREAKOUT GROUPS 89
12.1. Economic analysis and ecological production functions 89
12.2. Group Expressions of Value: Referenda; Citizen Juries 92
12.3. Deliberative Approaches for Modeling, Valuation, and Decision Making 108
12.4. Social/Psychological Methods for Ecosystem Values Assessments 122
12.5. Spatial Representation of Biodiversity and Conservation Values and Ecological
Services 125
2
-------
1. Workshop Background and Objectives
The EPA Science Advisory Board (SAB) held a public workshop on December
13-14, 2005 in Washington D.C. on "Science for Valuation of EPA's Ecological
Protection Decisions and Programs." The purpose of the workshop was to discuss the
initial work of the SAB's Committee on Valuing the Protection of Ecological Systems
and Services (C-VPESS); to provide an opportunity for members of the SAB, the
Advisory Council on Clean Air Compliance Analysis (Council), and Clean Air Scientific
Advisory Committee (CASAC) to learn from each others' work relating to ecological
valuation; and to feature feedback and insights from Agency clients and outside subject
matter experts. The agenda included presentations and discussions with advisory
committee members, Agency personnel, and invited experts (agenda and list of invited
participants included in section 11 of this workshop report).
Background
Protecting human health and the environment is the core mission of EPA. EPA's
Strategic Plan lists protecting "healthy communities and ecosystems" as one of EPA's
five major goals. Two environmental statutes administered by the Agency (the Toxic
Substances Control Act and the Federal Insecticide, Fungicide, and Rodenticide Act)
mandate the assessment of benefits. Cost-benefit analysis is also required by Executive
Order 12866 for economically significant regulations. EPA is required, by good
management practice and by federal law, to assess the effectiveness of its programs. In
addition, effective environmental protection requires communication about the value of
ecological protection decisions at the regional and national levels.
The Science Advisory Board (SAB) and the Advisory Council on Clean Air
Compliance Analysis (Council) have recognized the need for science-based approaches
for valuing ecological protection and have undertaken many projects in recent years. In
some projects, advisory committees have advised the Agency in reporting on the
environment. Other projects have focused on methods for identifying critical ecosystems
at the regional level. Some have focused on benefit assessment issues in particular
programs, such as the Clean Air Act programs or the Superfund program.
One SAB project, initiated in October 2003, led to the establishment of the C-
VPESS. This multi-disciplinary committee was charged with conducting a broad
assessment of Agency needs and the state of the art and science of valuing protection of
ecological systems and services and identifying key areas for improving knowledge,
methodologies, practice, and research. In addition to providing advice for the Agency on
its draft Ecological Benefits Assessment Strategic Plan, the C-VPESS has also planned
reports to help strengthen EPA's approaches for valuing the protection of ecological
systems and services, use of such information by decision makers, and the key research
areas needed to strengthen the science base.
3
-------
Intended Audience for the Workshop
There were multiple audiences planned for the workshop: the members of the
SAB, Council, and CASAC; EPA managers concerned with decision-making affecting
ecological resources and documenting ecological benefits; EPA ecological scientists and
risk assessors who support those decisions; EPA social and behavioral scientists
supporting those decisions; EPA economists responsible for regulatory impact analyses
and other economic analyses supporting ecological protection; EPA regional staff
concerned with demonstrating the benefits of protecting and restoring specific
ecosystems and ecological resources; and scientists in other federal agencies concerned
with characterizing ecological benefits.
4
-------
2. Introductory Remarks at the Workshop
Dr. M. Granger Morgan, Chair of the SAB, welcomed meeting participants and
expressed his appreciation for the involvement of a wide range of experts from different
SAB committees, the Clean Air Scientific Advisory Committee, and the Council, as well
as invited experts from universities, consulting firms, EPA and other federal agencies.
He recognized the efforts of the C-VPESS in preparing materials for the workshop and
expressed his hopes for a lively intellectual exchange on the challenging topic of
ecological valuation.
Dr. Dr. Barton H. (Buzz) Thompson, Jr., Chair, SAB C-VPESS, then briefly
discussed the mission of C-VPESS and the goals of the workshop. He noted that the
overall C-VPESS charge was "to assess Agency needs and the state of the art and science
of valuing protection of ecological systems and services, and then to identify key areas
for improving knowledge, methodologies, practice, and research." The workshop was
designed as a peer involvement workshop to give committee members an opportunity to
present some initial findings and conclusions and receive feedback from the Agency and
other participants.
He described the format, which included presentations and question and answer
periods in plenary sessions, breakout groups, and panel discussions.
5
-------
3. Keynote Presentation - Global View from the Perspective of the Millennium
Ecosystem Assessment
Dr. Walter Reid, Former Director of the Millennium Ecosystem Assessment, gave
the keynote presentation at the workshop. In a slide presentation (attached below), he
provided a global perspective on valuing the protection of ecosystems and their services,
based on his work and that of over 1,300 scientists from 95 countries in the study, the
largest assessment ever undertaken of the health of ecosystems. He noted that the
Millennium Ecosystem Assessment project encountered many of the same issues and
reached many of the same conclusions about valuation approaches as the SAB C-VPESS.
Both share the goals of bringing the findings of science to bear on the needs of decision-
makers
Dr. Reid noted that the assessment focused on the linkages between ecosystems
and human well-being and, in particular, on "ecosystem services." The Millennium
Ecosystem Assessment dealt with the full range of ecosystems, from those relatively
undisturbed to ecosystems intensively managed and modified by humans, such as
agricultural land and urban areas. He defined ecosystem services as the benefits people
obtain from ecosystems, which include: provisioning services such as food, water,
timber, and fiber); regulating services that affect climate, floods, disease, wastes and
spiritual benefits; cultural services that provide recreational, aesthetic and spiritual
benefits; and supporting services such as soil formation, photosynthesis and nutrient
cycling.
The four main findings reached were that:
humans have radically altered ecosystems in the last 50 years
changes have brought gains but at growing costs that threaten achievement of
development goals
degradation of ecosystems could grow worse but can be reversed, and
workable solutions will require significant changes in policy.
He noted a need to incorporate nonmarket values of ecosystems in resource
management and investment decisions. The Millennium Ecosystem Assessment focused
on utilitarian values but recognized that considerations of intrinsic value also influence
the actions people affecting ecosystems. Analytical challenges are formidable, even in
focusing on utilitarian approaches to valuation because many ecosystem services are not,
and many cannot be, internalized in markets. In addition, many trade-offs associated
with ecosystem services are expressed in areas remote from the site of ecological
degradation. Economic values of non-marketed services are often substantial but rarely
included in management decisions. As a result, public goods are being excessively
degraded.
From the perspective of the Millennium Ecosystem Assessment, decisions can be
enhanced if they are informed by more complete information about economic and non-
economic values. Economic valuation can be most useful to policy in the context of
comparing alternative options. Economic valuation can be used to enhance
6
-------
understanding of the importance of ecological services, to provide a basis for payments
for ecosystem services and for establishing markets, and for national accounting. Dr.
Reid also noted that the Millennium Assessment considered spiritual and cultural values
of ecosystems as important as other services for many local communities. Deliberative
decision making processes provide a mechanism enabling the articulation of these
different types of value consideration.
In the short question-and-answer session that followed, Dr. Reid noted that the
Millennium Ecosystem Assessment drew on existing peer-reviewed studies. The
Assessment conducted both global and sub-global studies. He also noted that the
assessment focused on sendees with a biological nexus.
^ Millennium Ecosystem Assessment
A Global Perspective on Valuing the Protection of
Ecosystems and their Services:
The Millennium Ecosystem Assessment
Walter Reid
Consulting Professor, Institute for the Environment, Stanford University
Former Director, Millennium Ecosystem Assessment
wwwjikllkimkimaisossinunt.org Strengthening Capacity to Manage Ecosystems SustainaMy for H
What is the Millennium Ecosystem
Assessment?
¦ Largest assessment ever undertaken of the
health of ecosystems
Prepared by 1360 experts from 95 countries;
extensive peer review
¦ Consensus of the world's scientists
¦ Designed to meet needs of decision-
makers among government, business, civil
society
Information requested through 4 international
conventions
Synthesis Reports
Board Statement
MA Conceptual
Framework
Technical Assessment Volumes
Millennium
Assessment
(Pages end
to end)
Eiffel
Tower
7
-------
Science Assessment
A social process designed to bring the findings of
science to bear on the needs of decision-makers
Assessment
Decision-makers
¦ Governments
¦ Private Sector
¦Civil Society
¦ Individuals
Monitoring
Science
A scientific assessment applies the judgment of
experts to existing knowledge to provide
scientifically credible answers to policy relevant
questions
Core Questions
1. What is the rate and scale of ecosystem change?
2. What are the consequences of ecosystem change for human-well
being?
3. How might ecosystems and their services change over the next 50
years?
4. What options exist to conserve ecosystems and enhance their
contributions to human well-being?
What was unique?
Ecosystem services
Provisioning
Gool
^y\
fit *
Regulating
Benefit
Cultural
Supporting
for >ther
§s
E
1
1
V
UJ
c
i
c
I
1
Provisioning Services
Goods produced or provided by ecosystems
Food
¦ Crops
¦ Livestock
¦ Capture Fisheries
¦ Aquaculture
- Wild Foods
Fiber
¦ Timber
¦ Cotton, hemp, silk
¦ Wood Fuel
Genetic resources
Biochemicals
Freshwater
Regulating Services
Benefits obtained from regulation of ecosystem
processes
Air Quality Regulation
Climate Regulation
¦ Global (C02 sequestration)
¦ Regional and local
Erosion regulation
Water purification
Disease regulation
Pest regulation
Pollination
Natural Hazard regulation
Cultural Services
Non-material benefits obtained from ecosystems
Spiritual and Religious Values
Knowledge Systems
Educational values
Inspiration
Aesthetic Values
Social Relations
Sense of Place
Recreation and Ecotourism
8
-------
What was unique?
Consequences for People
Supporting
¦ Nutrient
Cycling
¦ Soil Formation
¦ Primary
Production
Provisioning
¦ Food
¦ Water
¦ Fiber
Regulating
¦ Climate regulation
¦ Disease regulation
¦ Water purification
Cultural
¦ Spiritual
¦ Religious
¦ Aesthetic
Ecosystem Services
Life on Earth: Biodiversity
*
Security
¦ Personal safety
¦ Secure from
Material
¦ Livelihoods
¦ Food
¦ Shelter
Freedom of
Choice and
Action
Opportunity to be
able to achieve
individual values
doing and being
Health
¦ Strength
¦ Feeling well
¦ Clean air and water
Social Relations
¦ Social cohesion
¦ Mutual respect
¦ Ability to help others
Constituents of Well-Being
MA Conceptual Framework
Human Well-being and
Poverty Reduction
Basic material for a good life
Health
Good Social Relations
Security
Freedom of choice and action
5
Indirect Drivers of Change
Demographic
Economic (globalization, trade,
market arid policy framework)
Sociopolitical (governance and
institutbnal framework)
Science and Technology
Cultural and Religious
J-4
Life on Earth:
Biodiversity
Direct Drivers of Change
Changes in land use
Species introduction or removal
Technology adaptation and use
External inputs (e.g., irrigathn)
Resource consumption
Climate change
Natural physical and biological
drivers (e.g., volcanoes)
What was unique?
Multi-Scale Assessment
Main Findings
fcCMMCf! lASfcHI
rjp -IP*- ¦ .
Eopto. Fi
Iwik
CUMteci
¦£3e£"* T ^
1* % Iffc"
t
1. Humans have radically altered
ecosystems in last 50 years.
mAhci A tomK n
T igpfetf F«*W Utrpm
Tfj#. pw«fiy. VA nmtfiXWt Ml *1 CM», CWlt HMtpie*'. Wm«, Serf- AJnet. ».* ttfl Nim
$52 trillion in 2003 [
$10 trillion in 1967
$1 trillion in 1900 J
9
-------
Teragrams of Nitrogen per Year
Natural Sources
Total Human
Additions
Fertilizer
Agroecosystems
Fossil Fuels
Pre-peak
Harvest peak
Post-peak
Year of Peak Fish Harvest
Main Findings
1. Humans have radically altered
ecosystems in last 50 years.
2. Changes have brought gains but
at growing costs that threaten
achievement of development
goals.
¦ Degradation of many ecosystem
services
¦ increased risk of abrupt changes in
ecosystems
¦ Growing harm to poor people
ar*1
The Balance Sheet
Change in benefits over last 50 years
Enhanced Degraded
Crops
Livestock
Aquaculture
Carbon sequestration
Capture fisheries
Wild foods
Wood fuel
Genetic resources
Biochemicals
Fresh Water
Air quality regulation
Regional & local climate
regulation
Erosion regulation
Water purification
Pest regulation
Pollination
Natural Hazard
regulation
Spiritual & religious
Aesthetic values
Timber
Fiber
Water regulation
Disease regulation
Recreation & ecotourism
Bottom Line: 60% of Ecosystem
Services are Degraded
10
-------
Increased likelihood of abrupt changes
(established but incomplete evidence)
Fisheries collapse
Eutrophication
Coral reef regime shifts metB
Disease emergence
j
Species introductions »«¦
A
Regional climate change
J \ ~
..
MM
Atlantic Cod off Newfoundland
24
impact on Poor People
Critical concern - drylands
40% of land surface and more
than 2 billion inhabitants
Lowest levels of human well-
being
10-20% of drylands degraded
Only 8% of renewable water
supply
Highest rate of population
growth
Main Findings
1. Humans have radically altered
ecosystems in last 50 years.
2. Changes have brought gains but
at growing costs that threaten
achievement of development
goals.
3. Degradation of ecosystems
could grow worse but can be
reversed.
MA Scenarios
World Development
Globalization
Global Orchestration
TechnoGarden
Regionalization
*"<3B
Order from Strength
Adapting Mosaic
Mediterranean Forests
Temperate Grasslands &
Woodlands
Temperate Broad leaf Forest
Tropical Dry Forest
Tropical Grasslands
Tropical Coniferous Forest
Tropical Moist Forest
Habitat Loss to 3890
under MA Scenarios
Some services improved in one or more
of the MA scenarios
Examples:
Freshwater
Water regulation
Erosion control
Water purification
Storm protection
Aesthetic values
Recreation
Percent of habitat (biome) remaining
11
-------
Main Findings
1. Humans have radically altered
ecosystems in last 50 years.
2. Changes have brought gains but
at growing costs that threaten
achievement of development
goals.
3. Degradation of ecosystems
could grow worse but can be
reversed.
4. Workable solutions will require
significant changes in policy
Responses
Economics and Incentives
Institutions and Governance
¦ Planning and Management
Technologies
¦ Greater efficiency of resource use
¦ Less harm to other services
Social and Behavioral
Knowledge
Incorporation of nonmarket values of ecosystems in
resource management and investment decisions.
MA perspectives on valuation
MA focused on utilitarian values but recognized that
the actions people take that influence ecosystems
result also from considerations of intrinsic value.
Many of the values associated with ecosystems are
not (and many can not be) internalized in markets -
problem of public goods. This is leading to
excessive degradation.
Services are treated as 'free' and limitless
Significant externalities
Trade-off of provisioning services against others
Trade-offs Among Services
Provisioning
Regulating
Cultural
Crops
Livestock
Aquaculture
Carbon sequestration
Degraded
Capture fisheries
Wild foods
Wood fuel
Genetic resources
Biochemicals
Fresh Water
Timber
Fiber
Water regulation
Disease regulation
Recreation & ecotourism
Air quality regulation
Regional & local climate
regulation
Erosion regulation
Water purification
Pest regulation
Pollination
Natural Hazard
regulation
Spiritual & religious
Aesthetic values
Provisioning services are being enhanced at the cost
of regulating & cultural services
MA perspectives on valuation
Economic values of non-marketed services are
often substantial but rarely included in management
decisions
¦ Annual recreation value of Hawaii marine protected areas:
$300,000 to $35 million ea.
¦ Increased income from forest-based pollinators on one coffee
farm in Costa Rica: $60,000/yr
¦ Value of marsh in Sri Lanka for flood control: $1,750/ha
¦ Benefits per household of preserving neighboring open space
in Maryland: $1,000 -3,300/ha
MA perspectives on valuation
4. Economic (and health costs) of degradation of
services can be substantial
¦ Cost of damage of UK agriculture to other ecosystem services:
$2.6 billion (10% of farm receipts)
¦ Introduction of Zebra mussels into aquatic ecosystems in the
US: $100 million annual costs to power industry
¦ Annual cost to fisheries of mangrove deforestation in
Campeche Mexico: $279,000
12
-------
Significant scope for greater estimation
of economic value
Spiritual & religious
Aesthetic
Genetic Resources
Disease regulation
Flood/Fire regulation
Climate regulation
Freshwater
Water purification
Recreation & tourism
Fiber
Food
Economic
Valuation
Difficult or
impossible
Established
methods
Economic Value ($)
Economic value of non-marketed
services can be high
Forests in Italy Forests in Croatia
[
¦
Economic Value^gqowrhEd^Hfej^ ($ per hectare)
Marketed and non-marketed values
Value of timber and
fuelwood less than 1/3
of total economic value
of forests in 5 of 8
Mediterranean
countries
Source. M*ennium Ecosystem Assessment
|t«m«
1
§s
E
1
O
V
UJ
c
i
c
I
1
MA perspectives on valuation
5. Decisions can be enhanced if they are informed by
more complete information on economic and non-
economic values.
Greatest policy relevance of valuation is in the
context of comparing alternative policy or
management options
Trade-offs among ecosystem services
Mangrove Services:
nursery and adult
fishery habitat
fuelwood & timber
carbon sequestration
traps sediment
detoxifies pollutants
protection from
erosion & disaster
1
Mangrove ecosystem
cropsQ
PtJ bttte HteSt FFtBae8Ttt\X& Lies {par
hectare 1989
|ve: $91000 to $3,600
¦m: $2^000 to $200
$4000
Va ue
(per hectare
CesaiBHDsiidies (-$1,700)
Pollution Costs (-$230)
imber products (
Restoration (-$8,240)
Mangrove
13
-------
Economic case for conservation
The total economic
value associated
with managing
ecosystems more
sustain ably is often
higher than the
value associated
with conversion
iirrro
J
WetuiMl Tropical Foml
MA perspectives on valuation
Economic valuation uses:
Choices among different management or policy
options
Public understanding
Basis of Payments for Ecosystem Services
Basis for establishing markets (if value can be
captured).
National accounting
National Accounting
Net National Savings
in 2001 Adjusted for
changes in Human
and Natural Capital
For countries shown here,
resource depletion
(minerals, oil, forests) and
damage from C02
emissions accounted for
10-25% decline in savings
DECLINE IN WEALTH GROWTH IN WEALTH
-*OK-30% -20«i-1Wfc 0 Ift as m
¦ MangiiipraiiitlfiMn
¦ A4ua*0n«M«*0> mptfcantoiaM «*< mmj, rtidkr
MmM <* »w««n cjp*> |» a and lp|Mi KMMIt
acKOttttg tfttuiUri^ferMlty ar*gy**.CO,pc*rttt)
Venezuela
Mauritania
ayiran
Ecuador
Indonesia
EiTwptt
Buindi
Malays*
Viet ten
Mongolia
Bolivia
MA perspectives on valuation
7. Serious shortcomings in availability of economic
information
Most studies focus on only one or two services and
most studies are not in peer reviewed literature
Major gap:
Landscape scale studies providing valuation of multiple
ecosystem services (full 'bundle'of services)
Availability of ecological production functions often
is the limiting factor
MA perspectives on valuation
MA perspectives on valuation
8. Spiritual and cultural values of ecosystems are as
important as other services for many local
communities
9, Need for deliberative decision-making processes
¦ Additional information on economic values of ecosystem
services could improve decision-making
¦ But, not all ecosystem services that matter to people can be
valued in economic terms (esp. cultural services and
considerations of intrinsic value)
¦ And, different stakeholders will place different weights on
different attributes of ecosystems
¦ Deliberative decision-making processes provide a mechanism
enabling the articulation of these different types of value
considerations.
14
-------
MA perspectives on valuation
"lO.Scale dependence and stakeholder dependence of
value considerations
Carbon sequestration Spiritual values
Bioprospecting Food
Genetic resources Water
48
Findings and data:
MAweb.org & Island Press
Publications
Synthesis Reports
¦ Synthesis
¦ Board Statement
¦ Biodiversity Synthesis
¦ Wetlands Synthesis
¦ Health Synthesis
¦ Desertification Synthesis
¦ Business Synthesis
Technical Volumes and MA Conceptual
Framework (Island Press)
¦ Ecosystems and Human Well-being:
A Framework for Assessment
¦ State and Trends
¦ Scenarios
¦ Multi-Scale Assessments
¦ Responses
Financial and in-kind support
(full list available at www.MAweb.org)
Global Environment Facility
Environmental Management Authority of Trinidad and
United Nations Foundation
Tobago
David and Lucile Packard Foundation
Ford Foundation
World Bank
Government of India
Consultative Group on International Agricultural
International Council for Science
Research
International Development Research Centre
United Nations Environment Programme
Island Resources Foundation
Government of China
Japan Ministry of Environment
Government of Norway
Laguna Lake Development Authority
Kingdom of Saudi Arabia
Philippine Department of Environment and Natural
Swedish International Biodiversity Programme
Resources
Asia Pacific Network for Global Change Research
Rockefeller Foundation
Association of Caribbean States
U.N. Educational, Scientific and Cultural Organization;
British High Commission, Trinidad & Tobago
UNEP Division of Early Warning and Assessment
Caixa Geral de Depositos, Portugal
United Kingdom Department for Environment, Food and
Canadian International Development Agency
Rural Affairs
Christensen Fund
United States National Aeronautic and Space
Administration
Universidade de Coimbra, Portugal
50
Cropper Foundation
15
-------
4. Introduction to C-VPESS Work on an "Expanded and Integrated Approach" for
Valuing Ecological Protection
Dr. Kathleen Segerson, C-VPESS Vice Chair, gave a short overview presentation
of the conceptual approach developed by the committee (presentation slides found at the
conclusion of this section of the workshop report). She described the committee's goal:
to assess EPA's needs and the state of the science used for valuing ecological protection
and to identify key areas for improvement. The committee to date has focused on EPA's
needs in three areas: national rulemaking, regional decision-making, and program
evaluation and assessment.
A key issue for the committee was the use of terms "value" and "valuation."
"Value" is understood by the committee as a broad concept including both instrumental
and non-anthropocentric values and thus includes utility-based, moral, religious, and
spiritual values. The committee's approach was to recognize the many possible sources
of value and seek methods for characterizing or measuring them. "Valuation" methods
therefore include methodologies, based on theory and data, for characterizing and
measuring different kinds of value. The methods can differ in their focus and in some
cases in their underlying premises.
She summarized the key features of an integrated and expanded approach as: a
focus on impacts of most concern to people; an integration of ecological analysis and
valuation; and the use of an expanded set of valuation approaches. To implement such an
approach, she described a proposed process for valuing ecological systems and services
under consideration by the committee (See Figure 4-1).
Figure 4-1
Valuation
and Preference
s. Methods
Actions
1 B. Identify what
matters to people
1 A. Identify possible
ecological effects
2. Identify ecological
effects that matter
3. Characterize/ quantify
ecological effects
4. Characterize/ quantify
human consequences of
ecological effects
5A. Estimate value of effects
in non-monetary terms
5B. Estimate monetary
value of effects
LEGEND
T ools
A Proposed Process for Valuing Ecological
Systems and Services:
Ecological
Models
6. Communicate
results to public
and decision makers
16
-------
The approach would use ecological models and valuation and preference methods
to identify and then characterize and measure the ecological effects that matter in ways
appropriate to each type of decision context faced by EPA. The preliminary conclusions
and recommendations emerging from the committee encourage EPA to move toward an
expanded range of important ecological effects and human considerations by:
recognizing many types and sources of value;
thinking about ecosystem "services" and mapping from ecological
endpoints to services;
expanding the range of services;
focusing on the services that are most likely to be important to people
exploring the use of different methods;
ensuring interdisciplinary collaboration throughout the process throughout
the process; and
soliciting public input early in the process.
In the question and answer period that followed the presentation, a participant
asked whether the committee viewed the value of ecological systems as the sum of the
ecological services they provide. Dr. Segerson responded that for some committee
members, the term "ecological services" was sufficiently broad to encompass spiritual
values and non-utilitarian values. For other members, the term "services" did not capture
all the different types of values associated with ecological systems and their components.
Workshop participants also emphasized the importance of providing advice to improve
estimates of uncertainty and providing a way to address ecological effects from a long-
term perspective, even if such a view increases uncertainties in ecological modeling and
valuation, as the values people hold change over time.
One participant raised a question about the appropriate spatial scale for analysis.
Dr. Segerson noted that ecological effects can be viewed very differently at the national
scale, as opposed to local scale. She noted that the committee was focusing on a variety
of decision contexts and not solely on national-level values.
Another participant asked how the committee advice could help EPA work within
its limited constraints of time and funding for analysis. Dr. Segerson responded that the
committee was concerned with providing practical advice and how to qualify benefit
transfer information so it can be used in the best ways. It was also considering
alternatives to traditional economic methods, to see if other approaches might also offer
useful information for decision makers.
17
-------
An Expanded and Integrated
Approach for Valuation at EPA
U. S. Environmental Protection Agency
Science Advisory Board (SAB)
Workshop
Goal of CVPESS Project:
» assess EPA needs and state of the art and
science in valuing the protection of ecological
systems and services
» identify key areas for improving knowledge,
methodologies, practice, and related research at
EPA.
» Disciplinary composition of C-VPESS: decision
science, ecology, economics, engineering,
philosophy, and psychology
» Key focus: the need for an expanded and
integrated approach to valuing EPA efforts to
protect ecological systems and services.
EPA's Mission Regarding
Ecosystem Protection
> EPA's mission is to "protect human health and the
environment."
> EPA has historically focused decision making and much
of its expertise on human health from environmental
stressors, with relatively little attention given to
ecosystems effects not directly linked to human health.
> "Healthy Communities and Ecosystems" - One of the
Five Goals in EPA's Strategic Plan
> Basic premise of C-VPESS work: EPA has a mission to
protect ecological systems because of the services they
provide and the collective responsibility to protect the
environment.
Concept of Ecosystem Services
> EPA's ecological risk assessment paradigm focuses on
identification of ecological assessment endpoints, not
ecological services.
> Central theme of C-VPESS: the need to move beyond
consideration of ecological endpoints to the
consideration of "ecosystem services," which reflect the
direct or indirect contributions of ecosystems to human
well-being.
> Ecosystem services can be defined very broadly, but in
the context of valuation, it is important to distinguish
between intermediate services and final services to
avoid double counting.
Concept of Value
> The term "value" means different things to different
people, and has different meanings within different
disciplines.
> A fundamental distinction exists between instrumental
values (means) and intrinsic values (ends).
> While all values are anthropogenic, there is
disagreement over whether all values are
anthropocentric.
> C-VPESS recognizes that there are many possible
sources of value derived from ecosystems and the
services they provide, and seeks methods for
characterizing these various sources of value.
Use of Terms
> "Value" is used broadly to include values that stem from
contributions to human well-being as well as values that
reflect other considerations, such as social and civil norms
(including rights) and moral, religious, and spiritual beliefs and
commitments.
> "Benefit" is used to refer more narrowly to the contribution of
ecosystems and their services to human well-being.
> 'Valuation" is used to refer to the process of characterizing,
estimating, or measuring either the value of, or the value of a
change in, an ecosystem, its components, or the services it
provides. Includes both monetary and non-monetary
valuation.
> There are a number of methods that can be used for
valuation, which differ in their focus and, in some cases, in
their underlying premises.
> C-VPESS is exploring a range of possible valuation methods.
18
-------
Major challenges in Ecological Valuation
of EPA policies or programs
» understanding the many sources of value that ecosystems
generate
» predicting the ecological effects of alternative EPA actions
» linking those effects to changes in the dimensions of
ecosystems or the service flows that people value
» developing methods that can be used to characterize and/or
measure the value of protecting ecological systems and
services
» aggregating to a national level using local or regional
studies
» finding measures or means of representing ecological
values that are commensurable with values of non-
ecological changes, such as human health
Ecological Valuation at EPA
Valuation Contexts:
> National rule making
> Regional programs and decisions
> Program evaluation and assessment (GPRA)
> The specific information needs and institutional
constraints differ in these different contexts.
> Ecological valuation at EPA must be conducted within
a set of institutional, legal, organizational, and practical
constraints.
Observations regarding the current state
of ecological valuation at EPA
Observations from interviews with EPA staff (focusing on
rule making):
> Ecological valuation practices vary considerably across
program offices, reflecting differences in mission, in-
house expertise, etc.
> The timing of the process largely determines the kinds of
analytical techniques that are employed (e.g., court-
imposed deadlines and need forOMB review).
> There is a tendency to use methods that have passed
review in the past and a disincentive to explore new or
innovative approaches.
> Extent and nature of the integration between social
scientists and biophysical scientists at EPA is unclear.
Observations regarding the current state of ecological
valuation at EPA (continued)
Observations from an illustrative example:
New CAFO regulations
> Focus on a limited set of environmental benefits, driven primarily by
the ability to monetize these benefits using generally accepted
models and existing value measures (due to time and resource
constraints).
> Sole focus on the use of economic valuation methods.
> Mention of non-monetized benefits, but no attempt to characterize or
quantify them in any way.
> Use of highly leveraged benefit transfers.
> Little attempt to model systematically and in detail rule's ecological
impact early in the assessment.
> Little, if any, consultation with public, especially early in process, to
help identify effects that are likely to be most important to them.
> Limited use of peer review, especially early in the process, to
provide feedback and advice.
An Integrated and Expanded Approach
Components of Valuation: A General Framework
law* prwlpies. p^,
mandates and - rwi.taJuH
putac «tnc*rn» Decision-making
Problem Formulation communicate
Specify attematfcw ,¦ ~ 1
actwn* RflttwsVV \
Ecological Systems . Values
Ecotogcal analyses EvsJuaMns sk«i
proauetwn turvmons * _ . arKj economic effects
Ecosystem services
J
Key features of an integrated and expanded
approach:
1 Focus on impacts of most concern to people.
Key issues:
> Recognition of the many possible sources of value
from ecosystem protection
> Need for information about values early in the process
> Requires an expansion of the types of services to be
characterized, quantified or valued
19
-------
Key features of an integrated and expanded
approach (continued):
2. Integrate ecological analysis and valuation.
Key issues:
> Need for collaboration throughout process, beginning at
early stages.
> Identification of relevant ecosystem services and
mapping of effects on ecological assessment endpoints
to effects on services
> Design of ecological analysis so that outputs provide
usable inputs for value assessments
> Design of valuation techniques that address important
ecological/biophysical considerations
Key question:
How to implement such an approach?
Preliminary Conclusions and
Recommendations
The Committee encourages EPA to move towards covering
an expanded range of important ecological effects and
human considerations by:
> Recognizing many sources of value, both instrumental and
intrinsic
> Thinking about instrumental values in terms of ecosystem
"services" and mapping from ecological assessment endpoints to
services (using concept of an ecological production function)
> Expanding the range of services to which valuation is applied,
focusing on the services that are most likely to be important to
people
Key features of an integrated and expanded approach
(continued):
3. Use of an expanded set of valuation approaches.
Key issues:
> Recognition that different approaches provide different
ways of characterizing or providing information about
values
> Different approaches could be used at different stages of
valuation process (e.g., providing information to guide
focus of study vs. characterizing benefits specific to the
EPA action)
> Approaches to be used could vary with the specific
policy context, reflecting differences in information
needs, the underlying sources of value, data availability,
and methodological limitations.
A Proposed Process for Valuing Ecological
Systems and Services:
Preliminary Conclusions and
Recommendations (continued):
> Exploring and expanding the use of different methods for
characterizing or measuring both intrinsic and
instrumental values
> Involving, from the beginning of the process, an
interdisciplinary collaboration among of
physical/biological and social scientists
> Soliciting, from the beginning of the process, input from
the public or representatives of individuals affected by
ecological changes
20
-------
5. Panel Discussion with EPA Senior Managers
A member of the C-VPESS, Dr. James Boyd, introduced the panel of five senior
Agency managers who had been asked to respond to three questions:
The C-VPESS observes that EPA has only been able to conduct valuations for
a narrow range of ecological effects, compared to the wide range of
ecosystems and ecological resources affected by EPA decisions and
polices. In your own experience, have EPA decisions and programs been
affected by limits on the ability to appropriately measure and value
ecological effects?
C-VPESS recommends that EPA expand valuation efforts to reflect more
different types of values than the commercial and recreational values
captured by most kinds of traditional economic analyses. The Committee
advises EPA to explore supplementing economic methods with other
kinds of methods to reflect a fuller range of values. What do you see as
the potential for such efforts? What do you see as the barriers to exploring
these options?
C-VPESS intends to use case examples to illustrate the integration of multiple
methods as a way to describe and measure a broader range of values
related to protection of ecological systems and services. From your
vantage point, what would make such examination of case examples most
useful?
Mr. Robert Brenner from EPA's Office of Air and Radiation began the panel
discussion with a short presentation (slides attached below) and a quote from Albert
Einstein: "Not everything that can be counted counts, and not everything that counts can
be counted." He noted that EPA's air program has succeeded at documenting health
effects for pollutants, but that often they do not motivate local efforts for environmental
protection. He noted that environmental protection near Denver and in the Smoky
Mountains were driven by public desire to see important landmarks. Ecological concerns
drove those decisions, but were not captured in monetized cost-benefit analysis.
He viewed the C-VPESS integrated framework presented by Dr. Segerson as
consistent with the approach being taken by OAR in its Section 812 study of the costs
and benefits of implementing the Clean Air Act. His office is interested in strengthening
economic data and approaches for valuation through benefit-cost or cost-effectiveness
analyses, as well as exploring other assessment methods. In his view, however, serious
impediments to the success of the integrated, expanded approach to ecological valuation
remain. Given EPA's regulatory needs, there is competition for limited analytical
resources. There appears to be ongoing resistance to using currently available
approaches, such as the recent contingent valuation study of natural resource
improvements in the Adirondacks. He concluded by calling for help defining "best
21
-------
practices" for a full range of quantifiable effects, including effects that can only be only
characterized qualitatively. He suggested that the committee look at an upcoming rule
and identify current best practices, so the Agency can take advantage of them.
SAB C-VPESS
EPA Senior Managers
Panel Discussion
Rob Brenner
Director, Office of Policy Analysis and Review
Office of Air and Radiation
December 13, 2005
"Not everything that can
he counted counts, and not
everything that counts can
he counted."
Albert Einstein
Quantified Ecological Effects vs Mortality
2010 CAAA90 Benefits
Billions
~ Low
~ Central
¦ High
Acidification Timber N-Dep Mortality
Tools and Approaches Needed
¦ Some argue we should focus on expanding methods and
data for economic valuation through benefit-cost or cost-
effectiveness analysis
¦ Others argue economic data and methods will never give
full and adequate treatment to important ecological service
flows so other, non-economic paradigms are needed to
characterize the value of ecological effects
¦ OAR interested in both approaches
- Continue research in both ecological sciences and economics to
bridge gaps in economic analyses of ecological effects
- Explore other assessment methods to provide information on
ecological effects currently assigned an implicit value of $0
¦ e.g., "Natural Systems Impact Assessment"
4
22
-------
CVPESS and OAR Approaches
Prospects for Success
Clean Air Act Sec. 812 Studies CVPESS Integrated Framework
Broad assessment of ecologically a. Identify the context and scope of the
important air pollutants benefit assessment
1. In-depth assessment of selected b. Identify the ecological services that
ecological endpoints, esp. will be considered in the assessment
economically significant service flows
c. Characterize, repres ent or m easure
Wide-ranging evaluation of those impacts in biophysical, human,
potentially significant ecological and/or monetary terms
effects at various spatial scales (e.g.,
cellular, individual, population, local
ecosystem, etc)
5
¦ Serious impediments remain
¦ Much competition for limited analytical resources
- RIAs, ElAs, RFAs, Circular A-4 probabilistic analysis
¦ Ongoing resistance to using tools already in hand
- "Most people ... place little confidence in ... CV studies."
- Example: RFF CV study on Adirondacks is out there,
shows big benefits, but we can't use it
¦ Need support for using existing as well as new
tools and techniques
¦ Need help defining "best practice"-
- For full range of quantifiable effects
- For effects we can only characterize qualitatively
Dr. Albert McGartland, Director of EPA's National Center for Environmental
Economics, was the next speaker. He acknowledged that EPA sometimes appears as the
"cancer protection agency," rather than an environmental protection agency because it
more often frames issues in terms of human health, rather than ecological protection . He
noted that analyses supporting a recent mercury decision were framed nearly entirely in
terms of human health effects, while the impacts of the chemical had a broad range of
ecological impacts. He acknowledged the need for increased cooperation across agencies
and across disciplines to improve valuation of ecological endpoints. He also cautioned
against "letting the perfect be the enemy of the good" and noted that his entire extramural
budget would be exhausted by a single high-quality contingent valuation study. He
suggested that the committee focus its attention on issues of marginal benefits related to
specific decisions, rather than total benefits of ecosystems. He pointed out the Agency's
need for cheaper and more efficient ways to conduct valuations. He noted that valuation
of human health benefits was successful because dose-response information could be
linked to epidemiology and to the economic analysis of marginal changes needed by
rulemaking. Without those kinds of links, ecological valuation will always be difficult.
Given those missing links, he asked the committee to focus on benefit transfer or other
kinds of best practices to follow.
Ms. Kathleen Callahan, Deputy Regional Administrator from EPA Region 2,
presented a different perspective. She noted that the regions are intensely interested in
ecological valuation issues because they reflect day-to-day decisions encountered in
implementing national rules and policies and working with state and governments. Staff
in her region face tight deadlines for decision-making. They encounter ecological values
that differ across their region (which includes New York, New Jersey, the Virgin Islands,
and Puerto Rico) and sometimes also differ from national values. Public health concerns
drive many decisions, but ecological health also "plays in." She described a Superfund
site, where the region faced a decision about whether to protect an old stand of trees or
remove them so children would not be attracted to play on contaminated soil. Another
example was Long Island Sound, where the region attempted an ecological valuation
23
-------
study in 1990. Because that study was not peer reviewed and was challenged, the
regional office did not attempt an update needed in 2000. The region could benefit
generally from understanding the value of its ecological resources, especially estuaries,
but the unsuccessful effort in 1990 was the only example where such an effort was tried
in Region 2.
Ecological values also appear as issues when the Agency works with other tribal,
state, and federal partners. In a clean-up decision for Onondaga Lake, contaminated with
mercury and other hazardous waste residues, the Onondaga Indian Nation objected to the
fairly expensive (approximately half-million-dollar) remedy selected by the region. The
tribe objected because the lake was sacred; any remedy that left the lake less
contaminated than when the Onondaga's Peacekeeper sanctified it was unacceptable. A
different issue arose when the region worked with the State of New Jersey and other
federal agencies on the Special Area Management Plan for the Hackensack wetlands. In
discussing aims for the plan, colleagues in other federal agencies raised issues in
landscape ecology. EPA Region 2 had no landscape ecologist and no framework for
factoring such values and science into a decision where the region traditionally relied on
wetland characterization guidance and information about property values.
Ms. Callahan expressed a hope for an "iterative process" that would gradually
improve the information about ecological values supporting regional decisions. She also
called for a clearer understanding of how to conduct public dialogue about ecological
values in a manner consistent with all partners' roles in governance and how to translate
that dialogue into effective planning for environmental protection.
Dr. Michael Shapiro, Principal Deputy Administrator of EPA's Office of Water,
began his comments by acknowledging that his office has a "very high stake" in
ecological values. Within his office, policies and rules under the Safe Drinking Water
Act are driven by public health concerns, but ecological values are the driving concern
for the Clean Water Act.
He echoed his colleagues' views that information about ecological values play
into Agency decisions at different levels in different ways. At the national level, lack of
valuation information affects strategy and priority setting. In the Agency's five-year
planning cycle, the Agency determines investments in programs and rulemaking efforts
that offer the greatest benefits. When ecological programs are unable to project
ecological benefits, they can get short-changed in the planning and budgeting process.
Dr. Shapiro noted the value of teamwork and a holistic approach described in the
C-VPESS integrated and expanded approach, but emphasized that for the most part, at
the national level, dollar values are critical when hard, difficult decisions on major rules
must be made. As much as the committee may wish for the Agency to adopt an analysis
that factors multiple concerns through multiple metrics, dollar values for benefits are
important, because most analyses quickly are dominated by monetized metrics. Given
the Agency's time constraints, it needs to rely on benefit transfer and has only limited
opportunities to conduct studies to "add to the bookshelf." If there is no agreement on
24
-------
contingent valuation, then key tools for analysis are removed.
Dr. Shapiro observed that EPA has completed many technology-based
rulemakings under the Clean Water Act and that regulatory efforts were shifting to the
regions for implementation. Regional and local-level decisions offer more significant
opportunities for long-term studies of particular ecosystems and extended dialogue on
ecological values. Such a context can provide more opportunities for a holistic approach
that can address aesthetic and spiritual values that cannot be easily aggregated or
analyzed at a national level.
Dr. George Gray, Assistant Administrator for EPA's Office of Research and
Development, began his remarks with recognition of the challenge before the committee,
as well as appreciation of the difficulties as well as the challenges of working across
disciplines. He noted that the SAB plays a useful role in motivating the Agency to
innovate, to try new methods, and to overcome the inertia associated with reliance
methods that have "passed review" in the past.
He saw merit in the expanded and integrated approach proposed by the
committee. In his view, it offered a challenge to biophysical sciences and even to human
health metrics and not just to economics. He then briefly reviewed the role of the Office
of Research and Development in ecological valuation. He noted that his office has
supported Agency efforts in ecological risk assessment. As part of the Agency's
extramural grant program, Science to Achieve Results (STAR) program, the Office of
Research and Development has focused a part of its Economics and Decision Science
efforts on ecological valuation. That program has called for inter-disciplinary
collaboration, and has encouraged studies at different geographic scales; studies using
different methodologies; and studies examining the appropriate use of benefit transfer.
He encouraged the SAB Committee to identify both long- and short-term research
priorities that ORD could consider for funding. He also suggested that the committee
explore case studies at different scales (e.g., local, regional, and national) and in the
process take advantage of the ecological case studies completed by an inter-disciplinary
research team in the Office of Research and Development.
The panel then took questions from the audience. The first question
acknowledged Dr. Shapiro's comment that increasingly benefits that cannot be monetized
aren't considered by decision-makers. The questioner asked about the "pressures" and
reasons behind this development, which appears in the environmental arena more
prominently than in the defense, security, and public health areas. Mr. Brenner
responded that over time decision-makers are asked to consider many different factors in
rulemaking, such as small business impacts and other effects. Monetizing benefits
reduces the complexity for decision makers. Non-monetized effects are listed, but until
there is a framework for addressing them, they are unlikely to be explicitly addressed in
rule making. Dr. McGartland made a different point. He noted that despite the inability
of the Agency to characterize ecological benefits fully, all rules from the Office of Water
that were finalized in the last three to five years had monetized costs that exceeded
25
-------
benefits. He observed that the Agency should be more systemic in its analysis of non-
monetized benefits and that decision makers could better trained in the use of such
analysis.
Ms. Callahan observed that reliance on quantifying values is a "trap our society
falls into." In the face of competing values, there is no good process for dialogue and
decision-making. Decision-makers easily favor quantitative justifications for their
decisions. In a litigious society, quantitative evidence overwhelms qualitative evidence,
unless qualitative evidence can be communicated in an extraordinary way.
Another question concerned whether issues similar to eco-valuation arise when
human health values are monetized. Mr. Brenner responded that at times it is difficult to
know if human health values are monetized fully. In the acid rain program, for example,
EPA is finding human health benefits previously unsuspected. Dr. Shapiro responded
that the public generally shared an intuitive sense of the metric of live saved or health
events averted, but lacked such consensus in the ecological arena. Dr. Gray noted that
EPA has a history of decision-making related to un-quantified, un-monetized health
effects, where non-cancer health events are at issue and data principally involve reference
doses or reference concentrations related to hazardous effects. In his view, such history
shows a willingness to take different approaches.
Mr. James Laity, a member of the workshop's expert panel on December 14,
2006, spoke from the audience and identified himself as an examiner of Office of Water
rulemakings at the Office of Management and Budget. He characterized the non-
monetized benefits he reviewed as generally presented in the format of a "laundry list"
where it is difficult to distinguish important effects from those less important. He
encouraged the Science Advisory Board to stimulate work on meaningful, objective,
possibly quantified ways to evaluate non-monetized benefits.
A final questioner asked about the merits of a tiered approach to analysis of
ecosystem services and, in a separate question, about the possibility of an ecological
equivalent to the "statistical value of a human life" used in the 812 Study. Mr. Brenner
agreed that the Agency did need a less resource-intensive way to identify major benefits.
Other panelists were intrigued by the notion of a "statistical value of an ecosystem" but
couldn't envision what that would be.
26
-------
6. Overview of Methods Being Considered by C-VPESS
Dr. Gregory Biddinger introduced this joint presentation with Drs. Terry Daniel
and Stephen Polasky to provide an overview of the total suite of methods related to
valuation being considered by the C-VPESS (see presentation slides at the end of this
section of the workshop report). He noted that the workshop agenda planned for
workshop participants to focus on five selected sets of approaches from that suite of
methods in the afternoon breakout sessions. In the short presentation, Dr. Biddinger
summarized eco-centric approaches; Dr. Terry Daniel summarized socio-psychological
methods; and Dr. Stephen Polasky summarized methods designed to obtain group
determination of values and economic methods. They each related methods to the
process diagram introduced by Dr. Segerson (see figure 4.1) and provided some brief
detail about several of the methods discussed.
Workshop participants then had the opportunity for several questions. One
participant asked whether the committee had considered research on the information
needs of decision-makers and legislators. He asked whether there was a need for tools to
help decision makers with their own value clarification and tools to help them
communicate to the public and with each other. Dr. Daniel responded that analysis of
decision-making in Congress was outside the scope of the committee's charge. Other
committee members seconded the view that the committee was focusing on public
values, not the values of leaders, because a focus on the latter would add to the
complexity of analysis and raise issues of amplification feedback.
Another workshop participant asked about issues of feasibility in selection and
use of methods and how the Agency could be advised to use its resources to do a better, if
not perfect, job of valuation. He asked whether the "80/20 percent" rule could apply to
help the Agency gain practical benefits. The panel noted that such an approach was
important.
A participant questioned how the use of surveys or "polls," described by Dr.
Daniel as part of the "socio-psychological methods" might be used to measure values
discussed in the morning's keynote presentation, which called for increased sustainability
and stewardship of natural resources. The participant expressed concern that results of
polls could be trivial because they would express little understanding of ecological
effects. Dr. Daniel responded that surveys that express public opinion do not make
decisions; they provide one source of information for decision makers. If results of
surveys show a divergence between public views and values and those of experts, then
there may be an education or information lag that decision-makers might choose to
address. Dr. Polasky responded that effective valuation is multi-disciplinary. Valuation
runs a risk if decisions are made on the basis of a single type of valuation study in
isolation. Dr. Biddinger also commented that it was important to find new ways to
communicate ecological values effectively.
A question was then posed about the time preferences associated with public
27
-------
values. Dr. Polasky responded that considerations of long-term ecological effects raise
issues of dynamics and sustainability. He noted that the committee has plans to address
this issue. Dr. Daniel noted that social-psychological research shows that people do
consider future environmental impacts in how they express values. Another committee
member spoke from the floor about the importance of recent research on discounting. If
discounting rates are uncertain, people give future conditions more weight. He
considered this result important if the goal is sustainable human welfare.
Another question pertained to the implementation of multiple approaches and
whether they might be tested against each other; how they might be used at different
scales of decision-making; and whether it was appropriate to take a tiered approach that
included a screening step to allow for budgeting of resources. A committee member
responded that the committee's view of the valuation process was designed to link
appropriate tools to the specific decision context faced by the Agency.
The next question concerned how to value possibly irreversible ecological effects,
such as introduction of exotic species, where there are many unknowns. Dr. Polasky
responded that there is a need for valuation methods that are appropriate for dynamic
ecological systems where there are major uncertainties and possible long-term impacts.
He noted that decision-making approaches exist for such situations. Analysts would
estimate the rate of introduction, the likelihood of harmful effects, and the relative
outcome of continuing an existing vs. alternatives proposed. Dr. Biddinger and Dr.
Daniel responded that early collaboration between economists, ecologists and social
scientists to identify significant ecological effects and early screening of possible benefits
against costs were important first steps to take to screen for such cases.
28
-------
tc.sr4?
Overview of Methods Being
Considered by C-VPESS
SAB Workshop: Science for
Valuation of EPA's Ecological
Protection Decisions and Programs
Valuation Process Diagram
1A. Identity possible
Identity what
matters to people
3cological effects
l. Identity ecological
effects that matter
3. Characterize/ quantify
ecological effects
4. Characterize/ quantify
human consequences of
ecological effects
5A. Estimate value of effects
etary terms
Communicate
results to public
and decision makers
LEGEND Actions
Valuation Process Diagram - Eco-Centric Approaches
I A. Identity possible
ecological effects
dentity what
matters to people
2. Identify ecological
effects that matter
valuation
* ^ and Preference
Methods
J. Characterize/ quantity
ecological effects
4. Characterize/ quantity
human consequences of _ 3SS
ecological effects «»
3A. Estimate value of effects od. Estimate monetary
etary terms value of effects
Communicate
results to public
and decision makers J
LEGEND Actions
Categories of Eco-centric Approaches
Ecological Models
- Production Functions
Energy and Materials Flow
- Embodied Energy Value
- Emergy
- Ecological Footprint Analysis
Spatial Representation
- Spatial Representation of Biodiversity and
Conservation Value
Indicators Approach
- Ecosystem Benefit Indicators
Habitat Approaches
- Habitat Equivalency Analysis (NEA)
Decision Frameworks
- Net Environmental Benefit Analysis (NEBA)
Ecological Models - Production Function
~ Ecological science can deliver predictions of
ecological change (or prevention of change)
associated with effects of agency actions
~ Numerous Ecological production models
developed and primers exist (e.g. Primer of
Ecological Theory - Roughgarden 1998)
~ Can connect material outputs to stocks and
services if services are well defined.
(Research focus area)
~ Ecologists may not have data available on
shelf to parameratize every ecological system
for EPA's use. (Research focus area)
The Environment
Society
Societal
Benefit
K Compensatory
: Mechanisms
Extinctions !
pressure
Ecological
Service
Production Function
Stress |
¦ Stressor-
exposure fl
: response
profile
B function
Ecological System
Ecological
Element
Stress production
function. (Loading)
Social
Service
Action
(Public, Private)
I Engineering control
I or practice
1cos,s I
[ Waste j Energy 2
29
-------
Research Proposal
Identify ecosystem service provider
Determine aspects of community structure
that influence function
- Compensatory Response Mechanism
- Non-random extinction sequences
Assess key environmental factors
influencing provisions of services
Measure the spatial-temporal scales at which
services operate
Claire Kremen - Ecological Letters (2005) 8:468-479 - Managing Ecosystem
Services: What do we need to know about their ecology?
Energy and Material Flow Analysis
Quantifies energy and material flow through complex
ecological and economic systems
Input-Output Analysis or flow accounting methods
Produce estimates of the cost of goods in energy terms
~ Embodied Energy Value - Solar energy is the only Primary input to
global ecosystems. Focus on estimating total (Direct and indirect)
energy consumption for an economy (Costanza, 1980)
~ Emergy - Considers all systems to be networks of energy flow and
presents an energetic basis for quantification or valuation of
ecosystem goods and service Odum et. El. (2000).
~ Ecological Footprint Analysis - A variation of energy and material
flow that converts impacts to units of land. Total area of productive
land to ... (support population) - [Costanza. R (ed.) 2000]
Spatial Representation
Focus on Biodiversity and Conservation
value
Numeric representations of uniqueness,
irreplaceability and level of imperilment
Scale linked ecological target(s)
Cumulative biological, ecological and
conservation value
Stoms et.al. 2005 - Choosing surrogates for biodiversity conservation in
complex planning environment. Journal of Conservation Planning. 1: 44-63
Ecosystem Benefits Indicators
Quantitative and visual Approaches to assessments
and Land Use
Summarize and quantify complex information and
employ economic principles
Benefits are expressed as bundles of indicators,
both biophysical and socio-economic
Indicators mapped in Geographic Information
System (GIS) context
Indicators can be utilized as input to Trade-off
Analysis
Scale of assessment application can range from
regional to national.
James Boyd - What's nature Worth? Using Indicators to open the
Black Box of Ecological Valuation. 2004. Resources, pp. 18-22
Habitat Analysis
Habitat Equivalency Analysis was developed for use in
Natural Resource Damage compensation assessments
Quantifies units of habitat needed to provide same level
of service over time that was lost due to an injury
Simultaneously quantifies injury and scales the size of
restoration
Calculates Net Present Value (NPV) measure of service
in Discounted Service Acre Years (DSAYs)
HEA has mostly been applied at the local or watershed
levels
Net Environmental Benefits Analysis is a
management framework and some advocate NEBA and
HEA be linked in remediation / restoration activities
Social-Psychological Methods
1 A. Identify possible 1B. Identify what
ecological effects matters to people
1
2. Identify ecological
effects that matter
Ecological 3- Characterize/ quantify
Valuation 1
Models ecological effects j |
4. Characterize/ quantify
human consequences of '-5 1
effects
H 5A. Estimate value of effects 5B. Estimate monetary
1 in non-monetary terms value of effects
1 6. Communicate
y Methods 1
LEGEND
Actions Tools
30
-------
Background Issues
Human judgments as basis for ecosystem values
- Publics and stakeholders are relevant for EPA
- May include biocentric, moral and other values
Whose judgments?
- Experts, stakeholders, general publics, citizens
- Well-informed to ill-informed, rational to emotional
Context matters
- Values uncovered versus values constructed
Value metrics
- Multiple value dimensions: preference, importance ...
- Relative, incommensurate
Resolving value conflicts and tradeoffs
- Negotiation versus calculation
Socio-Psychological Methods -1
Surveys
- Standardized formal questionnaires
- Large representative samples
- Mail, phone, face-to-face, intfijjjSSr
- Closed responses (choice, ate)
Focus Groups
- Facilitated discuss' \ 'ieliberation
- Small relevant group cused topics
- Open respor. (comments)
Narrative i ijHMN's
- Loosely sti ed conversations
- Selected info.mants
- Emphasis on depth over breadth
Socio-Psychological Methods - 2
Behavior Observation / Behavior Trace
- Visitor/user behavior in the environment
- Direct, cameras, counters / trail erosion
- Correlate A behavior, A environment
Interactive Games
- Computer simulations, visualizations, VR
- A behavior A environment models
Referenda and Juries
- Sanctioned decisions/verd^^^3alues
- Social/civil context
Deliberative
- Facilitat^^^Sliied tradeoff analysis
- Value ^SLjtion/construction
- Consensus (?)
Identify What Matters to People
1B. Identify what
matters to people
I
S-P methods provide tested means for
systematically identifying and articulating
public desires and concerns relevant to
public environmental policy
I US DA Forest Service Survey
19. The most important role for the public lands is providing
jobs and income for local people.
15. Forests have a right to exist for their own sake, regardless
of human concerns and uses.
Shields et al 2002
u
Identify What Matters to People I
S-P methods provide tested means for Ns
systematically identifying and articulating s
public desires and concerns relevant to
public environmental policy
Pending NOAA Fisheries Survey (draft)
State and federal marine waters are managed for the benefit of
current and future generations. Which of the following should be
emphasized in the management of our marine waters?
~ Improving their natural conditions, such as wildlife, water
and scenery
~ Developing commercial opportunities such as commercial
fishing, energy development and shipping
~ Balancing natural conditions and commercial opportunities
about equally
~ I am unsure
Identify Ecological Effects
That Matter to People
7. Conserving and protecting forests and grasslands that are
the source of our water resources, such as streams, lakes, and
watershed areas.
Not at all
important 1
9. Protecting ecosystems and wildlife habitats.
Very
important
Shields et al 2002
31
-------
Estimate value of effects
in non-monetary terms
5. Developing new paved roads on forests and grasslands for
access for cars and recreational vehicles.
Very Very
unfavorable 1 2 3 4 5 favorable
1. Expanding access for motorized off-highway vehicles on
forests and grasslands (for example, snowmobiling or4-wheel
driving).
6. Designating more wilderness areas on public land that stops
access for development and motorized uses.
Shields et al 2002
Group Determination of Values
Referenda and Juries
-Sanctioned decisions/verdicts =>
values
-Social/civil context
Deliberative Groups
- Facilitated, informed tradeoff analysis
-Value elicitation/construction
- Consensus (?)
Group Determination of Values
Decision processes where outcome is
determined by groups (not individuals)
Economic approaches to valuation
typically aim to ascertain the values of
individuals
People acting as citizens in group
decisions may respond differently than
people acting as consumers in individual
decisions
Group Determination of Values
Examples of group decision processes
from which one might ascertain
information about the value of ecosystem
services
-Voting on referenda
- Deliberative value elicitation/citizen juries
- Civil court jury awards
Voting on Referenda
Referenda are formal solicitations to the
public to determine the public's willingness
to pay
Example: ballot initiative to purchase
open space
- 1,373 votes on community funding for parks
and open space, 1996-2004
- 1,062 passed; $26.4 billion funding committed
- Source: Trust for Public Land
Voting on Referenda
Advantages
- Direct expression of public preferences
- Evidence about median voter preferences
- Responses as citizens about a public good
Disadvantages
-Votes may reflect views on multiple subjects
(e.g., views on taxes)
- Can't directly infer aggregate valuation
32
-------
Voting on Referenda
Aggregation issue
example
- 3 person community
- Each has a value of the
public good
- Each faces a property tax
(tax price)
Total value of public
good: 220
Total cost: 300
Referendum passes 2-1
Voter
Value Tax
Price
Voting on Referenda
Reverse example
Total value of public
good: 300
Total cost: 220
Referendum fails 2-1
Voter
Value
Tax
Price
A
200
100
C 50 60
Deliberative Value Elicitation
& Citizen Juries
A problem with asking people their views
on the value of ecosystem services:
- Many people are not well informed: complex
issues that are largely ignored by the general
public
- Many people will not have thought carefully
about values or have well formed views
Deliberative Value Elicitation
& Citizen Juries
Deliberative approaches
- Form a small group (voluntary participation)
- Provide group with technical information
- In-depth discussion
- Structured decision-making
Deliberative Value Elicitation
& Citizen Juries
Advantages
-Thoughtful valuation based on information
and careful deliberation
Disadvantages
-Self-selection bias: is group representative of
general public
- Dominant individual effect
- Effect of how process is structured may affect
the results
Civil Court Jury Awards
Liability rules: require an entity that damages
ecosystems to
- restore
- replace with a functional equivalent
- pay monetary damages equivalent to lost value
Example: Natural Resource Damage
Assessment (NRDA)
- CERCLA (Superfund)
- Oil Pollution Act
Concern: high cost of litigation
33
-------
Economic Methods
1 A. Identify possible 1B. Identify what 1
ecological effects matters to people 1
/ 2. Identify ecological
1 cr-nimimi / 3. Characterize/ auantifv
Valuation 1
Models ecological effects j |
4. Characterize/ quantify 1
1 Methods 1
ecological effects 1
5A. Estimate value of effects 5B. Estimate monetary
in non-monetary terms value of effects
LEGEND
Actions Tools
Economic Approaches to Valuation
Dominant methodology for quantitative
assessment of valuation of ecosystem services
Built on microeconomic theory that has been
well developed
Translates values into a common currency
(monetized value) that makes comparisons easy
to comprehend
Some object to converting "priceless"
environmental attributes into money equivalents
Conceptual Foundation: Utility
Theory and Welfare Economics
Key concept: what tradeoffs makes an
individual equally well off?
Willingness-to-pay: how much money
would an individual give up to buy an
ecosystem service?
Willingness-to-accept: how much money
would an individual need to receive in
exchange for taking away an ecosystem
service?
Market-based Valuation
Some ecosystem services provide
marketed commodities, directly or
indirectly
Examples:
-Value of increased fish harvest from improved
water quality or protection of coastal wetlands
-Value of increased crop production from
pollinators
Non-Market Valuation
Revealed Preference
-Travel Cost Method
- Hedonic Approach
-Averting Behavior
Stated Preference
-Choice Experiments
Contingent Valuation,
Conjoint Analysis
Travel Cost Method
Opportunity cost for participating in outdoor
recreation: travel time and out-of-pocket
expenses
Opportunity costs provide an implicit price for
recreational trips
Estimate how trip choices change on the basis
of trip costs and other variables (e.g. site
qualities...)
Derive a willingness-to-pay (demand) for trips
34
-------
Hedonic Approach
Revealed Preference Methods
Some purchased goods as composite goods
whose values depends on many characteristics
Example: value of a house depends upon
- Structural characteristics (e.g., sq feet, age, # of
bedrooms...)
- Environmental characteristics (e.g., air quality, access
to open space...)
Controlling for other characteristics, how does
willingness-to-pay vary with environmental
characteristic of interest
Advantages:
- Based on observable behavior for decisions
with real consequences
Disadvantages:
- May only apply to a small set of ecosystem
services (e.g., travel cost - recreation)
- Questions about specification of empirical
equation (explanatory variables, functional
form...)
-Are individuals fully informed about choices?
Stated Preference Approaches
Choice experiments: survey asking
individual to make choices
- Contingent valuation: offer a choice about
whether individual would pay a specified price
for a specified increase is an ecosystem
service
- Conjoint analysis: offer bundles of services
and price and ask which is preferred
Stated Preference Approaches
Advantages
- Direct question about values
- Applicable to ecosystem services for which there is
no direct observable behavior ("non-use" values)
Disadvantages
- Hypothetical - would people really pay what they say
they will?
- How well informed are respondents?
- How much are responses influences by question
format?
Other Approaches
Rather than try to estimates value directly,
use evidence on cost to infer something
about value (similar to averting behavior)
Replacement Cost
Marketable Permit Prices
Replacement Costs
What would it cost to replace an ecosystem
service with human engineered solution?
Example: Catskills/New York City water supply
To be valid, must meet three conditions:
- Human engineered solution provides equivalent
quality/quantity of service
- Solution is least cost alternative of providing the
service
- Individuals in aggregate would be willing to incur the
cost if ecosystem service were not available
35
-------
Marketable Permit Prices
Marketable Permit Prices
Cap-and-trade systems
-Tradable emissions permits (pollution)
- Individually transferable quotas (fishing)
Observable price for permit/quota
Examples: S02 and C02 markets
Can the price of a permit/quota be used to
infer the value of an ecosystem service?
Price of emissions permit reflects the
marginal cost of meeting the cap
Price will depend on stringency of cap
-European Exchange: 21.18/metricton
-Chicago Climate Exchange: $1.65/metricton
(Prices as of Dec 9, 2005)
Cross-Cutting Issues
Benefits T ransfer
Dynamics
Uncertainty
36
-------
7. Addressing Uncertainty in Ecological Valuation and Expert Elicitation
Dr. William Ascher provided a short presentation (see slides at the end of this
section of the workshop report) focused on the analysis of uncertainty in ecological
valuation and how to convey uncertainty to policy makers. He noted different kinds of
uncertainty and then discussed how uncertainty analysis might proceed and be
communicated to policy makers, given their use of uncertainty (e.g., for edging,
contingency planning, and communication with the public and other policy makers) and
their preference for greater certainty and tendency to equate uncertainty with weak
analysis.
He suggested that analysts need to keep the range of uncertainty prominent. He
provided some initial conclusions, under discussion by the C-VPESS, about how
ecological analyses can convey uncertainty effectively to policymakers through
appropriate formats, methods, building on existing models, and processes for working
with policymakers or "gatekeepers" such as the Office of Regulatory and Information
Affairs at the Office of Management and Budget.
He then turned to a short discussion of expert elicitation as one method to
determine degrees of uncertainty and disagreement, understand their bases, and reduce
uncertainty and disagreement. He defined expert elicitation broadly as the "use of expert
judgment in an analysis" that entails "second-hand" analysis of available data. He
described four different kinds of expert elicitation [compilations of existing judgments;
individual-expert syntheses; expert-interaction approaches, such as the Delphi approach;
and the Technical Facilitator/Integrator (TFI) approach]. He described these approaches
as differing in cost, formality, extent and nature of any interactions among experts, and
the form of the final information provided to policy-makers from the elicitation.
After Dr. Ascher concluded the presentation, Dr. Robert Costanza provided brief
remarks. He noted that ecological risk assessment has many different kinds of
uncertainties, but that only some of them are adequately captured through quantitative
methods such as Monte Carlo analysis. As a result, analysts need to "bracket" other
kinds of uncertainties for decision-makers. The goal of uncertainty analysis, in his view,
was not to reduce uncertainties (because additional research might paradoxically increase
uncertainty about ecological effects). Instead the goal is to represent uncertainties in
ways that allow better decision-making in the face of uncertainty. In his view, this issue
was linked to the question "who bears the burden for uncertainty?" If there is no full
discussion of uncertainty, the public will typically bear the burden of adverse effects. He
suggested that the burden might better be place on parties standing to benefit from
creating uncertainties. Assurance bonds might be one mechanism where uncertainty
analysis could be used to foster better decision making. If polluters were required to
purchase and hold bonds until potential damages identified as possible in a worst-case
analysis were not demonstrated, the public might be better protected against high-impact,
highly uncertain risks.
37
-------
A workshop participant noted that the SAB might consider this idea in its review
of EPA's annual research budget and research planning efforts. There is a science
component to adaptive management that can help foster learning from policy
experiments, if decisions are structured appropriately and their impacts studied.
Another workshop participant spoke of the importance of uncertainty analysis in
ecological valuation and recounted his experience providing advice to the Agency for its
812 Study. He noted that the Agency had a high threshold for committing a Type 1 (false
positive) error in monetizing ecological benefits associated with ecological protection
resulting from implementing the Clean Air Act. When EPA does not include those
highly uncertain ecological benefits in its monetary estimates, the Agency may commit,
in his view, a Type 2 (false negative) error.
Another participant asked if there are points in the valuation process where expert
elicitation may best be involved. Dr. Ascher responded that the C-VPESS is wrestling
with this issue. Expert elicitation is probably more useful in the production function part
of an analysis and less useful in expressing how the general public values some
ecological change. Dr. Costanza also commented that often there are differences in
knowledge of experts and the public about the connections between ecological impacts
and human welfare. He suggested that it would be useful to find a role for expert
opinions, a role especially important until the public is better educated about ecological
processes. The committee's emphasis on "methodological pluralism," where results are
triangulated across several methods to assess ecological value, can provide better
analyses to support decisions.
A final questioner asked whether there was a role for fuzzy logic and sensitivity
analysis in the committee's exploration of uncertainty methods to support ecological
valuation. Dr. Costanza responded that the committee should explore those ideas.
38
-------
How to conduct valuation,
given uncertainty
1. Different valuation approaches
for different types of
uncertainty?
Stochastic uncertainty
Theory uncertainty
Data limitations
2. Probability-estimating
techniques within the valuation
(e.g., Monte Carlo, expert
elicitation)
Addressing Uncertainty in
Ecological Valuation; Expert
Elicitation
SAB Workshop: Science for
Valuation of EPA's Ecological
Protection Decisions and Programs
How to conduct valuation
given uncertainty
3. How to most efficiently reduce
uncertainty?
Maybe requires a diagnosis
of types & sources of
uncertainty
Budnitz et al. on
earthquake prediction
How to convey uncertainty
to policymakers
Requires knowing how policymakers do &
should use understandings of uncertainty
- For hedging
- For contingency planning
- For communication with public & other
policymakers
How to convey uncertainty
to policymakers
Requires knowing how policymakers do &
should use understandings of uncertainty
- Tendency to prefer greater certainty
-Tendency to fold estimate & uncertainty
indications into prior assessment
- Tendency to equate uncertainty with weak
analysis
How to convey uncertainty
to policymakers
Implications:
1. Need to keep range of uncertainty
prominent
2. Need to convey assumptions so that
policymakers can judge credibility &
convey to audiences
3. Need to provide enough information so
that bca & other decision aids can
incorporate the uncertainty
39
-------
How to convey uncertainty
to policymakers
1 .Appropriate formats (e.g., confidence
intervals, probabilities, etc.)?
2.Uncertainty tests (e.g., Monte Carlo,
expert elicitation)
3.Useful models adaptable from other
applications (e.g., "risk characterization"
for health risks)?
4.How much information to be conveyed?
How to convey uncertainty
to policymakers
5. What processes of interaction between
analysts & policymakers?
6. How to overcome institutional or other
obstacles to conveying uncertainty?
7. Choice of models & visual aids
8. Should policymakers (or gatekeepers
such as OIRA) insist on protocols for
expressing uncertainty?
Expert Elicitation: Defining Expert Elicitation
Use of expert judgment in an analysis
Also called "multi-expert opinion"
methods
Entails "second-hand" analysis
Wide range of different methods
Purposes of Expert Elicitation
1. Undertake analysis using others' expertise
Typically less costly in time & money
Potential to incorporate broad range of
wisdom & insights
2. Determine degree of uncertainty &
disagreement
3. Understand the bases of uncertainty &
disagreement
4. Reduce uncertainty & disagreement
Dimensions of Differences
1. breadth of expertise of the individuals
involved in the exercise
2. degree and nature (if any) of interaction
between the experts and those
conducting the expert elicitation
3. degree and nature (if any) of interaction
among the experts
- e.g., face-to-face vs. mail responses
Dimensions of Differences
4. formality
- from brain-storming to highly-structured,
often computer-aided
5. degree of synthesis and interpretation by
study conductors
6. [for multi-round methods]: nature of
feedback that the experts are provided from
one round to the next
40
-------
Dimensions of Differences
7. how judgments are aggregated & presented
- from simple means or medians, to full
probability distributions, confidence
intervals, & explanations for the
differences
- some use Bayesian methods to
incorporate quantitative probability
estimates
Specific Elicitation Methods:
1. Compilations of existing judgments
E.g., "consensus economic
forecasts"; biodiversity hotspots
Can present means, medians &
ranges
Very inexpensive & fairly quick
Specific Elicitation Methods:
1. Compilations of existing judgments
E.g., "consensus economic
forecasts"; biodiversity hotspots
Can present means, medians &
ranges
Very inexpensive & fairly quick
Specific Elicitation Methods:
1. Compilations of existing judgments
Risk of including obsolete judgments
Recency is highly correlated with
accuracy, at least for forecasts
Risk of mixing different concepts
No interaction among experts
No direct gauge of uncertainty
Spread only partially revealing
Specific Elicitation Methods:
2. Individual-Expert Synthesis
Granger Morgan
Individual interviews (no interactions
among experts) synthesized by study
conductors
Experts of similar expertise
Therefore gauging degrees of
disagreement
Presents reasons for judgments
Specific Elicitation Methods:
3. Expert-interaction approaches
Brainstorming:
- Unstructured
- Potentially rich interactions
- Vulnerable to intimidation,
groupthink, etc.
41
-------
Specific Elicitation Methods:
3. Expert-interaction approaches
Specific Elicitation Methods:
3. Expert-interaction approaches
Delphi
Multiple rounds of requests for
judgments, feedback on medians &
ranges, opportunities for revisions
No direct, face-to-face interactions
among experts
Avoids groupthink, etc.
Delphi
Premium on wide range of expertise
Typically, some degree of
convergence
Therefore not a measure of pre-
existing agreement/disagreement but
rather of agreement potential with
multi-disciplinary interactions
Specific Elicitation Methods:
3. Expert-interaction approaches
Delphi
- Sacrifices richness of direct interaction
Risk that different understanding of
terms will lead to false indications of
differences in judgments
Specific Elicitation Methods:
3. Expert-interaction approaches:
Technical Facilitator/Integrator (TFI):
4 stages:
1. Literature review & assessment +
own integrator's expertise
2. Integrator interacts with experts to
assess judgments, issues,
distribution of judgments
3. Integrator facilitates expert
interactions
4. Expert panel assesses overall record
& distribution of judgments
Specific Elicitation Methods:
3. Expert-interaction approaches:
Technical Facilitator/Integrator (TFI):
Number of stages depends on
resources
Obvious parallels to other methods
Huge burden on integrator/facilitator
42
-------
8. Discussion of Specific Methods Featured in Workshop Breakout Groups
Dr. Buzz Thompson introduced the topics of the five breakout groups (Economic
Analysis and Ecological Production Functions; Group Expressions of Value: Referenda
and Citizen Juries; Deliberative Approaches for Modeling, Valuation, and Decision
Making; Social/Psychological Methods for Ecosystem Values Assessments; and Spatial
Representation of Biodiversity and Conservation Values and Ecological Services). He
emphasized that the purpose of the breakout sessions was for the C-VPESS to receive
input and feedback from meeting participants. He pointed meeting participants to the
suggested questions for breakout sessions included in their meeting materials1 and the
background materials from breakout sessions provided for them (see section beginning
on page 89 of this report).
Breakout sessions met from 3:30-5:30 p.m. on December 13, 2005. Meeting
participants reconvened in plenary session at 8:30 a.m. on December 14, 2005 to hear
brief reports about the previous day's five breakout sessions and to participate in brief
question and answer sessions for each. Dr. A. Myrick Freeman, a C-VPESS member,
introduced the speakers and moderated the session.
8.1. Break-out Report: Economic Analysis and Ecological Production Functions
Dr. Joan Roughgarden and Dr. Stephen Polasky reported on the breakout session
on economic analysis and ecological production functions (materials for this breakout
session may be found on page 89). Dr. Roughgarden noted that ecological science has
advanced to the stage where EPA could take greater advantage of linking outputs of
ecological models to economic valuation. They noted the need for "quick and cheap"
methods but Dr. Roughgarden observed that ecological studies cannot be "second rate"
and encouraged EPA to "aim high, despite constraints." They noted that often there is a
mismatch between the research efforts of academic ecologists and EPA's needs. There
could be great benefits if models known to work for specific ecosystems and
parameterize were paramaterized for other places of regulatory concern.
A workshop participant asked whether there are well-articulated criteria for
quality of ecological studies and data in the regulatory arena and whether practical short-
cuts, given constrained resources, might critically compromise the quality of studies. The
breakout presenters did not respond with specific criteria. They did discuss, however,
strategies for targeting ecological research. They discussed EPA's taking advantage of
What methods or aspects of methods seem most promising for EPA to adopt or explore? In what kinds
of decision contexts would EPA use those methods? Where would they fit in a valuation process?
-What should be added to or changed in the methods discussed to make them more credible and more
useful?
-What are the barriers to adoption of promising new methods and how might they be best overcome?
-Which issues would benefit from additional exploration and research? How could the SAB, CASAC, and
Council be helpful in providing such advice?
-Given whatever approaches to valuation you are considering, how should be degree of uncertainty be
gauged and conveyed?
43
-------
the research surveyed for thee Millennium Assessment. They discussed the need to
consider investments in peer review in light of the overall need for ecological-economic
modeling, with a balanced allocation of effort in both dimensions. Dr. Roughgarden
noted that breakout participants had observed that advances in computing technology for
economic analyses might provide needed resources for extending ecological analysis.
A questioner responded with a question about the appropriateness of using
"Knowledge Networks" to identify survey respondents. This web-enabled strategy
results in high response rates, in his view a pre-requisite of a "good study," but he
expressed concern that the survey groups identified by "Knowledge Networks" might not
reflect the population as a whole. He asked about the availability of guidance that might
help EPA manage the tension between "doing a study that meets technical requirements"
and taking "shortcuts that make science practical in the real world."
Dr. Polasky responded that the 80/20% rule might apply. He noted that the C-
VPESS must wrestle with ways EPA can get most of the information needed for a limited
set of benefits and avoid emphasizing "the perfect over the practical." Dr. Roughgarden
agreed that there is a difference between "getting it right" and "getting it perfect." She
noted a need for a "systematic push" to minimize type 1 errors and avoid egregious type 2
errors. Nevertheless, in her experience, ecological studies generally take five years to
complete. Dr. Roughgarden suggested that the Agency pursue quality research and that
the Agency should adopt "stopgap" policy options until ecological data is available.
A participant asked whether there was also a shortage of technically competent
people willing and able to engage in applied work. Dr. Roughgarden responded that
there is a large pool of young Ph.D. ecologists now untapped. Graduate ecologists in her
experience never think of EPA as a potential employer. She suggested that better
outreach by EPA would tap this enormous potential. Dr. Polasky similarly acknowledged
the capability within EPA's National Center for Environmental Economics and also noted
that many economists outside EPA wanted to work with ecologists. He believed that the
tradition of multi-disciplinary collaboration was stronger outside the Agency than within.
The last question concerned the importance of an accessible platform to make
research data available to outside academic researchers, as well as to the Agency. A
workshop participant noted that if a common platform were conveniently designed with
good spatial data information, a wide variety of social, economic, and ecological data
could be used to enhance the science supporting ecological valuation. Data developed for
local or special purposes could be cross-checked, studied for their alignment with other
data, and appropriate standards developed. He noted that such platforms were routine in
labor and health economics. Dr. Roughgarden responded that such a platform could be
linked to the well-established Long Term Ecological Research network of data stations.
This existing network has a 20-year tradition of data-sharing and standards that might
accommodate research on ecological production functions and other needed ecological
valuation research.
8.2. Group Expressions of Value: Referenda and Citizen Juries
44
-------
Dr. William Ascher reported on the breakout group focused on referenda,
initiatives, and citizen juries (slides summarizing the breakout discussion are included at
the end of this section of the workshop report; background materials and presentations
used in this breakout session may be found on page 92 below). He noted that while these
approaches may be considered "unorthodox," they have the potential to capture values
not reflected in actual markets and reflect group expressions of value. In addition,
referenda and initiatives involve actual, not hypothetical, decisions, reached on particular
issues. He reported that the break-out group criticized the use of analyses of referenda
and initiatives due to concerns about the politics associated with voting and the
complications associated with interpreting results and using them in benefit transfer. The
breakout group, however, did discuss their use in validating more conventional methods.
Dr. Ascher reported on the breakout group's discussion of citizen juries. Despite
concern about the novelty of this approach, lack of possible representativeness of
membership injuries, and standards for legitimating the approach, the group saw some
potential in capturing types of values that elude other methods. The group shared a
strong sentiment for multiple methods and a concern for the problem of under-estimation
of hard-to-quantify ecosystem components.
Dr. Ascher's presentation slides are included at the end of this section of the
workshop report.
In the question and answer period, a meeting participant asked about the use of
the term "jury," since American society gives legitimacy to jury decisions, which usually
make decisions in an adversarial context. The participant observed that the juries
described operated more like grand juries in the American system and suggested that the
literature on grand juries might be helpful.
Another meeting participant suggested that group values relating to ecological
protection might be very well captured through the approach developed by Henry Willis
and Michael DeKay of Carnegie Mellon, who have developed a framework for ecological
risk-ranking, which has proven robust through many applications.
A final participant expressed concern over the target audience intended for the
group approaches described. She saw problems associated with limiting involvement to
the middle-class individuals likely to participate.
45
-------
Inferring Values from Public
Choices
W. Ascher
VPESS
December 2005
Vehicles of Public
Input Reflecting Values
Individual-response
valuation
public opinion polls
non-governmental fora
public hearings
public notice and
comment
concertation
quasi-governmental
commissions
direct community
decision-making through
town halls, etc
elected representation
public opinion polls
letter writing, emailing or
calling an elected
representative
taking existing public
policies as the revealed
preferences of society
contributions to non-
governmental efforts
willingness to accept
negotiations
referenda/initiatives
1.
Revealed Preferences from
"Public Choices"
Referenda/initiatives
Referendum: legislature calls for a
public vote
Initiative: citizen petition
Usually for eco-system improvements
2. Willingness-to-accept negotiation
outcomes
Best if voted; but could have other
indications of "close call"
Different conception of
what value is:
Intensity
Median: the majority (or close to majority)
of sufficiently engaged people believe that
the expense is worth it
-Closer to 50-50, the better, though floors
or ceilings can be estimated regardless
Different conceptions:
Intrinsic validity
-IF one accepts that society's decisions
have standing as expressions of value
-Whether private utility or public
regardedness
Different conceptions:
Conception of democracy & representation
-Anti-Burkean, non-Benthamite
Burke: representatives, not citizens,
choose what is good for the people
Bentham: greatest [private] good for
the greatest number
-Government should do what the public
wish government to do
46
-------
Criticisms
1. Referenda & initiatives are subject to
intense politicking
- But politicking is pervasive &
democratic
2. Perceptions of benefits & costs may
diverge from actual stakes
- But it is possible to follow up with
surveys to determine how the stakes
are perceived
Criticisms
3. Other issues determine the vote
-Popularity of backers; partisan
maneuvering
-For willingness-to-pay: restrict to simple-
issue referenda or initiatives
-For willingness-to-accept: simple-issue
Coasean negotiations
Criticisms
Not capable of determining the option of
greatest aggregate utility
60% favor because their net gain is
+$100
40% against because their net gain is -
$200
-Usually true, but logic is simply different
-Assuming 0 value for opponents, a floor
on mean value is possible
Complications:
Different benefits transfer complications
-Disentangling objectives if multiple
issues
Or, contingent valuation keyed to
actual cases of pending decisions
Complications:
More than 50% vote margin will
underestimate the community's collective
valuation
-Result is therefore a floor
Multiple issues obscure the willingness to
pay for any single benefit
-Go for simple-issue referenda or
initiatives
Different benefits transfer complications
Validation of More Conventional
Valuation Methods
Several studies predict referendum votes
from contingent valuation estimates; check
whether the predictions are borne out
47
-------
Validation of More Conventional
Valuation Methods
NOAA Panel:
[E]xternal validation of elicited lost passive use
values is usually impossible. There are however
real-life referenda. Some of them, at least, are
decisions to purchase specific public goods with
defined payment mechanisms, e.g., an increase
in property taxes. The analogy with willingness
to pay for avoidance or repair of environmental
damage is far from perfect but close enough that
the ability of CV-like studies to predict the
outcomes of real-world referenda would be
useful evidence on the validity of the CV method
in general.
Validation of More Conventional
Valuation Methods
The test we envision is not an election poll of the
usual type. Instead, using the referendum format
and providing the usual information to the
respondents, a study should ask whether they
are willing to pay the average amount implied by
the actual referendum. The outcome of the CV-
like study should be compared with that of the
actual referendum. The Panel thinks that studies
of this kind should be pursued as a method of
validating and perhaps even calibrating
applications of the CV method
8.3. Deliberative Approaches for Modeling. Valuation, and Decision Making
Dr. Harold Mooney reported on the breakout-group focused on decision-aiding
approaches and mediated modeling (slides summarizing the breakout discussion are
included at the end of this section of the workshop report; breakout materials and a
related presentation may be found on page Error! Bookmark not defined, of this
workshop report). He presented a short overview of the approaches featured by Drs.
Joseph Arvai and Robert Costanza and quickly summarized the discussions of the
breakout group. The group discussed the issue of identifying and engaging
representative stakeholders, gaming and bias issues, and noted that the SAB published a
science and stakeholder involvement study in 2001. The breakout group also posed the
following questions:
How can these techniques be used at the national level?
How do we get more social science involved in EPA in order to do these
activities meaningfully? How to get EPA to take non economic valuation
more seriously?
Who is going to own and rerun the models through time?
How complex can the models be and still be transparent?
How are uncertainty, non-linearities, dealt with?
Limitations of costs and available talent
Value measures are relative (no common metric)
Dr. Mooney reported that the breakout group saw many benefits associated with
the methods:
Emergent values result
Local focus makes it easier for decision-making
Social learning an important by-product of process
Benefits in relation to cost are high
48
-------
Transparency
Procedural equity
Collaborative
Dr. Mooney invited Dr. Bruce Hull to provide some additional commentary. Dr.
Hull noted that the approaches described can work well but may feel difficult at first for
experts. He saw these deliberated and mediated processes offering benefits when
working with local stakeholders in specific places. The social learning that results
through these processes builds the capacity of local groups to solve environmental
problems.
A workshop participant added that mediated modeling approaches have been used
by the Army Corps of Engineers in collaboration with Sandia National Laboratories.
Although the process is expensive, it can help obtain full buy-in and support for decision
options.
A different point was made by another workshop participant. She noted that
almost all the information needed for mediated modeling is required for expert valuation.
She viewed mediated modeling as an alternative decision process, not an alternative form
of valuation. Yet another participant took issue with this view. He noted that group
processes entail substantial value synthesis and is different in nature from expert
valuation and decision making.
Dr. Mooney invited Dr. Arvai to make several remarks. Dr. Arvai commented
that values need to be considered across several alternatives and are only meaningful in a
comparative context, where there are consideration of the multiple attributes relevant to a
decision. Decision-aiding approaches are ways to synthesize information across several
valuation methods to infer value from their results.
A workshop participant noted that EPA has experience with such methods
through negotiated rulemakings and pointed to the Agency's experience with the Phase
Two Microbial Disinfectant Byproduct (MDB) Rule. He acknowledged the merit of such
a process, but offered two cautions. Stakeholders involved in such processes have
difficulty balancing their genuine desire to arrive at a socially acceptable outcome against
representing their constituency. Therefore, the outcome of such efforts depends greatly
on the negotiating skills of individuals representing their constituencies. He also noted
that stakeholder identification is critically important. He also noted that such processes
are lengthy and expensive. The MDB Rule developed a near-consensus regulation that
opened up only slight relief to small drinking water systems that were the focus of the
rule-making and who were not directly involved in the negotiation.
Dr. Arvai acknowledged that stakeholder identification is always critical. He
disagreed with the previous commenter's remarks and noted that decision-aiding
approaches are not aimed at negotiation, but at identifying the relevant components of
values and describing the range of relevant values and their components to decision-
49
-------
makers.
50
-------
Deliberative Approaches for
Modeling, Valuation and Decision
Making
Joseph Arvai
Robert Costanza
Discussion leaders and early departers
The essence of Arvai
The value of ecological systems and services
emerges from stakeholder deliberation
The overall value of ecological systems and services
is multiattribute in nature.
The value of systems or services is established
through an analysis of management alternatives:
thus, the value of ecological systems and services is
relative and reflects the tradeoffs that people are
willing to make.
Joe's Procedure
Analyses of people's (stakeholders, public, experts,
etc.) objectives identifies the attributes of systems
and services that deserve attention in "valuation".
Consultation with technical experts (economists,
ecologists, etc.) identifies appropriate measures for
these attributes including:
natural measures (eg value of electricity); proxy
measures (habitat quality for fish); constructed
measures (indexes of accessibilitiy for cultural
and spiritual purposes)
OBJECTIVES, ATTRIBUTES, MEASURES
Objectives
Attributes; Measures
Recreation
Access to recreation opportunities; Weighted User Days
Environmental Health
Erosion levels; Weighted Erosion Days
Environmental Health
Flow levels; Weighted Flood Days
Environmental Health
Habitat quality, % Available Habitat, IBI
Environmental Health
Water quality; Multiattribute Index (particulates, PCBs, etc.)
Cultural
Regular access to sites; Consistency Index
Economic
Revenues; Annual Revenues M$/ Year
ESTABLISHING VALUE
Objectives
Attributes/
Measures
Hydrograph
Enhanced Summer
Releases
Enhance
recreation
opportunities
Access/
Weighted user
days
1400
1200
Enhance
environmental
health
Habitat Quality/
% Available
50%
20%
Maximize
economic returns
Revenues/
$Mil/Yr
$60
$80
ESTABLISHING VALUE
Attributes/
Measures
Enhance
recreation
opportunities
Enhance
environmental
health
Access/
Weighted user
days
Revenues/
$Mil/Yr
Habitat Quality/
% Available
saq
Instead, the value of a given
option exists in the tradeoffs
that people are willing to make
across not just their objectives,
but also the level of achievement
with respect to them. .
$60 $80 $55
51
-------
ADVANTAGES
Provides both preference orders for management
options.
MuItiattribute, inclusive, and transparent.
Useful for both decision making and retrospective
evaluation.
CHALLENGES
Big effort and potentially time consuming.
Decision makers may wish to protect their autonomy.
Not explicitly geared towards current OMB
requirements for regulatory evaluation.
LANDSCAPE SIMUlATfO
MODELING
rA
Three Step Modeling Process*
1. Scoping Models
high generality, low resolution models produced
with broad participation by all the stakeholder groups
affected by the problem.
2. Research Models
more detailed and realistic attempts to replicate the Increasing
dynamics of the particular system of interest with the Complexity
Mphads o« clibrto andtesfing: ^
and Precision
3. Management Models
medium to high resolution models based on the
previous two stages with the emphasis on producing
future management scenarios - can be simply exercising
the scoping or research models or may require further
elaboration to allow application to management questions
*fromCostanza, R. andM. Ruth. 1998. Using dynamic modeling to scope environmental pr
and build constmMiwironmental Managem&&:\83-195.
1
Gund Institute for Ecological Economics, University of Vermont
Degree of Understanding of the System Dynamics
J
High
EXPERT MODELING
MEDIATED MODELING
Typical result: Specialized
Typical result: Ccnsensus
model whose
onbothproblems/gcals and
recommendation never get
process - leading to
implements! because they
effective and
lade stakeholder support
implementable policies
L
Hl 9h^ Degree of Consensus
STATUS QUO
MEDIATED DISCUSSION
Typical result:
Typical result: Consensus
Confrontational debate
on goals or prcblems but no
and no improvement
help on how to achieve the
L
goals or solve the problems
Patuxent Watershed Scenarios'1
|ASTO lAhRos IF^I |d1p |S^c iH-TN-nhJ
* From: Costanza, R., A. Voinov, R. Boumans, T. Maxwell, F. Villa, L. Wainger, and
H. Voinov. 2002. Integrated ecological economic modeling of the Patuxent River
watershed, Maryland. Ecological Monographs 72:203-231.
52
-------
Spatial
Modeling
Framework
So what were the group issues??
The issue of identifying and engaging
representative stakeholders, gaming and bias
issues,etc. SAB did a stakeholder
involvement study in the past.
How can these techniques be used at the
national level?
How do we get more social science involved
in EPA in order to do these activities
meaningfully? How to get EPA to take non
economic valuation more seriously ?
More issues
Who is going to own and rerun the models
through time?
How complex can the models be and still be
transparent?
How are uncertainty, non-linearities, dealt
with?
Limitations of costs and available talent
Value measures are relative (not common
metric)
Benefits
Emergent values result
Local focus makes it easier for decision-making
Social learning an important by-product of
process
Benefits in relation to cost are high
Transparent
Procedural equity
Collaborative
8.4. Social/Psvetiological Methods for Ecosystem Values Assessments
Dr. Ann Fisher reported on the breakout-group focused on social and
psychological methods for ecosystem values assessments (slides summarizing the
breakout discussion are included at the end of this section of the workshop report;
breakout materials and a related presentation may be found on page 122 of this workshop
report). The scope of methods included surveys, focus groups, narrative interviews,
behavioral observations/behavioral traces, and interactive games. The group identified
possible EPA applications of methods that have potential for characterizing elusive non-
market values, especially important values that people resist expressing in terms of trade-
offs. They also suggested several additional ways social and psychological data could be
linked to eco-valuation. One breakout member described how the British government
was using standard multi-attribute approaches to risk for classifying reactions to a wide
variety of potential risks, including ecological risks. The group discussed barriers to
EPA's exploring such approaches, including overcoming the hurdles involved in
collecting data from more than 10 people. They discussed how the Agency might partner
with the US Forest Service and the National Oceanic and Atmospheric Administration,
53
-------
which uses social surveys to understand the value of different ecological protection
options.
In the question-and-answer-period that followed, a workshop participant noted
that it was important to decide when information about trade-offs are need, and when
other kinds of information about values can be helpful to decision makers. Another
workshop participant suggested that it would be useful for an SAB panel to look at the
scientific issues associated with data collection and offer advice regarding several
questions, including: when and how web panels might be appropriate; alternatives to
"Knowledge Networks;" criteria for good surveys, including the role of high response;
and formats for providing non-economic information to the Agency.
A workshop participant commented that at the local level, decisions regarding
values are not expressed in an economic framework. Social-psychological approaches
are needed to characterize ecological values in that domain.
A federal expert invited to the workshop noted that the U.S. Forest Service uses
social-psychological approaches as part of a major effort to understand communities. He
expressed the view that it would be useful for the Forest Service to share both "success
and horror stories."
Dr. Terry Daniel, one of the breakout leaders, made special note that social-
psychological methods are useful to discriminate responses by groups. They can show
where there are differences and convergence in values across groups so that decision-
makers can develop effective policies.
Socio-Psychological
Methods Breakout
54
Possible Applications at EPA
"Elusive" non-market values
Early: ID values to analyze, public buy-in for
expert analysis
ID needs for education about ecological
impacts
So public feels concerns are heard
Combine with other methodsto supplement
and validate results
Useful at regional level
Can measure preferences over policies AND
outcomes - explore "means" as well as
"ends"
-------
Additional Soc/Psych Methods
Research documenting public health,
psychological, and public safety benefits of
ecological amenities
Complex systems analysis for emergent
behavior
Recent British "Orange Book" - Standard
Multi-Attribute Model for Managing Risks to
the Public
Cultural risk metrics adapted from USFS
Monitoring news coverage of ecological
resources
Barriers/Possible Strategies
Representative results?
- Use techniques that ensure representative
samples
EPA's lack of non-economist social
scientists
- Possible strategy: ?
OMB review a major hurdle
- Plan ahead for survey needs
- Partner with other agencies (NOAA, USFS)
Issues
What Soc/Psych methods offer that
economic methods don't
-People resist thinking about tradeoffs
for some values
-Preference ratings can express
information that trade-offs can't (e.g.,
rationale for values) or gather
information in more understandable
ways
lssues-2
How can Soc/Psych methods establish the
importance of bio-physical impacts that can't
be monetized? Metrics?
Use value-of-information approach to identify
when additional ecological valuation is
needed for a decision?
Useful to identify when a decision needs
public value info vs. when expert input is
enough
Key: learn the mental models of populations
to be studied
8.5. Spatial Representation of Biodiversity and Conservation Values and Ecological
Services
Dr. Robert Johnston reported on the breakout-group focused on spatial
representation of biodiversity and conservation values and ecological services (slides
summarizing the breakout discussion are included at the end of this section of the
workshop report; breakout materials and a related presentation may be found on page 125
of this workshop report). He quickly summarized the presentations made by Drs. James
Boyd and Dennis Grossman. He noted common presentation themes emphasizing
standardized, transparent, spatially explicit information and models for use by policy
makers. He reported that breakout discussions touched on issues of simplicity vs.
richness; making assumptions underlying models clear; the importance and difficulty of
communicating effectively across disciplines; and making appropriate scientific use of
existing spatial data. He noted that workshop participants agreed on priorities and the
importance of multi-disciplinary collaboration. In his view, the most important keys to
successful use of such spatial approaches were transparency, consistency,
standardization, and clarity. Dr. Johnston's presentation slides are included at the end of
55
-------
this section of the workshop report.
A brief question and answer session followed. The first question concerned
whether the breakout group discussed whether the data presented were normative or
positive and the appropriate use of those different kinds of spatial data in Agency
decision-making. Dr. Johnston responded that there was no explicit discussion related to
this question. A workshop participant asked whether issues of confidentiality arose in
discussion, because the issue of confidentiality of data arose when Dr. Grossman
attempted to link models. Dr. Grossman responded that the NatureServe model does
contain confidential information about the specific location of some endangered species.
Another workshop participant noted recent research by Robert Dodds that
surprisingly suggests that weights are not as important as previously thought in technical
assessment schemes.
Dr. Boyd concluded the discussion with a comment that increased clarity about
the nature of ecosystem services was necessary, so that both ecologists and economists
can more transparently identify and count ecosystem services.
Breakout Session Report:
Spatial Representation of Biodiversity,
Conservation Values and Ecological
Services
Robert J. Johnston
Department of Agricultural and Resource Economics
University of Connecticut
US EPA Science Advisory Board Workshop: Science for
Valuation of EPA's Ecological Protection Decisions and
Programs. December 13-14, 2005
Seeds of Discussion
~ Two Presentations:
- James Boyd, Accounting for Ecosystem Services:
Spatial Units & Measurement
~Need for spatially explicit, standardized definitions
of ecological services, benefit indicators, and
improve value estimation.
Conservation Value
databases and linked models, and potential use for
56
-------
Common Presentation Themes
~ Standardized, transparent and spatially explicit
information and models that can:
- Offer a source of information to policymakers,
independent of any subsequent valuation efforts.
- Serve as a standardized starting point for more
- Promote more rigorous, standardized economic and
non-economic valuation efforts.
workidentify systematic elements of value.
m
Challenge #1: Simplicity vs. Richness
~ Despite shortcomings, reducing benefits to a
single metric (monetary valuation) has advantage
of formality, simplicity, and clear acceptance in
the policy process.
economic valuation or ways of improving
economic valuation?
the former.
Challenge #2: What are the Black Boxes?
~ Issues taken for granted or suppressed by some
disciplinary approaches are exactly those
considered most important by others.
M Examples:
I - How are weights defined when comparing/aggregating
- What are intermediate inputs versus final valued services?
Challenge #3: Playing Nicely with Others
~ Communication and convergence between
disciplines is critical and sometimes lacking.
Examples:
- What assumptions are implicit in maps of "ecological
- To what extent are certain types of ecological services
- What is a "value"? An ecological "service"?
implied by anthropocentric valuation? By non-
Challenge #4: Using What is Out There
~ Lack of awareness of spatially explicit databases
that can, at the very least, serve as an input to
valuation.
- These data can be better utilized by ecologists, economists
^aiK^therc^^
I - Availability of data does not imply availability of models,
capacity there on place-specific basis to support valuation?
Challenge #5: The Usual Suspects
~ Anthropocentric versus non-anthropocentric
approaches to ecological value.
M What ecological services are subject to economic
tradeoffs, substitution and prioritization?
I
57
-------
Significant Convergence
~ More promising to focus on complementary
approaches to ecological value rather than substitute
approaches.
M Estimated values should provide a basis for
~ We can't do a "core dump" on policymakersthere
~ Sine qua non: Need for transparency, consistency,
standardization and clarity from all disciplines.
'jigiiifioat]I: Coiiv6rg6ij06
~ benefits of ^llghorajjm muItjdjSCjjjljjjSjy work hg.ys
not been sufficiently realized or exploredf.hic is an
ubstantial promise.
msaasmmi
Need to standardize communication, units <
measurement, arid reporting.
lity anc
issues, and have
not been addressed sufficiently.
Need to clearly communicate what is providing
ices, how services are defined, and who is
alizing benefits.
58
-------
9. Panel Discussion: Experts' Feedback on Valuation Methods
A member of the C-VPESS, Dr. Harold Mooney, introduced the panel of four
experts who had been asked to respond to three questions:
1. Given the C-VPESS call for EPA to expand valuation to include a wide suite
of ecological values, how well do the methods discussed at this workshop
capture the range of methods currently available and in development?
2. What methods seem most practical and implementable for use by EPA in
characterizing or measuring values not reflected in traditional markets?
3. If you were to choose two topics for specific attention by the C-VPESS and
the SAB to provide advice to help EPA expand valuation efforts, what would
they be?
59
-------
Dr. Trudy Cameron, Raymond F. Mikesell Professor of Environmental and
Resource Economics at the University of Oregon, was the first to present remarks. She
addressed the three questions above in a slide presentation, captured in the three points
immediately below:
1. How well do the methods discussed here capture the range of methods currently
available and in development?
Pretty well
2. Which seem most practical and implementable for non-market ecosystem
services?
Available data will determine preferred methods within economic toolkit
(prefer observed choices; make do with stated preferences)
Deliberative processes: very helpful for scoping out feasible and
potentially attractive policy options, and for "first cuts" concerning
relevant versus irrelevant attributes of options for affected populations;
will not yield info on average tradeoffs in population as a whole (non-
representative)
Energy and materials flow analysis: helpful in capturing the constraints
faced by society, but not preferences (demand; benefits)
To a certain extent, you get what you pay for...
Different alternatives for valuation have different attributes: cost, quality
of information. Agency will have to make tradeoffs. "Best" method for a
particular context maximizes net benefits (total benefits from the valuation
information obtainedi.e. "better policy decisions"minus the total costs
of arriving at it).
3. Two topics for specific attention?
Ecological production functions - how ecosystem properties contribute to
ecosystem services that we, as a society, perceive and care about (either
for our own instrumental uses, or for their "intrinsic" value)
Misinformation produces bad choices which imply incorrect valuations.
Perfectly informed constituency is an impossible goal. Elicit subjective
information sets, even if "wrong," so that it is possible to use models to
"back out" repaired valuations under simulated "correct" information.
Dr. Cameron then made eight additional points, clarifications about "valuation"
that she did not see mentioned in C-VPESS draft reports. Text from her presentation
slides appears below.
1. Non-economists often use: "economic" versus "non-economic" ecosystem
services to describe what economists consider to be: "commercially exploited" versus
"non-commercial" ecosystem services
60
-------
... Any problem about how to allocate scarce resources (here, ecosystems)
to different uses (commercial exploitation, development, preservation,
etc.) is an "economic problem." The economics tent is huge. Economics
is not just Alan Greenspan (or Ben Bernanke). It is the "study of how to
allocate scarce resources across competing end uses" (SOHTASRACEU?)
"Resources" can be anything, not simply money.
Beyond the term "economics," there are many other common words that
economists recycle and use as technical terms. We should have invented
new jargon, as other disciplines sometimes do, but we haven't. Makes
things very confusing for outsiders who think they know what we mean
when we use words like "cost" or "capital," for example.
2. Beyond valuation? i.e., The case of an ecosystem service that it is "impossible to
put a dollar value upon"?
Property rights and refusal to contemplate WTP:
Some people may be unwilling to think about willingness to pay to
preserve or enhance an ecosystem service because this implies that they do
not have an inalienable "property right" to those services. Need to get
people to imagine "IF you did not have a right to this ecosystem service,
how much would you be willing to give up to preserve or enhance it?"
A very practical question, but some people won't play this game.
May refuse to contemplate tradeoffs because it is "not fair" that they
should be asked to make such a tradeoff, perhaps because they have never
had to do it before and it makes them uncomfortable to think about it. The
"invaluable 41-year marriage." May have the luxury of viewing a
marriage as beyond valuing until you have had to trade off against itas
in "your career versus your marriage."
Social stigma and refusal to contemplate WTA:
If someone does have a property right to something (e.g. a clean river),
then the correct measure of social value is "compensation demanded" or
"willingness to accept (compensation to give up that right)". But some
people will refuse to play that hypothetical game as well, even if they
would (privately and anonymously) be willing to take some finite amount
of compensation. Reason? They do not wish to incur the social stigma
associated with "selling out" (a term that is pejorative in itself). The
amount of compensation you would have to give them would need to be
enough to make up for not only for the loss of the clean water, but also for
the loss of their reputation for solidarity with their community. The value
of a good depends upon the availability of substitutes, among other things.
There may be substitutes for a clean river, but no substitute for their
community standing and reputation (their "good name").
61
-------
Empirical problem: can't distinguish between the amounts of
compensation needed for each thingthe resource and their good name.
3. Economists don't usually question the reasons why people are willing to give up
other things to preserve or enhance ecosystem services. All that matters: the fact that
they ARE willing, and the extent to which they are.
This doesn't mean we are not curious about things that might seem to
explain variations in WTP. We often explore how our WTP estimates
seem to vary systematically with observable individual attributes (gender,
age, ethnicity), or with measures of attitudes or reported behaviors from
the same people ("How well-informed are you about environmental
issues?" "How often do you go fishing for sport?") When we do this, we
are looking for logical consistency between estimated WTP and things that
intuition suggests should be correlated with it. Or, we are seeking to
forecast how average WTP might differ across sub-populations with
different characteristics.
4. There is a difference between what is (a "positive" question) and what should be
(a "normative" question). Economists study the tradeoff decisions that people do make,
conditional on their characteristics and the attributes they perceive for the alternatives
they are considering.
Rich enough empirically estimated economic models of choice can be
used to simulate, counterfactually, what choices people would probably
make if their characteristics were different along the same dimensions (e.g.
age) or if the alternatives they face had different levels of the same
attributes (= positive analysis).
Without data on the choices people make, no theoretical economic model
will tell you what choices society should make.
Best we can do: Conditional on knowledge of preferences, we can point
to the alternative that is likely to be considered most desirable,
But: economic models cannot tell us what preferences should be.
5. Important role for environmental advocacy: to help people's perceptions match
the scientific facts about ecosystems services.
6. An unfortunate but common misconception about economists' motives:
18th century definition of "wealth" = well-being (e.g. the commonweal)
.. .the opposite of "illth" (mentioned even in a 1915 intro textbook)
i.e. Adam Smith's (1776) "The Wealth of Nations" was concerned
with the "well-being of nations"...
62
-------
21st century definition of "wealth" = stock of financial assets
.. .may or may not be an indicator of "wealth" in the 18th century
sense
Due to this evolution in common usage, the idea of "maximizing wealth"
now sounds money-oriented and just plain greedy. That's why economists
now call it "maximizing social welfare."
7. Misconceptions about economists' notions of "value"particularly when it is
characterized as "monetization" (as an epithet). After all, moneychangers have had a bad
reputation for a couple of thousand years.
"Value" is another common word recycled by economists to serve as a
technical term. Here, it refers to a particular "marginal rate of
substitution": i.e. how much of "all other goods" would you be willing to
give up, to get one more unit of the environmental good in question?
The composite commodity we call "all other goods" (AOG) is measured
in convenient units such that one unit costs $1. This is where the
"monetization" step occurs. We can then use dollars as a measure of the
quantity of "all other goods."
If you had to pay for the environmental good, but you chose none of it,
you could afford a number of units of "all other goods" equal to your
number of dollars of income. In consumer theory, we view money income
simply as a measure of the quantity of other (market) goods you can
consume, given your budget constraint.
"Value" of the environmental good is derived from tradeoffs willingly
made between quantities of the environmental good ("envgood") and
quantities of all other goods ("AOG")
Suppose utility depends on the marginal utility of each thing, times the
quantity of that thing (in the case of a simple linear utility function):
Utility = Utility _ + ^Utility ^ enVgGO(j + (other terms?)
A $AOG Aenvgood
To hold utility constant (A Utility = 0 ) when we increase the
environmental good by one unit, by how much could we decrease
consumption of all other goods (income)? Solve this equation for
/S$AOG-
63
-------
0 = AUtility =
AUtility
A %AOG
[A$^OG] +
I [ ?
marginal
utility of
AOG (=income)
AUtility
Aenvgood
¦ [Aenvgood ] + 0
] } [
marginal
utility of
envgood
]
Economists don't dictate the sizes of these marginal utilitiesthey have to
attempt to measure them by studying the tradeoffs people are willing to
make. Here, the linear utility function implies that marginal utilities are
constant (an oversimplification in most cases). This model implies that
"value"= "willingness to pay" is also a constant, given by:
A %AOG
r AUtility ^
Aenvgood
[i]
AUtility
A$AOG J = «_ rajj0 marginal utilities'
Note that units of Utility cancel (fortunately) and result is in
"dollars-worth of AOG willingly given up, for one more unit of
envgood"
Note that this is a quantity of AOG, given current prices. Money is just a
convenient metric for the quantity of this composite good. Economic
"values" are about tradeoffs (substitutions) people are willing to make.
or (more tersely) "willingness to pay, in dollars
envgood
per unit of
Generalizing the valuation approach:
The ratio of marginal utilities is called the "marginal rate of substitution"
between the environmental good and all other goods (income).
^ "Marginal utility of envgood" ^
A %AOG =
"Marginal utility of income"
More-interesting and more-realistic utility functions are not simply linear
in their arguments (the quantities of each good). They are likely to display
diminishing marginal utility (DMU), such that the extra utility
from and extra unit declines with additional units, and
diminishing marginal rates of substitution, such that the amount of
another good that you are willing to give up to get one more unit of good
64
-------
X declines, the more you already have of good X.
Implication: "value" of one unit of envgood will depend on how much
you are currently consuming of this good and other goods (income). It is
also likely to differ across people, because preferences differ.
8. Energy Theory of Value
The "energy requirements" per unit of output is a dimension of the
constraints faced by society.
The "production opportunity set" describes all the possible combinations
of goods and services we can enjoy. It is defined by the quantities of each
available resource, and the technologies available to convert these
resources into things that we want or need.
If "resources" are heterogeneous, it is easy to argue that the production
possibility frontier is "bowed outwards" (has an increasingly negative
slope from left to right) the more of any one good you try to produce,
the more costly each unit becomes in terms of other things you can no
longer produce.
If we can reduce all resources to equivalent and homogeneous units of
"emergy" (solar radiation), then the production possibility frontier has a
constant (negative) slope. Relative emergy requirements per unit of each
possible good we could produce define the cost of good A in terms of the
units of other goods that would need to be foregone in order for us to
produce that one unit of good A.
Relative emergy requirements for production define "What we HAVE to
give up in terms of other goods to produce another unit of A."
This is called a marginal rate of transformation in production.
Relevant concept for valuation is "What we ARE WILLING to give up in
terms of other goods to consume another unit of A."
This is defined by preferences over the different possible goods we
could produce (the marginal rate of substitution in consumption).
Only at the "optimal" allocation are these two amounts equal. For non-
marketed goods such as ecosystem services, we do not observe
competitive equilibrium. Instead, it is our task to try to figure out where it
might be, so we get some idea of how much to provide.
We need to know BOTH
a. marginal rates of transformation (which the idea of emergy might
be helpful in identifying), and
b. marginal rates of substitution (based on some type of aggregation
65
-------
of individual preferences... a social welfare function)
In the question-and-answer session that followed her presentation, a workshop
participant asked whether framing questions in terms of dollars distorts respondents'
answers. Dr. Cameron responded that economists can use other metrics as a numeraire.
Any pair-wise trade-off can be used. Monetary estimates can be "backed in" at a later
time. In response to another question, she affirmed that economics is the study of choices
among competing uses. Another participant asked whether questions about ecological
value should be framed "absent reference to rights." Dr. Cameron suggested that
researchers can reframe questions when people refuse to indicate choices. She was asked
whether there were some domains where it may be inappropriate as a society to build
utility functions. Another workshop participant commented that any time society makes
a choice, society has revealed values about alternatives. Dr. Cameron concluded the
discussion with the comment that the task of economists are to reveal people's choices.
Once that's done, it is possible to study whether those choices relate to government's
choices.
Dr. Jon Krosnick, Frederic O. Glover Professor in Humanities & Social Sciences
& Professor of Communication, Political Science, and Psychology, Stanford University,
was the second panelist to speak and expressed his appreciation for the workshop and
work of the C-VPESS. He presented several slides to demonstrate that response rates are
not a valid indicator of survey accuracy. He emphasized that reliability depends on
surveys being administered to a representative sample and designed well (so that they
controlled for recall errors, comprehension errors, reporting errors, intentional
omission/addition, and nonresponse).
Internet surveys can be as reliable as telephone surveys and can be less costly to
conduct. He emphasized that the general public informs policy-making via social science
measurement and cited numerous examples of continuing surveys conducted by other
federal agencies. He then provided a variety of examples showing the accuracy of social
science measurements that are well conducted. He noted that contingent valuation
surveys can also have high validity and reliability. Among social science methods, he
cautioned against data collection in groups, as opposed to collecting data from
individuals, and cautioned also against collection technique that emphasize introspection.
He characterized all methods as having value and urged EPA to explore and evaluate
them all. Dr. Krosnick was only able to take one question. It concerned whether
response to surveys differ by the degree to which related questions involve controversy.
Dr. Krosnick responded that surveys can be framed well to deal with controversial issues.
Dr. Krosnick's slides appear below.
66
-------
Me 20 years ago:
Deeply skeptical
-About survey measurement of values
-About CV
Me today:
Deeply skeptical but...
- Impressed by how well surveys work.
(if designed and implemented well)
- Impressed by how well CV works.
EPA is not Alone
The general public
informs policy-making
via social science measurement
Examples of Continuing Federal Surveys
~ Survey of Income and Program Participation (Census Bureau) 1984 -
~ Consumer Expenditure Surveys (Census Bureau) 1968 -
~ Annual Housing Surveys (Census Bureau) 1973 -
~ Survey of Consumer Attitudes (NSF) 1953-
~ Health and Nutrition Examination Surveys (NCHS) 1959 -
~ National Health Interview Surveys (NCHS) 1970 -
~ American National Election Studies (NSF) 1948-
~ Panel Study of Income Dynamics (NSF) 1968 -
~ National Longitudinal Surveys (BLS) 1964 -
~ Behavioral Risk Factor Surveillance System (CDC) 1984 -
~ Monitoring the Future (NIDA) 1975-
~ American national Election Studies (NSF) 1948-
~ General Social Survey (NSF) 1972-
~ National Crime Victimization Survey (DOJ) 1973-
~ Continuing Survey of Food Intake by Individuals (USDA) 1965-
~ National Survey of Distracted and Drowsy Driving (NHTSA)
Potential Threats to Accuracy
* Recall errors
» Comprehension errors
* Reporting errors
* Intentional omission/addition
* Nonresponse
How Accurate is
Social Science Measurement?
67
-------
100%
50%
|Gallup Estimate Actual Vote|
Gallup final survey estimates of support for winning presidential candidate
correlated with actual election outcome, r = .85.
Home Buying Attitudes vs.
Sales of Single Family Homes
Consumer Buying Altitudes Actual Sa&s
100
ILk a."
5
1
fill tn rt .tfh-
X
3
f 50
4 i
5
lAt Jf iA u/ M/
t
e eo
!K\} y\r "
3
3 i.
1
f It \ m
f
40
? i M
2
1970 1975 1980 1985 1990 1995
Survey estimates of home buying attitudes correlated with actual home sales,
time series correlation = .77.
8
Buying Conditions for Cars vs.
Total Light Vehicle Sales
Unemployment Expectations vs.
Changes in the Unemployment Rate
-Consumer Buying AH*u
-------
Expected Annual Inflation Rate vs.
Changes in the Consumer Price Index
(4-Quartef moving average*)
14
Conwmwr EareUtfiont ~~ Actual Ctoiw
A
14%
1 «
a h\
t2%
I 10
i
1 8
AA
X0% g
-!
? *
aA f r Vifc. />a
I A
111
*%
2
2%
tses 1970 1975 1 980 1994 1990 1995
Survey estimates of inflation expectations correlated with actual changes in
the Consumer Price Index, time series correlation = .90. 13
Accuracy of Face-to-Face,
Telephone, and Internet Surveys?
Never Married
60% --
40% -
Married
100%
80% f
60%
40%
20% f
0%
100%
80% -
60% -
40%
20% -
Divorced
10%
Two People in Household
60% --
69
-------
Three People in Household
100%
0%
1 S%-
_ _ JZ%
18%
Four People in Household
100%
60% --
40% --
20% -
Children Present in Household
100%
80% -
60% -
40% -
20% -
Homeowners
60% -
40% --
20% --
40% --
20% --
Two Bedrooms in House
100%
80% -
60% -
40% -
20% -
Census 2000
70
-------
Three Bedrooms in House
39%
46%
46%
- - 1 1
I--
Census 2000 RDD KN
Four Bedrooms in House
- -15%"
20%
-16%-
Census 2000 RDD KN
Four Bedrooms in House
- -15%
20%
-16%-
Census 2000 RDD KN
One Vehicle
Census 2000
Two Vehicles
Census 2000
Impact of Response Rates
on Accuracy:
7% vs. 70%
71
-------
'erage Absolute
Discrepancy
J 03 CO O
Average Absolute Gender Discrepancy
<
Low High
Response Rates
30
Average Absolute Race Discrepancy
Response Rates
31
Average Absolute Age Discrepancy
Response Rates
High
32
A\erage Absolute Education Discrepancy
Response Rates
Average Absolute Education Discrepancy
10 -i
O o
Low High
Response Rates
Validity of CV Measurements
35
72
-------
Results: 4 vs. 2 Species
Price
4 Species
2 Species
$10
50.3%
35.6%
$25
39.8%
24.7%
$80
29.9%
17.9%
$140
23.6%
14.9%
$215
18.6%
10.7%
Mean
$63
$34
% Yes
37%
21%
Regression Predicting WTP
Price (log) -.40**
Income (if <$35,173; log) .17**
I ncome (if > $35K, < $150K; log) .15**
I ncome (if > $150K; log) .11*
(continued ...)
+p<10 *p<05 **p<01
Regression Predicting WTP
Extremely important to .15*
protect coastal areas
Decrease government spending on -.27*
endangered wildlife
Strong environmentalist .24**
(continued ...)
Regression Predicting WTP
Natural recovery will take longer
.53**
Natural recovery will be quicker
2g**
Program will be completely or
.60**
mostly effective
Program will not be effective
-1.26**
(continued ...)
39
Regression Predicting WTP
Thinks special tax will be for more
than one year -.28**
No confidence in the State of -.21*
California
Opposes increased spending on -.32**
government programs
(continued ...)
Regression Predicting WTP
Participates in saltwater boating or
22**
fishing or goes to the beach
Birdwatcher
.18*
Often watches TV shows about
.19**
animals
Household often eats fish
.18*
Lives in Los Angeles or Orange Co.
.17*
Lives North of San Francisco
-.25*
73
-------
Reliability of CV Measurement
52% yes in 1991
53% yes in 1993
42
Interim Conclusions
Social science measurements appear to
be reliable.
Social science measurements appear to
be valid.
Methodological mistakes can easily be
made to compromise the reliability and
validity.
43
EPA (finally)
Organization
of Measurement Techniques
Surveys
Focus Groups
Narrative discussions
Observation of behavior
Games
Distinctions in Data Collection
Methods
Sampling
- Representative
- Non-representative
Measurement vs. Experimentation
- Experiment within subjects
- Experiment between subjects
React to goods vs. Introspection
- Lots of work cautions about introspection
Respondents alone vs. in groups
Distinctions in Data Collection
Methods
Provide information?
- Provide information about the good only
- Provide other background info as well
Choices between ...
- Problems to solve
- Solution options for a single problem
Evaluations of...
- Problem seriousness
- Solution options
74
-------
Distinctions in Data Collection
Methods
Measurement metric
- Money
- Non-monetary
Mode of presentation
-Verbal
-Visual
All methods have value
Choice:
-An empirical issue
-Not intuitive prejudices
49
Be careful when interpreting
past work on measurement validity
Poor design undermines a study's value
EPA is lucky
- Can tell respondents it is a federal agency
seeking public guidance.
A half-page questionnaire handed to
people walking into a science museum will
not elicit the same respondent motivation
and decision quality.
What should EPA do?
Be brave!
- Don't choose among methods now.
- Commission an ambitious and thorough
methodological review.
New empirical comparison of methods in EPA
contexts.
-Work with other federal agencies.
Dr. Mark Schwartz, Professor in the Department of Environmental Science and
Policy at the University of California-Davis, acknowledged the important charge of
understanding more fully the value of protection ecological systems and services. He
linked the advisory effort to EPA's interest in broadening its mission to think about
ecosystem protection. He noted that the C-VPESS membership contained the appropriate
multi-disciplinary mix and range of ecological expertise to address the issue of ecological
valuation seriously. He agreed with Dr. Roughgarden on that ecological science can
provide meaningful production functions to help the Agency make ecological protection
decisions.
He suggested that an important priority would be to explore the potential of the
Long-Term Ecological Research Program and National Ecological Observatory Networks
(NEON) for predicting ecological outcomes. He also noted that from his perspective the
number and range of methods presented is almost "dizzy-ing." The challenge will be to
distill methods into a clear, coherent picture that the Agency can understand and use. He
suggested that the Sokal and Rohlf publication Biometry includes a diagram that provides
one way to represent data or methods for different types of applications that might be a
75
-------
model for the C-VPESS. He also noted that priority attention should be given to linking
ecological data and methods appropriate to different levels and types of decisions and to
increasing model integration. He observed that there ought to be clear guideposts
provided by facets of scale and attributes of EPA concern . For example, when doing an
economic valuation of clean water for national EPA policies, monetization via several
methods seems an obvious sort of recommendation. Similarly, assessing benefits for
attributes where people clearly perceive the amenity value (e.g., for a lake) lends itself to
some methods, while assessing other benefits (e.g., the contribution of the lake to
filtering pollutants out of downstream water bodies) might require expert inputs
regarding the magnitude of the effects and the consequences. He noted that even were
"public value" assessed, this may not reflect the same understanding of "value" as the
general understanding of the value of the housing industry in the U.S. economy.
He observed that ecosystem transfer function values may, or may not integrate
well at national scales, since some values may differ depending on local conditions, so
that either some average value must be assumed (benefits transfer), or ignored
(undervaluation problem). In either case, it would seem a worthwhile exercise for the
panel to enumerate the kinds of problems faced by the EPA and make specific
recommendations regarding approaches. This appears likely to be a difficult task given
some fundamental and philosophical disagreements regarding monetization and
valuation. Nonetheless, constrained by problems where EPA is required to provide a
cost-benefit analysis seems to help resolve some of these issues.
Mr. James Laity, Policy Analyst, Office of Information and Regulatory Affairs,
Office of Management and Budget, was the final speaker on the expert panel. He
thanked the SAB and EPA for taking on the issues discussed at the workshop and for
bringing them to the attention of the Office of Management and Budget. He noted that he
was speaking as an individual and not for the Office of Management and Budget.
The starting point of his remarks was his assumption that the goal of an eco-
valuation exercise was to provide useful, objective information for decision-making.
Therefore, in his view, valuation should be a descriptive, not normative process. He
identified himself as explicitly anthropocentric in orientation and posited that those who
advocate a different position were "jockeying for advantage."
In the context of monetized valuation, he personally advocated use of revealed
preference approaches, but he noted that OMB officially had endorsed the use of stated
preference approaches.
He urged the C-VPESS to look at best practices related to cost-effectiveness
metrics that would be useful for eco-valuation. He called for standards for non-
monetized metrics that could be used when it is difficult to monetize values. He
emphasized the need for standardization in this area that could make bio-physical metrics
objective, measurable, and verifiable.
Where values can neither be monetized nor quantified, he emphasized that the
qualitative information provided be descriptive, not normative, to express society's
76
-------
preferences, so that policy makers can make risk management decisions. He noted that
most environmental protection decisions are incremental in nature and do not raise issues
of absolute rights.
He noted that such terms as "educating the public, "conservative assessments"
(rather than use of central tendencies), and the "precautionary principle" do not have a
role in the debate over eco-valuation.
As scientists discuss and clarify the nature of ecological services, he wondered if
the benefits associated with some EPA programs may not be as high as some might
argue. He suggested that the Office of Water's technology-driven statutes, such as
effluent guidelines, might not be associated with significant ecological benefits.
He provided a few comments related to his own critical view of stated preference
studies. In his view, the budget constraint intended as part of stated preference surveys is
not perceived as real by respondents. In addition, respondents don't have a "mental
universe" of the other possible environmental amenities that could be met by the
investment of interest in the stated preference surveys. He also expressed a strong view
that efforts to educate people as part of the stated preference survey effort distort the
results of such surveys.
At the conclusion of his remarks he addressed the issue of data gaps and
uncertainty in eco-valuation. He identified the need for criteria and standards for benefit
transfer for eco-valuation as an important priority area for the committee to address. He
suggested that approaches minimize expert elicitation and strike a balance between
complexity and simplicity to make the results of an eco-valuation comprehensible to
decision makers. Standardization is needed. If dollars are not the universal metric, then
some other measure is needed that can bridge the gap between increasingly rich data and
theory in ecological sciences and the type of monetized information preferred by decision
makers. He suggested that some mechanism like the Bureau of Environmental Statistics
could provide standardization, strike the balance between complexity and simplicity, be
independent of EPA, and provide transparent scientific information could be useful for
regulatory support and for program evaluation needs under the Government Performance
and Results Act and for the Program Assessment Rating Tool.
In the question and answer period following his remarks a workshop participant
noted that the National Academy of Sciences has also called for a separate agency for
environmental statistics. Another participant challenged Mr. Laity's conjecture that the
value of ecological services might be low; she suggested that "existence values" might be
so high that they might be difficult to accommodate within existing regulatory
assumptions. Mr. Laity responded that there is a need to look at empirical evidence, not
hypotheticals.
Another workshop participant asked about how regulatory decision making can
accommodate and adjust to changes in environmental values. He suggested that the C-
VPESS should keep in mind that environmental values in American society have changed
historically and can change in major ways in the future. The final comment from a
77
-------
workshop participant concerned the proper role of education as linked to eco-valuation
and suggested that this topic deserved additional attention.
78
-------
10. Summary and Next Steps
Dr. Buzz Thompson concluded the workshop with brief remarks thanking
workshop participants for the valuable, constructive discussions. He noted that the
workshop generally endorsed the integrated, expanded approach proposed by the C-
VPESS. He expressed his sense that the plenary and breakout discussions had
emphasized that C-VPESS advice be practical and consider the budget and resource
limitations of the Agency when making recommendations about valuation methods and
processes. He also noted workshop participants' suggestions for the committee to balance
complexity and simplicity in its recommendations and to include suggestions for possible
institutional reforms in its advice. He stated that the next steps for the C-VPESS were to
focus on reports focusing on methods and applications. The workshop discussions would
provide valuable insights to help guide those efforts.
79
-------
11. Agenda and List of Invited Participants
U. S. Environmental Protection Agency Science Advisory Board (SAB) Workshop
December 13-14, 2005
Science for Valuation of EPA's Ecological Protection Decisions and Programs
Horizon Ballroom, Ronald Reagan Building, Washington, DC
Purpose: The SAB Committee on Valuing the Protection of Ecological Systems and
Services (C-VPESS) was charged to assess Agency needs and the state of the art and
science of valuing protection of ecological systems and services and then to identify key
areas for improving knowledge, methodologies, practice, and research.
The purpose of the workshop is to discuss the initial work of the C-VPESS; to
provide an opportunity for advisors across the SAB, Advisory Council on Clean Air
Compliance Analysis (Council) and Clean Air Scientific Advisory Committee (CASAC)
to learn from each others' work related to ecological valuation; and to feature feedback
from the Agency and outside experts.
December 13, 2005
9:00-9:20 Welcome
Purpose of Workshop and Agenda
Overview
Dr. M. Granger Morgan, Chair, SAB
Dr. Barton H. (Buzz) Thompson, Jr.,
Chair, SAB C-VPESS
9:20-9:30
Introductory Remarks
Mr. Marcus Peacock, EPA Deputy
Administrator
9:30-10:10 Global View from the Perspective of the
Millennium Ecosystem Assessment
10:10-11:00 Introduction to C-VPESS Work on an
"Expanded and Integrated Approach" for
Valuing Ecological Protection
11:00-12:00 Panel Discussion with EPA Senior
Managers
Dr. Walter V. Reid, Stanford
University
Dr. Kathleen Segerson, C-VPESS Vice-
Chair
Dr. James Boyd, Moderator
Panelists:
Mr. Robert Brenner
Office of Air and Radiation
Ms. Kathleen Callahan
EPA Region 2
Dr. George Gray
Office of Research and Development
80
-------
Dr. Albeit McGartland
Office of Policy, Economics and
Innovation
Dr. Michael Shapiro
Office of Water
12:00-1:30
Lunch
1:30-2:30 Overview of Methods Being Considered
by C-VPESS
2:30-3:10 Addressing Uncertainty in Ecological
Valuation and Expert Elicitation
3:10-3:15 Charge to Breakout Sessions
Dr. Gregory Biddinger, Moderator
Dr. Terry Daniel
Dr. Stephen Polasky
Dr. William Ascher
Dr. Robert Costanza
Dr. Barton H. (Buzz) Thompson, Jr
3:15-3:30 Break
3:30-5:30 Break-out Sessions on Specific Methods Session Leaders
A. Economic Analysis and Ecological
Production Functions (Concourse Level:
Hemisphere A)
B. Group Expressions of Value:
Referenda; Citizen Juries (Concourse
Level: Meridian C)
Dr. Joan Roughgarden
Dr. Stephen Polasky
Dr. William Louis Ascher
Dr. Barton H. (Buzz) Thompson, Jr.
C. Deliberative Approaches for Modeling, Dr. Robert Costanza
Valuation, and Decision Making Dr. Joseph Arvai
(Concourse Level: Hemisphere B)
D. Social/Psychological Methods for
Ecosystem Values Assessments
(Concourse Level: Meridian E)
Dr. Terry Daniel
Dr. Kathleen Segerson
E. Spatial Representation of Biodiversity Dr. Dennis Grossman
and Conservation Values and Ecological Dr. James Boyd
Services (Concourse Level: Meridian D)
5:30
Adjourn
81
-------
December 14, 2005
8:30 Opening of Second Day of Workshop
8:30-10:30 Reports from Break out Sessions Dr. A. Myrick Freeman, Moderator
10:30-10:45 Break
10:45-11:45 Panel Discussion: Experts' Feedback on Dr. Harold A. Mooney, Moderator
Valuation Methods
11:45-12:00 Summary of Workshop Dr. Barton H. (Buzz) Thompson, Jr.
12:00 Adjourn
82
-------
U. S. Environmental Protection Agency Science Advisory Board (SAB) Workshop
December 13-14, 2005
Science for Valuation of EPA's Ecological Protection Decisions and Programs
Ronald Reagan Building, Washington, DC
List of Participants
Please note: asterisks (*) denote members of the SAB Committee on Valuing the
Protection of Ecological Systems and Services (C-VPESS)
83
-------
Dr. Anna Alberini
University of Maryland
aalberini@arec.umd.edu
Ms. Ashley Allen
EPA Office of Water
Office of Science and Technology
allen.ashley@epa.gov
Dr. Henry (Andy)Anderson
Wisconsin Division of Public Health
anderha@dhfs. state, wi.us
Dr. Thomas Armitage
SAB Staff Office
armitage.thomas@epa.gov
*Dr. Joseph Arvai
Michigan State University
sknkwrks@msu.edu
*Dr. William Louis Ascher
Claremont McKenna College
william.ascher@claremontmckenna.edu
Dr. Mark Bain
Cornell University
Mark.Bain@Cornell.edu
*Dr. Gregory Biddinger
ExxonMobil
gregory.r.biddinger@exxonmobil.com
*Dr. Ann Bostrom
Georgia Institute of Technology
ann.bostrom@pubpolicv.gatech.edu
Mr. Robert Brenner
EPA Office of Air and Radiation
brenner.rob@epa.gov
*Dr. James Boyd
Resources for the Future
bovd@rff.org
Dr. Thomas Brown
US Forest Service
tcbrown@fs.fed.us
Dr. James Bus
The Dow Chemical Company
ibus@dow.com
Dr. Gilles Bussod
New England Research
gbussod@ner.com
Dr. G. Allen Burton
Wright State University
all en. burton@ wri ght.edu
Dr. Dallas Burtraw
Resources for the Future
burtraw@rff.org
Ms. Kathleen Callahan
EPA Region 2
callahan.kathy@epa.gov
Dr. Trudy Ann Cameron
University of Oregon
cameron@darkwing.uoregon.edu
Mr. James Canavan
EPA Office of Solid Waste
canavan.james@epa.gov
Mr. David Chapman
Stratus Consulting Inc.
dchapman@stratusconsulting.com
Ms. Lauraine Chestnut
Stratus Consulting Inc.
lchestnut@stratusconsulting.com
Dr. Joel Corona
EPA Office of Water
corona.joel@epa.gov
Dr. Deborah Cory-Slechta
University of Medicine and Dentistry of
New Jersey and Rutgers State University
84
-------
dcs@eohsi. rutgers.edu
gernes.marie@epa.gov
*Dr. Robert Costanza
University of Vermont
rcostanz@zoo. uvm. edu
*Dr. Terry Daniel
University of Arizona
tdaniel@u. arizona.edu
Dr. Ricardo DeLeon
Metropolitan Water District of Southern
California
ricardo deleon@mwdh2o.com
Mr. James DeMocker
EPA Office of Air and Radiation
democker ,i im@epa. gov
Dr. Robert Doudrick
US Forest Service
rdoudrick@fs.fed.us
Dr. Glenn Farber
EPA Office of Policy, Economics, and
Innovation
farber.glenn@epa.gov
Dr. Baruch Fischhoff
Carnegie Mellon University
baruch@cmu.edu
Dr. Ann Fisher
Pennsylvania State University
fi sherann@p su. edu
*Dr. A. Myrick Freeman
Bowdoin College
rfreeman@bowdoin.edu
Dr. James Galloway
University of Virginia
ing@virginia.edu
Ms. Marie Gernes
SAB Staff Office
Dr. William H. Glaze
Oregon Health & Science University
glazeb@ohsu.edu
Dr. Domenico Grasso
University of Vermont
dgrasso@uvm. edu
Dr. George Gray
EPA Office of Research and Development
gray.george@epa.gov
*Dr. Dennis Grossman
Nature Serve
dgrossman@natureserve.org
Dr. Richard Haeuber
EPA Office of Air and Radiation
haeuber.richard@epa.gov
Mr. Josh Hall
EPA Office of Water
hall .j osh@epa. gov
Dr. Barbara Harper
Confederated Tribes of the Umatilla Indian
Reservation (CTUIR)
bharper@amerion.com
Mr. Keith Harrison
KGH Environmental, PLC
kgh environmental plc@vahoo.com
Mr. Erik Helm
EPA Office of Water
helm.erik@epa.gov
Dr. Rogene Henderson
Lovelace Respiratory Research Institute
rhenders@lrri.org
Dr. Julie Hewitt
EPA Office of Water
hewitt.iulie@epa.gov
85
-------
Dr. Phil Hopke
Clarkson University
hopkepk@clarkson.edu
Dr. Bruce Hull
Virginia Institute of Technology
hullrb@vt.edu
Dr. Richard Iovanna
EPA Office of Policy, Economics and
Innovation
iovanna.rich@epa.gov
Dr. Sabrina Ise-Lovell
EPA Office of Policy, Economics, and
Innovation
ise-lovell.sabrina@epa.gov
Dr. DeWitt John
Bowdoin College
dj ohn@b owdoin. edu
Dr. F. Reed Johnson
Research Triangle Institute
fri ohnson@rti. org
Dr. Robert Johnston
Connecticut Sea Grant Office
robert.johnston@uconn.edu
Dr. Byung Kim
Ford Motor Company
bkim@ford.com
Dr. Dennis King
University System of Maryland
dking@cbl.umces.edu
Dr. K. Jack Kooyoomjian
SAB Staff Office
kooyoomjian.jack@epa.gov
Dr. Jon Krosnick
Stanford University
krosni ck@ Stanford. edu
Mr. James Laity
OMB Office of Information and Regulatory
Affairs
j laity @omb. eop. gov
Dr. Linda Langner
US Forest Service
llangner@fs.fed.us
Dr. Robert Lee
EPA Office of Pollution Prevention and
Toxics
lee.robert@epa.gov
Dr. Anthony Maciorowski
SAB Staff Office
maciorowski.anthony@epa.gov
Mr. Lawrence Martin
EPA Office of Research and Development
martin.lawrence@epa.gov
Dr. Genevieve Matanoski
Johns Hopkins University
gmatanos@i hsph. edu
Dr. Michael J. McFarland
Utah State University
farlandm@msn.com
Dr. Albert McGartland
EPA Office of Policy, Economics, and
Innovation
mcgartland.al@epa.gov
Dr. Joseph S. Meyer
University of Wyoming
meyerj@uwyo.edu
Dr. Jana Milford
University of Colorado at Boulder
milford@colorado.edu
Mr. Tom Miller
SAB Staff Office
86
-------
miller.tom@epa.gov
*Dr. Harold Mooney
Stanford University
hmoonev@stanford.edu
Dr. M. Granger Morgan
Carnegie Mellon University
granger.morgan@andrew.cmu.edu
Dr. Thomas C. Mueller
University of Tennessee
tmueller@utk.edu
Dr. Mark Nechodom
US Forest Service
mnechodom@fs.fed.us
Mr. David Nicholas
EPA Office of Solid Waste and Emergency
Response
nicholas.david@epa.gov
Dr. Angela Nugent
SAB Staff Office
nugent.angela@epa.gov
Ms. Gillian Ockner
David Evans and Associates, Inc.
gcf@deainc.com
Dr. Rebecca Parkin
The George Washington University
parkinr@gwu.edu
*Dr. Stephen Polasky
University of Minnesota
spolasky@apec.umn.edu
Dr. John Powers
EPA Office of Water
powers.john@epa.gov
Dr. Walter V. Reid
Stanford University
wrei d@ Stanford. edu
Dr. David Rejeski
Woodrow Wilson International Center for
Scholars
rei eski dw@ wwi c. si. edu
*Dr. Holmes Rolston
Colorado State University
rol ston@l amar. col ostate. edu
*Dr. Joan Roughgarden
Stanford University
rough@pangea. stanford.edu
Dr. James Salzman
Duke University
salzman@law.duke.edu
Dr. Herbert Schroeder
US Forest Service
hschroeder@fs.fed.us
Dr. Mark Schwartz
University of California - Davis
m wschwartz@ucdavi s. edu
*Dr. Kathleen Segerson
University of Connecticut
segerson@uconn. edu
Dr. Suhair (Sue) Shallal
SAB Staff Office
shallal. suhair@epa.gov
Dr. Michael Shapiro
EPA Office of Water
shapiro.mike@epa.gov
Dr. Benjamin Simon
Department of the Interior
benjamin simon@ios.doi.gov
*Dr. V. Kerry Smith
North Carolina State University
kerry smith@ncsu.edu
87
-------
Dr. Holly Stallworth
SAB Staff Office
stallworth.holly@epa.gov
Dr. Elizabeth Strange
Stratus Consulting Inc.
estrange@stratusconsulting.com
Dr. Deborah Swackhamer
University of Minnesota
dswack@umn.edu
Dr. Thomas L. Theis
University of Illinois at Chicago
theist@uic.edu
Dr. Valerie Thomas
Georgia Institute of Technology
valerie.thomas@isve.gatech.edu
*Dr. Barton H. (Buzz) Thompson, Jr.
Stanford University
buzzt@law. Stanford, edu
Mr. Steve Thur
NOAA Damage Assessment Center
steven.thur@noaa.gov
Dr. Robert Twiss
University of California-Berkeley
twiss@rtasc.com
Ms. Amy Upgen
EPA Office of Water
upgen.amy@epa.gov
Dr. Vanessa Vu
SAB Staff Office
vu.vanessa@epa.gov
Dr. Ann Watkins
EPA Office of Air and Radiation
watkins.ann@epa.gov
Dr. Paul C. West
The Nature Conservancy
pwest@TNC.org
Dr. William Wheeler
EPA Office of Research and Development
wheeler.william@epa.gov
Ms. Kathleen White
SAB Staff Office
white.kathleen@epa.gov
Dr. T.J. Wyatt
EPA Office of Pesticide Programs
wyatt.tj@epa.gov
88
-------
12. Background Material and Presentations for Breakout Groups
12.1. Economic analysis and ecological production functions
Session Leaders:
Dr. Joan Roughgarden, Professor, Biological Sciences and Evolutionary Biology,
Stanford University, Herrin Laboratory, Stanford, CA
Dr. Stephen Polasky, Fesler-Lampert Professor of Ecological Economics, Department of
Applied Economics, University of Minnesota
Contents:
Bibliography in Ecological Economics developed by Joan Roughgarden
Citations related to breakout topic (copies available at breakout session)
1998, Roughgarden, J., Production functions from ecological populations: a survey with
emphasis on spatially explicit models. In: Tilman, D. and P. Kareiva, (Eds.)
Spatial Ecology: The Role of Space in Population Dynamics and Interspecific
Interactions, Princeton University Press, pp. 296317.
2003, Armsworth, P. and J. Roughgarden. The economic value of ecological stability.
Proc. Nat. Acad. (USA) 100:7147-7151.
2004, Taylor H., Gretchen C. Daily, Paul R. Ehrlich, and Charles D. Michener. 2004.
Economic value of tropical forest to coffee production. PNAS (34): 12579-12582.
2005, Polasky, Stephen, Erik Nelson, Eric Lonsdorf, Paul Fackler, and Anthony Starfield.
2005. Conserving species in a working landscape: land use with biological and
economic objectives. Ecological Applications 15:1387-1401.
Diagram (see diagram below depicting ecological models, outputs, services,
valuation tools and decision approaches)
Diagram (see diagram below depicting ecological models, outputs, services,
89
-------
Bibliography in Ecological Economics developed by Joan Roughgarden
2003, Armsworth, P. and J. Roughgarden. The economic value of ecological stability.
Proc. Nat. Acad. (USA) 100:71477151.
2001, Roughgarden, J. and P. Armsworth. Managing ecosystem services. In: Press, M.,
N. Huntly, and S. Levin, eds. Ecology: Achievement and Challenge, Blackwell
Science, pp. 337356.
2001, P. Armsworth and J. Roughgarden. An invitation to ecological economics, Trends
in Ecology and Evolution 16:229234.
2001, Roughgarden, J. Guide to diplomatic relations with economists. Bull. Ecol. Soc.
America 82:8588.
1998, Roughgarden, J., Production functions from ecological populations: a survey with
emphasis on spatially explicit models. In: Tilman, D. and P. Kareiva, (Eds.)
Spatial Ecology: The Role of Space in Population Dynamics and Interspecific
Interactions, Princeton University Press, pp. 296317.
1998, Roughgarden, J., How to manage fisheries. Ecological Applications, 8(1):S160
S164.
1997, Brown, G. and J. Roughgarden, A metapopulation model with private property and
a common pool. Ecological Economics, 22:6571.
1996, Roughgarden, J. and F. Smith, Why fisheries collapse and what to do about it.
Proc. Nat. Acad. Sci., (USA), 93:5078-5083
1995, Roughgarden, J., Can economics protect biodiversity? In: T. Swanson, (Ed.), The
Economics and Ecology of Biodiversity Decline. Cambridge University Press, pp.
149-156.
1995, Brown, G. and J. Roughgarden, An ecological economy: notes on harvest and
growth. In: Perrings, C., K.G. Maler, C. Folke, C.S. Holling and B.O. Jansson
(Eds.), Biodiversity Loss: Ecological and Economic Issues, Cambridge,
Cambridge University Press, pp. 150-189.
90
-------
ECOLOGICAL MODELS OUTPUTS SERVICES
Habitable.
en
z
O
I
o
o
I
o
ZD
Q
O
Cd
CL
_l
<
o
o
o
_l
o
o
LU
Individual
Level
Space
Foraging
Cultural
Harvest
Population
Dynamic
Biomass
Provisioning
Community-
Level
Species
Supporting
Biodiversity
Biogeochemical\
h unctions
Regulating
Ecosystem
Global-Change) Biosphere
Preservation
ECONOMIC TOOLS
VALUATION TECHNIQUES
Market
Market Price
Consumer Surplus
Non-Market
Revealed Preference
Stated Preference
Static Optimization
Deterministic
Stochastic
Dynamic Optimization
Deterministic
Stochastic
Closed-Loop
Open-Loop
OPTIMIZATION APPROACHES
91
-------
12.2. Group Expressions of Value: Referenda; Citizen Juries
Session Leaders:
Dr. William Louis Ascher, Donald C. McKenna Professor of Government and
Economics, Claremont McKenna College
Dr. Barton H. (Buzz) Thompson, Jr., Robert E. Paradise Professor of Natural
Resources Law and Vice Dean , Stanford Law School, Stanford University
Contents:
Valuation Methods Based on Referenda and Other Public Decisions
Citations Related to Referenda and Citizen Juries (copies available at breakout
session)
Kahn, M.E., and J.G. Matsusaka. 1997. Demand for environmental goods: Evidence from
voting patterns on California initiatives. Journal of Law and Economics 40:137-
173.
RK Blarney, RF James, R Smith and S Niemeyer. 2000. Citizens' Juries and
Environmental Value Assessment. http://cjp.anu.edu.au/docs/CJl.pdf.: The first in
a series of reports to be published containing the results of the research project
Citizens' Juries for Environmental Management.
Valuation Methods Based on Referenda and Other Public Decisions
Referendum votes and other formal public decisions provide the basis for a set of
valuation approaches that can provide monetized values, but use somewhat different logic
than that of the conventional individually based revealed-preference and stated-
preference methods. The outcomes of referenda (measures placed on the ballot by a
legislative body), initiatives (ballot measures proposed by citizens), or other official
public decisions directly express what the body politic as a collectivity values in terms of
policy outcomes. These expressions may or may not correspond closely to the
aggregated values of the individuals in the community in terms of outcomes. Referenda
approaches (not to be confused with the "referendum format" often used for posing
questions to solicit contingent valuation responses) provide information about the policy
preferences of the median voter; under certain circumstances this information can tell us
about the median voter's valuation of specific environmental amenities, and can even
provide information, albeit weaker, about mean valuations of those who participate in the
voting process.
Referenda and initiatives are formal solicitations to the public to determine the
public's willingness to pay. In a referendum or initiative, officials or policy activists
present voting choices that formally specify environmental objectives, such as reducing
air pollution, establishing a wildlife preserve, or building a storm run-off system. In
some cases, these objectives are clearly specified in quantitative terms: number of tons of
sulfur dioxide expected to be removed, number of acres of reserve, or reduction of the
area subject to flooding. The costs of achieving these objectives are specified in various
ways, ranging from the financial costs in taxes or bonds, to the restrictions that would be
92
-------
expected to impose opportunity costs such as reduced employment opportunities or
restricted resource extraction.
The logic of using formal public outcomes to infer how much "society values"
particular outcomes has been used primarily in the literature on health and safety. For
example, the value of a "statistical life" has been estimated by calculating how much
public policies commit to spend in order to reduce mortality rates from health or safety
risks, or, conversely, how much economic gain is associated with public decisions that
reduce safety (e.g., by examining official decisions of U.S. states to raise or lower speed
limits, Ashenfelter & Greenstone [2004] estimated the market value of the time saved by
getting to the destination more quickly, and from that estimated the value of the
additional expected traffic fatalities). The logic of making valuation inferences from
referenda and initiatives has been addressed in a few publications, most directly in
Deacon & Shapiro, 1975; and Shabman & Stephenson, 1996.
In addition to taking the valuation derived from the analysis of public decisions as
an input in itself, the analysis of public decisions, particularly referenda and initiatives,
can be used to validate the results of other valuation methods. Several studies have
compiled the results of initiatives and/or referenda in order to try to validate more
conventional valuation techniques, especially contingent valuation (Kahn & Matsusaka
(1997), List & Shogren (2001; 2002), Murphy et al. (2003), Polasky, Gainutdinova &
Kerkvliet. (1996), Schlapfer, Roschewitz, & Hanley (2004). Vossler & Kerkvliet
(2003)). As Arrow et al. (1993) recommend:
The referendum format offers one further advantage for CV. As we have argued,
external validation of elicited lost passive use values is usually impossible. There are
however real-life referenda. Some of them, at least, are decisions to purchase specific
public goods with defined payment mechanisms, e.g., an increase in property taxes. The
analogy with willingness to pay for avoidance or repair of environmental damage is far
from perfect but close enough that the ability of CV-like studies to predict the outcomes
of real-world referenda would be useful evidence on the validity of the CV method in
general. The test we envision is not an election poll of the usual type. Instead, using the
referendum format and providing the usual information to the respondents, a study
should ask whether they are willing to pay the average amount implied by the actual
referendum. The outcome of the CV-like study should be compared with that of the actual
referendum. The Panel thinks that studies of this kind should be pursued as a method of
validating and perhaps even calibrating applications of the CV method... (emphasi s
added)
In comparing the valuations yielded by stated-preference approaches with those
derived from public decisions, the studies typically show the inferences from public
decisions to yield lower valuesnot surprising in light of the absence of the hypothetical
element in the public-decision results. Although systematic comparisons with
conventional revealed preference approaches are lacking, it is likely that the valuations of
eco-system components calculated from public decisions would be higher, because public
decisions do capture whatever elements of public-regardedness are present among the
voters. The valuations based on public decisions have intrinsic validity within the
93
-------
paradigm that gives standing to the community votes as reflecting the policies that the
public prefers.
Direct Referendum/Initiative Analysis
The valuation analyst can chose to take the referendum choices as they are
formally specified, in which case a winning proposal can be interpreted as having
standing as the electorate's choice. For example, a municipal government may propose a
referendum measure to purchase and maintain 500 acres of currently unused land as a
forest reserve costing $1,000,000 annually for a community of 10,000 households.
Assume that the measure is not significantly entangled in controversies over how it will
be financed (e.g., there is no opposition that a bond measure would simply saddle future
generations). The measure passes by 51%. The value can be metricized in various ways;
e.g., as
$1,000,000 per annum for the 500 acres for the community
$2,000 per annum per acre for the community
$100 per annum for the 500 acres per household
$.20 per annum per acre per household.
94
-------
If the initiative or referendum passes by a slim majority, this valuation can be
considered to be quite close to the "community's" valuation. If the vote is more strongly
in favor, then the valuation represents a floor on the community's value of the eco-system
benefits. If the initiative or referendum loses by a slim majority, then (more arguably)
one could assert that the community's valuation is also close to the value implied by the
proposed measure.
If the outcome is not close (e.g., the initiative or referendum passes by 70%), the
inferred value is a floor on the community's value. This is because a higher cost may
have still gained a majority, albeit probably a narrower one.
However, the fact that a referendum or initiative fails to pass does not necessarily
mean that the inferred value is a ceiling on the community's value, because other issues,
such as how the measure is to be financed, may lead to the rejection of a measure that
otherwise would have been accepted. The results will be most easily interpreted if the
initiatives or referenda are: a) as focused as possible on a single dimension of
environmental protection or amenity; b) free of ideological debate; c) confined to easily
identifiable government costs rather than diffused and uncertain costs such as job losses.
Note that the approach does not primarily address the mean value of the
ecosystem improvement or protection. This is because the electorate's choice is not the
conventional utilitarian notion of the total value summed across all individuals who vote.
It is possible to determine a very modest floor on this aggregate value (and therefore on
the mean value) by attributing to the "yes" voters the value of the benefit-cost ratio
specified by the proposal, and a value of zero to all voters who opposed the proposal.
For example, in the case of the forest reserve proposal described above, if the proposal
had received a 70% "yes" vote, the minimum mean value would be $1,400 per annum per
acre for the community (i.e., .7 x $2,000 + .3 x 0).
Making valuation estimates directly from referendum or initiative outcomes has
two advantages over conventional valuation methods. Unlike the standard revealed-
preference approaches, such as hedonic pricing or the travel-cost method, voting on
referenda or initiatives will reflect as much (or as little) public-regardedness as the voters
actually hold toward the objectives involved. Standard revealed-preference approaches
reflect the private-utility-maximizing decisions of individuals who purchase homes,
spend money to visit parks, etc.; these decisions do not reflect what individuals want for
their communities. Voting affirmatively for referendum- or initiative-proposed public
expenditures do elicit valuing on behalf of the community, insofar as the voters are so
disposed. Of course, a voter may vote for or against a referendum or initiative proposal
strictly out of concerns for herself and/or her family, but the outcome does not exclude
the existence value component if it exists.
Unlike the conventional stated preference approaches such as contingent
valuation, the analysis based on referendum or initiative outcomes is not subject to the
possible distortions of hypothetically-posed choices. If a voter supports the referendum
or initiative proposal, the vote contributes to the likelihood that the expenditures will
95
-------
actually occur and the costs will actually be borne. Some might argue that the chance
that any one vote will decide the outcome of the referendum or initiative is remote, and
therefore the vote is more of a symbolic act than a tradeoff choice. However, there are
two important responses to this point. First, whatever the mix of motives of the voters,
the outcome is the community's decision, and therefore has standing in and of itself.
This is the same logic by which we accept elected officials as legitimate even if we are
dubious about the motives or rationality of the voters. Second, even if a voter believes
that the chances that his or her vote will make the difference are negligible, the vote is
still an expression of support or opposition to the proposal. There is little reason to
believe that a "yes" vote would reflect just the gratification of voting "yes" (especially in
secret balloting) rather than a belief that the proposal merits support.
The most useful referenda or initiatives would propose direct costs to the voters,
typically in the form of taxes, fees, or bonds to finance actions designed to improve or
protect eco-systems. Referenda or initiatives that entail restrictions on development
(such as more stringent emissions or effluent standards) are less useful, because of the
uncertainty of the level and incidence of the economic impacts. Similarly, in order to
isolate the values attributed to particular ecosystem benefits, referenda and initiatives that
address only one objective, such as preserving habitats or reducing air pollution. With
multiple objectives, the analysis cannot assign the willingness to pay to each component.
Similarly, if it is clear that a referendum or initiative entails additional partisan political
stakes (e.g., if it is widely viewed as a political test of a government official), the results
are less illuminating in terms of the ecosystem values that the voters hold.
Another concern that some would level against inferences based on referenda or
initiatives is that these votes are often subject to intense efforts by interest groups,
advocacy groups, and even governments to manipulate public perceptions. This concern
has two aspects: whether the information on which voters base their decisions has been
distorted, and whether the votes are swayed by appeals on one side or the other. The first
aspect is more compelling: we certainly would be less willing to accept the validity of an
estimate derived from voting decisions driven by serious misconceptions of the proposed
benefits and/or costs. The outcome is still the official decision of that community, but the
justification for using the result as the basis of benefits transfer to other communities
would be very weak. On the other hand, the fact that referenda and initiatives are often
subject to intensive campaigns of persuasion may be considered a virtue rather than a
drawback, insofar as it would provide more information on both sides. In addition, the
fact that individuals are exposed to efforts at persuasion is by no means confined to
referenda and initiative contests: respondents to contingent valuation surveys have of
course been subjected to many years of promotional activities by environmental groups;
people who travel farther to a particularly popular national park such as Yosemite have
been influenced by all sorts of communications extolling its virtues. In short, efforts at
value persuasion are pervasive, and in any event should not be a basis for rejecting the
significance of decisions of individuals exposed to those efforts. The philosophical basis
underlying the use of referenda or initiatives, namely that the public's preferences are
legitimately shaped by the political process, and that the public's policy preferences are
important beyond how the public values the outcomes that these policies may produce, is
quite different from the so-called "progressivist" position that individuals' values should
96
-------
be determined in isolation of "politics" (Sagoff 2004: 177-178).
Another difference in philosophical basis is that the referendum and initiative
results reflect intensity of attention to the issue, at least insofar as those who do not care
enough to vote are excluded from the analysis. From the progressivist, technocratic
perspective, everyone's values ought to be incorporated, because the policies ought to
maximize utility (i.e., the consequences of public decisions) regardless of whether
specific individuals are mobilized to take action. On the other hand, prominent strains of
pluralist democratic theory regard intensity as a fully legitimate factor in determining
policy outcomes (Lowi 1964).
One limitation of estimating values from referendum or initiative outcomes is that
it is sometimes difficult for voters to assess the actual stakes involved. The benefits will
often have to be predicted (e.g., how much biodiversity will be reserve really safeguard;
how much less flooding will the flood-control system actually prevent?), entailing a
certain amount of uncertainty. The benefits that do occur will often be community-wide,
with some uncertainty as to how much an individual or particular household can take
advantage of the benefits. On the cost side, the burden of a tax increase or bond measure
on household expenditures may be very difficult for the typical voter to estimate, and the
impacts of development restrictions may be even more difficult in light of the uncertainty
as to which families would ultimately be affected. Insofar as the costs specified by the
referendum or initiative are not easily translatable into household budget terms, the
outcome, though it is still "the community's decision," is less revealing about the values
held by the voters.
Referendum/Initiative Analysis Followed by a Survey
Therefore another variant that relies on referendum and initiative outcomes to
make willingness-to-pay estimates consists of combining the voting outcome with a
follow-up survey to determine the perceptions of the stakes. This variant amounts to a
hybrid of the first variant and the "referendum format" contingent valuation approach.
The floor of the willingness-to-pay value of the proposed eco-system improvements is
estimated by determining the voters' perceptions of the eco-system improvements and
costs proposed by a recent referendum or initiative. The respondents are asked whether
they voted, how they voted, and what they believed the benefits and costs of the proposal
were. As with Variant 1, if the initiative or referendum passes by a slim majority, this
valuation can be considered to be quite close to the median voter's valuation. If the
initiative or referendum loses by a slim majority, then (more arguably) one could assert
that the median voter's valuation is also close to the value implied by the proposed
measure. (Note: again, a losing initiative or referendum does not necessarily mean that
the inferred value is a ceiling on the median voter's value, because other issues may lead
to the rejection of a measure that otherwise would have been accepted.) As with Variant
1, the results will be most easily interpreted if the initiatives or referenda are: a) as
focused as possible on a single dimension of environmental protection or amenity; b) free
of ideological debate; c) confined to easily identifiable government costs rather than
diffused and uncertain costs such as job losses.
97
-------
If, in addition to asking how respondents voted and their perceptions of the
benefits and costs of the proposal, the randomly-sampled respondents who opposed the
proposal are asked what (lower) cost would have induced them to vote for the proposal,
and those who supported the proposal are asked how much more they would have been
willing to pay, this approach also permits an estimate of aggregate and mean values, just
as a standard contingent valuation study would, with less potential distortion arising from
respondents' desire to be regarded in a favorable light. Thus the survey following a
referendum or initiative can provide an internal cross-check of how much correspondence
there is between the stated-preference approaches and the referendum or initiative
findings.
It should be noted that in focusing on the benefits and costs that respondents
report, rather than the actual benefits and costs that the referendum or initiative proposal
specifies, the results do not reflect the community's formal decision. This is a significant
difference in the philosophy underlying the standing of the results. That is, the first
variant, even if it does not necessarily reflect the values that voters perceive, it does
represent what the voters have chosen. Different logics underlie their standing.
Direct Analysis of Public Decisions to Accept Pollution or Resource Depletion
While the approaches outlined above provide information about willingness to
pay, there are some public decisions that can provide inferences for willingness-to-accept
decisions. These decisions involve a community's vote as to whether to permit the entry
of a new firm or a new (or increased) economic activity despite the expectation that such
permission will degrade the eco-system. Assuming that a) the vote is explicit; b) the
expected damage is well specified, c) property rights are clearly held by the community
(i.e., the community has the right to refuse entry), d) the community's gains can be easily
estimated, and e) the transactions costs are low, the payment represents the ceiling on the
community's valuation of the environmental amenities that are being relinquished. It is
a ceiling because of the possibility that the community would have accepted a lower level
of compensation, and if the community valued the forgone eco-system services more than
the compensation, then presumably it would not have accepted the compensation.
However, if there is a vote and the outcome is close, the calculated valuation can be
considered to be close to the community's valuation.
The estimation task involves assessing the amount of environmental damage in
physical terms and the amount of compensation in monetary terms. Typically this
compensation will come in the form of additional sources of taxes, the value of
infrastructure that the new entrants provide for the community, additional income earned
by community members, etc. The per-household as well as per-community
compensation would be relevant. For example, the entry of an air-polluting factory may
be accepted only after the factory's owner commits to a certain number of jobs for the
community, building a park, upgrading roads, contributing to the community's vocational
program.
Obviously many "community decisions" to permit the entry of polluters or other
activities that degrade the ecosystem are not amenable to this approach, because
98
-------
community leaders negotiate the level of benefits that the community will receive without
a vote being taken, or the benefits or costs are difficult to estimate.
Public Decisions to Accept Pollution or Resource Depletion Followed by a Survey
Just as the analysis of referendum and initiative outcomes can be augmented by
determining voters' perceptions of the stakes, the ceiling of the willingness-to-accept
value of eco-system deterioration can be estimated by determining the benefits perceived
by voters who supported the arrangement accepting the entry of a polluting or depleting
operation into the community, and their perceptions of the damage that would be done.
Like the direct analysis of willingness-to-accept votes, if the arrangement was approved
by the electorate, and the property rights clear and transactions are low, the ratio of the
perceived benefits and costs represents the ceiling of the median voter's valuation. The
survey, best administered as soon as possible after the actual vote, would reveal what the
community members interpreted the benefits and costs to be, thus bringing the valuation
closer to individual values; but again with the tradeoff that the results would not have
standing as the "community's choice." If the survey includes the questions of the
conventional contingent valuation survey questions regarding how much each respondent
would have been willing to accept, then the results would be even more robust in finding
mean and aggregate valuations as well as median valuations.
Uses and Limitations of All Four Variants
All of these approaches attempt to measure the sum total of values of improving
or protecting eco-systems and eco-system services; therefore both means and ends
(instrumental and intrinsic) values can be involved. All variants in principle could
measure the values attributed to all types of services, expressed in terms of monetary
values per unit of eco-system improvement or protection. The variants are flexible in
terms of levels of data, detail and scope, inasmuch as initiatives, referenda and other
public decisions have been made at all sub-national levels. The valuations can be
aggregated across benefits and with other methods, as long as the scale and magnitude of
benefits are roughly the same. While highly complex initiatives, referenda, and other
public decisions are not good candidates for estimating value, the valuations generated
from simpler cases can be used as inputs for complex applications.
Any EPA decision context calling for monetized valuation could employ any of
these variants, either singly or as cross-checks with conventional revealed preference or
sated preference approaches. Benefit transfer applications will be limited to cases of
similar magnitudes of benefits, because of the likelihood that community decisions are
highly sensitive to such magnitudes.
The first two variants, in analyzing referenda and initiatives, can evaluate
tradeoffs between community and/or household costs (higher taxes, possibly job losses)
and eco-system improvements (establishment or improvement of air, water, biodiversity
protection, etc.). The third and fourth variants can evaluate tradeoffs between community
and/or household benefits (increase in tax base, job creation, infrastructure
improvements, etc.) and eco-system deterioration (greater pollution, amenity reductions).
99
-------
In uses that apply valuations directly to the jurisdiction previously experiencing
the initiative, referendum or negotiation, the scale would be the same municipality,
country or state. For benefits transfer, the scale should also be the same, given the need
for similar magnitude of benefits and costs mentioned above.
The outputs of these approaches should be easy to understand and to
communicate to the public. It is a significant advantage to be able to say that the
valuation of an eco-system component has been estimated on the basis of how
community's have decided what these components are worth.
These approaches would work best when:
applied to the same jurisdiction (e.g., if Portland is considering another storm
control issue, the analysis of the Portland referendum would be most
appropriate), but can still be used via benefits transfer;
a unitary conservation or environmental benefit is involved;
the initiative or referendum outcome was a close vote (this yields stronger
inferences about the actual valuation, rather than floors or ceilings);
extraneous issues (such as whether the vote is a "political test" on particular
politicians, or the mode of financing is controversial) are unimportant;
surveys can be accomplished soon after the actual vote.
100
-------
The resources needed to implement the variants would depend on the
applications. If the purpose is to compile a set of initiative and referendum results, this
could be done for the first approach by a) assigning an EPA economist to oversee the
effort (perhaps 10% effort over a year); b) assigning an intern to compile as many U.S.
municipal, county and state initiatives and referenda related to environmental and
conservation held over the past half-decade, (perhaps 50% effort over a year). The
analysis to generate valuations would require 10% of the time of a two-person team of
EPA economists, perhaps one being a consultant. For the second variant, more effort is
required for each survey: two EPA analysts (or consultants) each devoting one month to
develop, administer, and analyze the survey results.
The major obstacle to the effective use of these approaches may be the lack of
familiarity within government of the approach of drawing inferences from public
decisions, although the method has had a respectable history of use in estimating the
value of a "statistical life." It is striking that despite the multiple studies of how
conventional valuation methods such as contingent valuation compare to initiatives or
referenda outcomes, there is apparently no literature that takes the outcomes of the
initiatives or referenda per se as valuations, except to study why different subunits (e.g.,
counties within California [Kahn & Matsusaka, 1997]) yield different outcomes. Perhaps
it is just too simple a findingthat a particular initiative or referendum that devotes X
dollars to gain Y enhancement or protection of the eco-systemto warrant publication.
Nevertheless, the paucity of literature may be an obstacle to adopting this approach.
Addressing the Uncertainty Entailed in these Approaches
The uncertainties involved in the variants (first and third) that focus on benefits
and costs specified in the proposals lie in the estimates of actual benefits and costs
entailed in the proposals. They should be analyzed with the standard methods of
projecting consequences, and conveyed through probability distributions and confidence
intervals. The uncertainties involved in the approaches that rely on surveys lie in the
potential for biased sampling in the selection of survey respondents, as well as poor
memory and response set (e.g., respondents may report that they voted). These can be
reduced through careful random sampling and cross-checks within the questionnaires.
References
Arrow, K., R. Solow, P.R. Portney, E.E. Learner, R. Radner, & H. Schuman (1993).
Report of the NOAA Panel on Contingent Valuation, Washington, D.C.:
Government Printing Office, January 11.
Ashenfelter, Orley, & Michael Greenstone, 2004. "Using Mandated Speed Limits to
Measure the Value of a Statistical Life," Journal of Political Economy, vol. 112:
S226-S267.
Deacon, R., & P. Shapiro. 1975. "Private preference for collective goods revealed
through voting on referenda," American Economic Review 65: 793.
Kahn, M.E., & J.G. Matsusaka. 1997. "Demand for environmental goods: Evidence from
voting patterns on California initiatives," Journal of Law and Economics 40:
137-173.
101
-------
Kahn, M.E., & J.G. Matsusaka. (1997). "Demand for environmental goods: Evidence
from voting patterns on California initiatives," Journal of Law and Economics
40: 137-173.
List, J. and Gallet, C. "What Experimental Protocol Influence Disparities Between Actual
and Hypothetical Stated Values? Evidence from a Meta-Analysis," Environmental
and Resource Economics (2001), 20 (3): pp. 241-254.
List, J. and Shogren, J. "Calibration of Willingness-to-Accept," Journal of Environmental
Economics and Management (2002), 43 (2): pp. 219-233.
Lowi, Theodore J. 1964, "American Business, Public Policy, Case Studies and Political
Theory," World Politics 15(3), 677-715.
Murphy, J., P.G. Allen, T. H. Stevens & D. Weatherhead. (2003). A Meta-Analysis of
Hypothetical Bias in Stated Preference Valuation. University of Massachusetts
Amherst, Department of Resource Economics, Working Paper No. 2003-8.
Polasky, S., O. Gainutdinova, and J. Kerkvliet (1996). "Comparing CV Responses with
Voting Behavior: Open Space Survey and Referendum in Corvallis Oregon,"
paper presented at annual U.S.D.A. W-133 meeting, Jekyll Island, GA, February,
1996.
Sagoff, Mark. 2004. Price, Principle, and the Environment. Cambridge: Cambridge
University Press.
Schlapfer, F., A Roschewitz, & N. Hanley (2004). "Validation of stated preferences for
public goods: a comparison of contingent valuation survey response and voting
behaviour," Ecological Economics, 51: 1-2 , 1 (November): 1-16.
Shabman, L., & K. Stephenson. (1996). "Searching for the correct benefit estimate:
Empirical evidence for an alternative perspective," Land Economics, 72(4)
(November): 433-49.
Vossler, C.A., & J. Kerkvliet. 2003. "A criterion validity test of the contingent valuation
method: Comparing hypothetical and actual voting behavior for a public
referendum," Journal of Environmental Economics and Management 45(3): 631 -
49.
Vossler, C. A., J. Kerkvliet, S. Polasky and O. Gainutdinova (2003), "Externally
Validating Contingent Valuation: An Open-Space Survey and Referendum in
Corvallis, Oregon," Journal of Economic Behavior and Organization 51: 261-77.
102
-------
Inferring Values from Public
Choices
W. Ascher
VPESS
December 2005
Vehicles of Public
Input Reflecting Values
Individual-response
valuation
public opinion polls
non-governmental fora
public hearings
public notice and
comment
concertation
quasi-governmental
commissions
direct community
decision-making through
town halls, etc
elected representation
public opinion polls
letter writing, emailing or
calling an elected
representative
taking existing public
policies as the revealed
preferences of society
contributions to non-
governmental efforts
willingness to accept
negotiations
referenda/initiatives
1.
Revealed Preferences from
"Public Choices"
Referenda/initiatives
Referendum: legislature calls for a
public vote
Initiative: citizen petition
Usually for eco-system improvements
2. Willingness-to-accept negotiation
outcomes
Best if voted; but could have other
indications of "close call"
Different conception of
what value is:
Intensity
Median: the majority (or close to majority)
of sufficiently engaged people believe that
the expense is worth it
-Closer to 50-50, the better, though floors
or ceilings can be estimated regardless
Different conceptions:
Intrinsic validity
-IF one accepts that society's decisions
have standing as expressions of value
-Whether private utility or public
regardedness
Different conceptions:
Conception of democracy & representation
-Anti-Burkean, non-Benthamite
Burke: representatives, not citizens,
choose what is good for the people
Bentham: greatest [private] good for
the greatest number
-Government should do what the public
wish government to do
103
-------
Criticisms
1. Referenda & initiatives are subject to
intense politicking
- But politicking is pervasive &
democratic
2. Perceptions of benefits & costs may
diverge from actual stakes
- But it is possible to follow up with
surveys to determine how the stakes
are perceived
Criticisms
3. Other issues determine the vote
-Popularity of backers; partisan
maneuvering
-For willingness-to-pay: restrict to simple-
issue referenda or initiatives
-For willingness-to-accept: simple-issue
Coasean negotiations
Criticisms
Not capable of determining the option of
greatest aggregate utility
60% favor because their net gain is
+$100
40% against because their net gain is -
$200
-Usually true, but logic is simply different
-Assuming 0 value for opponents, a floor
on mean value is possible
Complications:
Different benefits transfer complications
-Disentangling objectives if multiple
issues
Or, contingent valuation keyed to
actual cases of pending decisions
Complications:
More than 50% vote margin will
underestimate the community's collective
valuation
-Result is therefore a floor
Multiple issues obscure the willingness to
pay for any single benefit
-Go for simple-issue referenda or
initiatives
Different benefits transfer complications
Validation of More Conventional
Valuation Methods
Several studies predict referendum votes
from contingent valuation estimates; check
whether the predictions are borne out
104
-------
Validation of More Conventional
Valuation Methods
NOAA Panel:
[E]xternal validation of elicited lost passive use
values is usually impossible. There are however
real-life referenda. Some of them, at least, are
decisions to purchase specific public goods with
defined payment mechanisms, e.g., an increase
in property taxes. The analogy with willingness
to pay for avoidance or repair of environmental
damage is far from perfect but close enough that
the ability of CV-like studies to predict the
outcomes of real-world referenda would be
useful evidence on the validity of the CV method
in general.
Validation of More Conventional
Valuation Methods
The test we envision is not an election poll of the
usual type. Instead, using the referendum format
and providing the usual information to the
respondents, a study should ask whether they
are willing to pay the average amount implied by
the actual referendum. The outcome of the CV-
like study should be compared with that of the
actual referendum. The Panel thinks that studies
of this kind should be pursued as a method of
validating and perhaps even calibrating
applications of the CV method
105
-------
Juries
Buzz Thompson
C-VPESS
December 2005
Relevance of Initiatives/Public
Negotiations
~ Sources of Revealed
~ Public Valuations
1 Preferences
~ Citizen valuation
¦ Legislative decisions
juries
~ Conservation
Consensus
investments
conferences
¦ NGO acquisitions &
¦ Public-regarded
investments
surveys
~ E.g., conservation
easements
¦ Jury awards
~ E.g., NRD actions
¦ Judge-ordered
awards
Potential Advantages of Jury Awards
over Referenda/Initiatives
~ More complete information
~ More deliberative
¦ Also reflective & evolutionary
~ Often more focused issue
~ Continuum of choices
~ "Due process" protections
¦ Politicking expressly excluded
~ More likely to generate "public valuation"??
¦ Implicit role
¦ Judicial instructions
¦ Responsibility and impact
Potential Problems with Jury Awards
(shared with referenda/initiatives)
~ Limited availability
~ Legitimacy of valuation transfer
Public benefits <-4 private costs
¦ Context: legal violation
~ Voting rule
Majority voting
¦ Super-majority voting
¦ Consensus
Potential Problems with Jury Awards
(unique)
~ High variance
Small number of jurors
~ Jury representativeness
Bias toward old and poor
Do instructions help overcome?
~ Potential circularity
Two models of jury process:
~ Informed jury valuation
~ Judging expert credibility
Citizen Valuation Juries
~ Can organize around any valuation
question
~ Can specify decision-making rule &
model
~ Better control over representation
~ Valuation transfer less problematic
~ But still issues re:
Small size
Variance in results
106
-------
Potential Advantages Over Surveys
~ More complete information
~ More deliberative
Also reflective and evolutionary
~ More likely to generate "public
valuation"??
~ Opportunity for citizen involvement &
"empowerment"
Valuation Framework:
1 Consumers-Citizen Spectrum
Pure
Consumer
<
Pure
Citizen
>
~ Empirical evidence
¦ Greater valuation of public goods
¦ Greatest effect for environmental goods
~ Why different valuations?
¦ Differences in information considered
¦ Levels of deliberativeness
¦ Degree of "other regardedness"
Citizen Valuation Juries:
Design Issues
~ Jury composition
¦ Number of jurors
¦ Representation
~ Charge
~ Voting rules
~ Witness selection
107
-------
12.3. Deliberative Approaches for Modeling, Valuation, and Decision Making
Session Leaders:
Dr. Robert Costanza, Professor/Director, Gund Institute for Ecological Economics,
School of Natural Resources, University of Vermont
Dr. Joseph Arvai, Professor/Director, Skunkworks Lab Department of Community,
Agriculture, Recreation, and Resource Studies, Michigan State University and Principal
Investigator, Decision Research, Eugene, OR
Contents:
Conceptual Framework for the Decision Science Approach to Values
Deliberative Approaches
Mediated Modeling
Conceptual Framework for the Decision Science Approach to Values
The decision science perspective on valuing the protection of ecological systems
and services is, at its core, relativist. Form this perspective, the "value" surrounding
ecological systems and services is not an absolute concept, despite the fact that numerical
and narrative descriptions of individual components of it (absent a comparison) may be
obtained using a variety of economic and non-economic (e.g., psychological, biophysical,
etc.) methods. Instead, the decision sciences take the view that that the overall value that
is ascribed to the environment and its services can only be fully understood in a
comparative context; in other words, we can only say that a systemor indeed the suite
of services provided by that systemhas a high or low value in the context of:
(a) retrospective evaluations undertaken by analyzing the degree of change
experienced by the system relative to some previous or unaltered state (i.e., a system is
either more or less valuable because it performs either better or worse than it did before),
or
(b) decision making for management undertaken by comparing predictions about how
a system or its suite of services might behaveagain better or worse relative to its
current conditionafter it has been subjected to one or more possible management or
regulatory options.
The attributes across which these changesand hence, valuesare accounted for
are defined by the objectives of a given decision context. These objectives tend to be
diverse and simultaneously incorporate inputs from a wide variety of disciplines. It is not
atypical, for example, to ascribe an overall relative value to an ecological system or
service based on the extent to which it maintains some requisite level of ecological
function and productivity, provides security for endangered or threatened species,
facilitates the maintenance of key services such as nutrient cycling or decomposition,
yields economic outputs in the form of resource extraction and tourism, lends itself to
desired recreation opportunities, and supplies a sense of pride or awe (Gregory et al.
2001). In this sense, the decision sciences straddle the line between economic and non-
economic approaches to valuation in that inputs for a formal comparison of options in the
108
-------
case of management decisions, and current and previous conditions in the case of
evaluation, are required from fields such as economics, ecology, psychology, and
sociology. However, absent an explicit framework for comparison across attributes, and
options or alternative states, individual inputs from these sources have very little
meaningfrom an overall "value" standpointin their own right.
Thus, a decision science approach to valuing the protection of ecological systems
and services is explicitly multiattribute in nature. Absent this multiattribute view of
valuewith the various attributes of value tied to the concerns stated by stated by
technical experts and other key stakeholdersthe relative values obtained often fall short
of providing the requisite guidance for decision making and evaluation, and run the risk
of not meeting or surpassing the threshold of relevancy (Keeney & Raiffa 1993)
defined chiefly by those who will hold decision makers and agencies accountable. Of
course, a multiattribute and comparative view of value presents challenges to decision
makers and evaluators. For example, those who undertake valuations geared toward the
decision sciences must be prepared to work with multiple and diverse stakeholders
sometimes over extended temporal periods, conduct additional decision-specific technical
analyses that are linked to stated objectives, and address complex and often contentious
tradeoffs (Arvai et al. 2001; Gregory et al. 2001; Hammond et al. 1999; Keeney &
Gregory 2005).
References
Arvai, J. L., R. Gregory, and T. McDaniels. 2001. Testing a structured decision approach:
Value-focused thinking for deliberative risk communication. Risk Analysis
21:1065-1076.
Gregory, R. 2000. Using stakeholder values to make smarter environmental decisions.
Environment 42:34-44.
Gregory, R., J. L. Arvai, and T. McDaniels. 2001. Value-focused thinking for
environmental risk consultations. Research in Social Problems and Public Policy
9:249-275.
Hammond, J., R. L. Keeney, and H. Raiffa 1999. Smart Choices: A Practical Guide to
Making Better Decisions. Harvard Business School Press, Cambridge, MA.
Keeney, R. L., and R. Gregory. 2005. Selecting attributes to measure the achievement of
objectives. Operations Research 53:1-11.
Keeney, R. L., and H. Raiffa 1993. Decisions with multiple objectives: Preferences and
value tradeoffs. Cambridge University Press, Cambridge, UK.
Deliberative Approaches
Significant interest has been devoted to multi-stakeholder, deliberative processes
for environmental decision making both at EPA (e.g., EPA 2000) and elsewhere (e.g.,
Beierle and Cayford 2002; Beierle 2002). Much of this interest has focused on
deliberative processes as a means of legitimizing resulting policy decisions. To this end,
there have been several examples of both research and practice where deliberative
approaches to decision making have resulted in a high degree of participant satisfaction
109
-------
in a variety of different management contexts (McDaniels, Gregory et al. 1999; Arvai
2003). Results from these studies, and others (e.g., Kraft 1988; National Research
Council 1989; e.g., Heiman 1990; Vari, Mumpower et al. 1993), argue that people are
more likely to accept outcomes that result from decision making processes that seem fair,
reasonable, and amenable to allowing the public and other stakeholders an opportunity to
voice their feelings and concerns.
This argument is also in line with writing on "procedural justice", which suggests
that a higher degree of acceptance is be expected for decisions that seem fair to the
affected parties from the point of view of both the decision outcome and the process that
resulted in it (Lind and Tyler 1988; Kraft and Scheberle 1995). In other words, people
whose individual interests are adversely affected by an outcome may be more willing to
accept decisions because they perceive that they have been dealt with fairly, they
understand the other participants' positions, and they have had the opportunityeven if
comes indirectlyto contribute to the debate (Syme, Macpherson et al. 1991; Hillier
1998).
Why does this positive relationship between deliberative processes and support
for resulting decisions exist? Some have suggested greater stakeholder satisfaction
results from a frame shift during decision making from one that is imposed to one that is
voluntary (Slovic 1987). Others have suggested that greater stakeholder satisfaction with
decisions that are the product of deliberative approaches is simply the manifestation of a
halo effect (Thorndike 1920). In this latter case, people tend to judge multiple
dimensions of a stimulus in much the same was as they judge the most salient dimension.
In other words, when one judges a decision to be "good" in one dimension (i.e., because
it was made in a deliberative fashion), they are also likely to judge the same decision to
be good in other dimensions (i.e., the outcomes of that decision).
Beyond these "stakeholder relations" benefits, there are other reasonsreasons
that are of greater interest to this committeefor advocating the use of deliberative
approaches for valuation and decision making. Foremost among these is the fact that
these approaches work to foster the inclusion of differently formulated objectives,
concerns, and arguments in the valuation and decision making process (NRC 1996; Chess
and Purcell 1999; Renn 1999; Gregory 2000).
Indeed, EPA itself has acknowledged this point, stating in the past that the
American people are the agency's primary "customer" and to this end issued the
following policy statement (EPA 2000, p. 1): "We are committed to providing the best
customer service possible. We aim to achieve this through increased public participation,
increased access to information, and more effectively responding to customer needs."
This is a sweeping statement that applies to a wide variety of valuation contexts,
including both those that involve single valuation metrics (e.g., dollar responses obtained
via contingent valuation) and multiattribute inputs obtained via multi-stakeholder
approaches (e.g., such as mediated modeling and structured decision approaches).
For example, in the context of contingent valuation, a commitment to deliberative
110
-------
approaches implies that EPA will seek input from stakeholders regarding such things as:
the ecological systems or services that will be the subject of valuations, the aspects of
these ecological systems or services to be valued (e.g., the attributes by which an object
such as aesthetic quality might be defined), appropriate measures (e.g., dollars for
economic valuations; indices of quality for environmental attributes) for valuation
outputs, and appropriate ways to frame and implement valuation questions.
Likewise, in the context of multiattribute approaches, this commitment guides
EPA to seek input regarding: problem identification and framing, stakeholders'
objectives as they relate to a given decision or evaluation context, the range of options
that may be considered as part of a management decision, valuation inputs to consider
during decision making or evaluation; these include results from valuation processes that
include, but are not limited to CV, deliberative value elicitations, and the results from
(non-monetized) surveys, and information about the tradeoffs that must be addressed
when selecting one option over another.
References
Arrow, K., K. Solow, P. R. Portney, E. E. Learner, R. Radner, and H. Schuman. 1993.
Report of the NOAA Panel on Contingent Valuation. Federal Regsiter 15:4601-
4614.
Arvai, J. L. 2003. Using risk communication to disclose the outcome of a participatory
decision making process: Effects on the perceived acceptability of risk-policy
decisions. Risk Analysis 23:281-289.
Arvai, J. L., and R. Gregory. 2003. A decision focused approach for identifying cleanup
priorities at contaminated sites. Environmental Science & Technology 37:1469-
1476.
Arvai, J. L., R. Gregory, and T. McDaniels. 2001. Testing a structured decision approach:
Value-focused thinking for deliberative risk communication. Risk Analysis
21:1065-1076.
Arvai, J. L., T. McDaniels, and R. Gregory. 2002. Exploring a structured decision
approach for fostering participatory space policy making at NASA. Space Policy
18:221-231.
Arvai, J. L. (2003). "Using risk communication to disclose the outcome of a participatory
decision making process: Effects on the perceived acceptability of risk-policy decisions."
Risk Analysis 23: 281-289.
Beierle, T. and J. Cayford (2002). Democracy in Practice: Public Participation in
Environmental Decisions. Washington, DC, Resources for the Future.
Beierle, T. C. (2002). "The quality of stakeholder-based decisions." Risk Analysis 22:
739-749.
Bond, P. and P. Goldblatt (1984). "Plants of the Cape Flora." Journal of South African
Botany Suppl. 13: 1-455.
Ill
-------
Chess, C. and K. Purcell (1999). "Public participation and the environment: Do we know
what works?" Environmental Science & Technology 33(16): 2685-2691.
Costanza, R. and R. Matthias (1998). "Using Dynamic Modeling to Scope Environmental
Problems and Build Consensus." Environmental Management 22: 183-195.
Cowling, R. and R. Costanza (1997). "Valuation and Management of Fynbos
Ecosystems." Ecological Economics 22: 103-155.
Cowling, R. M. (1992). The ecology of fynbos. Nutrients, fire and diversity. Cape Town,
Oxford University Press.
Environmental Protection Agency (2000). Engaging the American People: A Review of
EPA's Public Participation Policy and Regulations with Recommendations for Action.
Washington, DC: 28.
Gregory, R. (2000). "Using stakeholder values to make smarter environmental decisions."
Environment 42(5): 34-44.
Gunderson, L. C., S. Holling, et al. (1995). Barriers and bridges to the renewal of
ecosystems and institutions. New York, Columbia University Press.
Heiman, M. (1990). "From "not in my backyard!" to "not in anybody's backyard!""
Journal of the American Planning Association 56: 359-362.
Higgins, S. I., J. K. Turpie, et al. (1997). "An ecological economic simulation model of
mountain fynbos ecosystems: dynamics, valuation, and management." Ecological
Economics 22: 155-169.
Hillier, J. (1998). "Beyond confused noise: Ideas toward communicative procedural
justice." Journal of Planning Education and Research 18: 14-24.
Hobbs, R. J., D. M. Richardson, et al. (1995). Mediterranean-type ecosystems:
opportunities and constraints for studying the function of biodiversity. Mediterranean-
type ecosystems. The function of biodiversity. G. W. Davis and D. M. Richardson.
Berlin, Springer: 1-42.
Kraft, M. E. (1988). Evaluating technology through public participation: The nuclear
waste disposal controversy. Technology and Politics. M. E. Kraft and N. J. Vig. Durham,
NC, Duke University Press: 253-257.
Kraft, M. E. and D. Scheberle (1995). "Environmental justice and the allocation of risk:
The case of lead and public health." Policy Studies Journal 23: 113-122.
Lind, E. and T. Tyler (1988). The Social Psychology of Procedural Justice. New York,
NY., Plenum Press.
112
-------
McDaniels, T., R. Gregory, et al. (1999). "Democratizing risk management: Successful
public involvement in local water management decisions." Risk Analysis 19: 497-510.
Myers, N. (1990). "The biodiversity challenge: expanded hot-spots analysis." The
Environmentalist 10: 243-255.
National Research Council (1989). Improving Risk Communication. Washington, DC,
National Academy Press.
National Research Council (1996). Understanding Risk: Informing Decisions in a
Democratic Society. Washington, DC, National Academy Press.
Rao, S. S. (1994). "Welcome to open space." Training(April): 52-55.
Renn, O. (1999). "A model for analytic-deliberative process in risk management."
Environmental Science & Technology 33: 3049-3055.
Slovic, P. (1987). "Perception of risk." Science 236: 280-285.
Syme, G., D. Macpherson, et al. (1991). "Factors motivating community participation in
regional water allocation planning." Environment and Planning A 23: 1779-1795.
Thorndike, E. L. A. (1920). "Constant error in psychological ratings." Journal of Applied
Psychology 4: 25-29.
van den Belt, M. (2004). Mediated Modeling: a systems dynamics approach to
environmental consensus building. Island Press.
Vari, A., J. L. Mumpower, et al. (1993). Low-Level Radioactive Waste Disposal Facility
Siting Processes in the United States. Albany, NY, Center for Policy Research. State
University of New York.
113
-------
A Decision Science Perspective on Valuing the
Protection of Ecological Systems and Services
Joseph Arvai
Michigan State University
Decision Research
T: 517-353-0694
E: arvai@msu.edu
OUTLINE
The value of ecological systems and services is
driven by people's objectives.
The overall value of ecological systems and services
is multiattribute in nature.
The value of systems or services is established
through an analysis of management alternatives:
thus, the value of ecological systems and services is
relative and reflects the tradeoffs that people are
willing to make.
OBJECTIVES
Analyses of people's (stakeholders, public, experts,
etc.) objectives identifies the attributes of systems
and services that deserve attention in "valuation".
Consultation with technical experts (economists,
ecologists, etc.) identifies appropriate measures for
these attributes.
ATTRIBUTES AND MEAUSURES
In response to the desire/need for quantification...
Natural Measures - Direct measures of an attribute
e.g., monetary value of electricity generated ($)
Proxy Measures - Indirect measures of an attribute
e.g., habitat quality as a measure of the health offish communities
Constructed Measures - Measures created for an attribute
e.g., index of accessibility for cultural or spiritual purposes
OBJECTIVES, ATTRIBUTES, MEASURES
Objectives
Attributes; Measures
Recreation
Access to recreation opportunities; Weighted User Days
Environmental Health
Erosion levels; Weighted Eroshn Days
Environmental Health
Flow levels; Weighted Flood Days
Environmental Health
Habitat quality; % Available Habitat, IBI
Environmental Health
Water quality; Multiattribute Index (particulates, PCBs, etc.)
Cultural
Regular access to sites; Cons/sfency Index
Economic
Revenues; Annual Revenues M$ / Year
MULTIATTRIBUTE VALUATION
Objectives
Attributes/
Measures
Option A
Enhance
recreation
opportunities
Access/
Weighted user
days
X DAYS
Enhance
environmental
health
Habitat Quality/
% Available
Y%
Maximize
economic returns
Revenues/
$Mil/Yr
SZ MiZYr
114
-------
ESTABLISHING VALUE
Objectives
Attributes/
Measures
Hydrograph
Releases
Enhanced Winter
Releases
Enhance
recreation
opportunities
Access/
Weighted user
days
1400
1200
1500
Enhance
environmental
health
Habitat Quality/
% Available
50%
20%
30%
Maximize
economic returns
Revenues/
$Mil/Yr
$60
$80
$65
ESTABLISHING VALUE
Objectives
Attributes/
Measures
Hydrograph
Enhanced Summer
Releases
Enhanced Winter
Releases
Enhance
recreation
opportunities
Access/
Weighted user
days
1400^
1200
1500
Enhance
environmental
health
Habitat Quality/
% Available
50%
^
"Value" is
function 0
°ta3
f a
sure
$
D%
Maximize
economic returns
Revenues/
$ Mi l/Yr
$60
V
$80
65
ESTABLISHING VALUE
Objectives
Attributes/
Measures
Hydrograph
Releases
Enhanced Winter
Releases
Enhance
recreation
opportunities
Access/
Weighted user
days
1400
12
00
Nor is
the fu
%mpc
1500
""N
it simply
tction of a
site score
J
-
Enhance
environmental
health
Habitat Quality/
% Available
50%
Maximize
economic returns
Revenues/
$Mil/Yr
$60
$80
$65
ESTABLISHING VALUE
Attributes/
Measures
Enhanced Summer
Enhance
recreation
opportunities
Enhance
environmental
health
Revenues/
$Mi l/Yr
1400-
1200
Habitat Quality/
% Available
-400
Instead, the value of a given
option exists in the tradeoffs
people 91a willing to njakg
'8ss notjueHneir objectives'0
but also the level of achievement
. with respect to them. «
$60 $80 $60
ESTABLISHING VALUE
Objectives
Attributes/
Measures
Hydrograph
Releases
Enhance
recreation
opportunities
Access/
Weighted user
days
1400
1200
Enhance
environmental
health
Habitat Quality/
% Available
50%
20%
Maximize
economic returns
Revenues/
$Mil/Yr
$60
$80
ADVANTAGES
~ Provides both preference orders (A>B>C) and
relative values (A=2B=3C) for management options.
~ Multiattribute, inclusive, and transparent.
~ Useful for both decision making and retrospective
evaluation.
~ High level of methodological precision.
115
-------
CHALLENGES
~ Effortful and potentially time consuming.
~ Overly formal for many EPA decisions; decision
makers may wish to protect their autonomy.
~ Not explicitly geared towards current OMB
requirements for regulatory evaluation.
Mediated Modeling
Brief Description of Method. Computer models of complex systems are
frequently used to support decisions concerning environmental problems. To effectively
use these models, (i.e. to foster consensus about the appropriateness of their assumptions
and results and thus to promote a high degree of compliance with the policies derived
from the models) it is not enough for groups of academic "experts" to build and run the
models. What is required is a different role for modeling - as a tool in building a broad
consensus not only across academic disciplines, but also between science and policy.
Mediated modeling is the involvement of stakeholders (parties interested in or affected by
the decisions the model addresses) as active participants in all stages of the modeling
process, from initial problem scoping to model development, implementation and use
(Costanza and Matthias 1998; van den Belt 2004). Integrated modeling of large systems,
from individual companies to industries to entire economies or from watersheds to
continental scale systems and ultimately to the global scale, requires input from a very
broad range of people. We need to see the modeling process as one that involves not
only the technical aspects, but also the sociological aspects involved with using the
process to help build consensus about the way the system works and which management
options are most effective. This consensus needs to extend both across the gulf
separating the relevant academic disciplines and across the even broader gulf separating
the science and policy communities, and the public. Appropriately designed and
appropriately used mediated modeling exercises can help to bridge these gulfs. The
process of mediated modeling can help to build mutual understanding, solicit input from
a broad range of stakeholder groups, and maintain a substantive dialogue between
members of these groups. Mediated modeling and consensus building are also essential
components in the process of adaptive management (Gunderson, Holling et al. 1995). An
extended description of this method can be found in Appendix. B.
Mediated Modeling and Value. Mediated models can contain explicit valuation
components. In fact, if the goal of the modeling exercise is to consider trade-offs, then
valuation of some kind becomes an essential ingredient. How these trade-offs and
valuations get incorporated into the model, varies, of course, from exercise to exercise.
116
-------
Perhaps the best way to describe this process is with an example. The South African
fynbos ecological economic model described by Higgins et al. (1997) is an illustrative
example.
The area of study for this example was the Cape Floristic Regionone of the
world's smallest and, for its size, richest floral kingdoms. This tiny area, occupying a
mere 90,000 km2, supports 8,500 plant species of which 68% are endemic, 193 endemic
genera and six endemic families (Bond and Goldblatt 1984). Because of the many threats
to this region's spectacular flora, it has earned the distinction of being the world's
"hottest" hot-spot of biodiversity (Myers 1990).
The predominant vegetation in the Cape Floristic Region is fynbos, a hard-leafed
and fire-prone shrubland which grows on the highly infertile soils associated with the
ancient, quartzitic mountains (mountain fynbos) and the wind-blown sands of the coastal
margin (lowland fynbos) (Cowling 1992). Owing to the prevalent climate of cool, wet
winters and warm, dry summers, fynbos is superficially similar to California chaparral
and other Mediterranean climate shrublands of the world (Hobbs, Richardson et al.
1995). Fynbos landscapes are extremely rich in plant species (the Cape Peninsula has
2,554 species in 470 km2) and plant species endemism ranks amongst the highest in the
world (Cowling 1992).
In order to adequately manage these ecosystems several questions had to be
answered, including, what services do these species-rich fynbos ecosystems provide and
what is their value to society? A two-week workshop was held at the University of Cape
Town (UCT) with a group of faculty and students from different disciplines along with
parks managers, business people, and environmentalists. The primary goal of the
workshop was to produce a series of consensus-based research papers which critically
assessed the practical and theoretical issues surrounding ecosystem valuation as well as
assessing the value of services derived by local and regional communities from fynbos
systems.
To achieve the goals, an 'atelier' approach was used to form multidisciplinary,
multicultural teams, breaking down the traditional hierarchical approach to problem-
solving. Open space (Rao 1994) techniques were used to identify critical questions and
allow participants to form working groups to tackle those questions. Open space
meetings are loosely-organized affairs which give all participants an opportunity to raise
issues and participate in finding solutions.
The working groups of this workshop met several times during the first week of
the course and almost continuously during the second week. The groups convened
together periodically to hear updates of group projects and to offer feedback to other
groups. Some group members floated to other groups at times to offer specific
knowledge or technical advice.
Despite some initial misgivings on the part of the group, the structure of the
course was remarkably successful, and by the end of the two weeks, seven working
117
-------
groups had worked feverishly to draft papers. These papers were eventually published as
a special issue of Ecological Economics (Cowling and Costanza 1997). One group
focused on producing an initial scoping (or mediated) model of the fynbos. This
modeling group produced perhaps the most developed and implementable product from
the workshop: a general dynamic model integrating ecological and economic processes
in fynbos ecosystems (Higgins, Turpie et al. 1997). The model was developed in
STELLA and designed to assess potential values of ecosystem services given ecosystem
controls, management options, and feedbacks within and between the ecosystem and
human sectors. The model helped to address questions about how the ecosystem services
provided by the fynbos ecosystem at both a local and international scale are influenced by
alien invasion and management strategies. The model consists of five interactive sub-
models: a) hydrology; b) fire; c) plant; d) management; and (e) economic valuation.
Parameter estimates for each sub-model were either derived from the published literature
or established by workshop participants and consultants (they are described in detail in
Higgins, Turpie et al. 1997). The plant sub-model included both native and alien plants.
Simulation of the model produced a realistic description of alien plant invasions and their
impacts on river flow and runoff.
This model drew in part on the findings of the other working groups, and
incorporates a broad range of research by workshop participants. Benefits and costs of
management scenarios were addressed by estimating values for harvested products,
tourism, water yield and biodiversity. Costs included direct management costs and
indirect costs. The model showed that the ecosystem services derived from the Western
Cape mountains are far more valuable when vegetated by fynbos than by alien trees (a
result consistent with other studies in North America and the Canary Islands). The
difference in water production alone was sufficient to favor spending significant amounts
of money to maintain fynbos in mountain catchments.
The model was designed to be user-friendly and interactive, allowing the user to
set such features as area of alien clearing, fire management strategy, levels of wildflower
harvesting, and park visitation rates. The model has proven to be a valuable tool in
demonstrating to decision makers the benefits of investing now in tackling the alien plant
problem, since delays have serious cost implications. Parks managers have implemented
many of the recommendations flowing from the model.
There are several other case studies in the literature of various applications of
mediated modeling to environmental decision-making, including valuation. Van den Belt
(2004) is the best recent summary and synthesis.
Decision contexts where this method can be used. As described above, the
method is fairly general and could be used to assess any value (means toward and ends)
that a group of stakeholders could identify and build into a model. Any decision context
that requires the estimation of the values of ecosystem goods or services could employ
this method, although to the committee's knowledge no EPA decisions have as yet
employed this technique. The method covers all elements of the diagram after the initial
identification of EPA needs, and could be used in conjunction with the full range of
118
-------
decision models. Prior applications have been at a broad range of scales, from
watersheds or specific ecosystems to large regions and the global scale. The method is in
principle broadly applicable to the full range of time and space scales.
Resource Inputs/Limitations. Resources needed to implement the method vary
from application to application. The method can deal with a broad range of available
data and resources, probably better that most other methods, since the model can adapt to
the resources available across different levels of data, detail, scope and complexity. As a
rule of thumb, one can produce a credible mediated model in 30-40 hours of workshops;
about 300-400 hours of organizing/modeling. Cost: about $40,000 - $100,000 depending
on side activities. The most serious obstacle seems to be the fact that this method is very
different from the top-down approach most frequently used in government. It requires
that consensus building be put at the center of the process, which can be very scary for
institutions accustomed to controlling the outcome of decision processes. The final
outcome of this process cannot be predetermined.
Uncertainty: In terms of uncertainty, there are all the usual sources, but the
difference is that the stakeholders are exposed to these sources as they go, and learn to
understand and accommodate them as part of the process. The method is compatible
with formal or informal characterizing of uncertainty, producing probability distributions
in addition to point estimates.
Other important dimensions:
The method is inherently dynamic - that is what it does best
The results can be aggregated to get a single benefits number as needed.
Participants in the mediated modeling process gain deep understanding of
the process and products. Those who have not participated can easily
view and understand the results if they invest the effort. Usually the
results can (with some additional effort) be made accessible to a broad
audience.
Since the method explicitly discusses and incorporates subjective or
"framing" issues, it is at least open and transparent to users. No research
has yet been done on whether application of the process to exactly the
same problem by two independent groups would yield "consistent and
invariant" results. One would expect general consistency, but some
variation between applications. This is an area for further research.
References
Bond, P. and Goldblatt, P. 1984. Plants of the Cape Flora. Journal of South African
Botany Suppl. 13:1-455.
Checkland, P. 1989. Soft Systems Methodology, in J. Rosenhead (ed.) Rational
Analysis for a Problematic World, John Wiley and Sons, Chichester, England.
Costanza, R. 1987. Simulation Modeling on the Macintosh Using STELLA, Bioscience,
Vol. 37, pp. 129 - 132.
Costanza, R., F. H. Sklar, and M. L. White. 1990. Modeling Coastal Landscape
119
-------
Dynamics. Bioscience 40:91-107
Costanza, R. and M. Ruth. 1998. Using dynamic modeling to scope environmental
problems and build consensus. Environmental Management 22:183-195.
Costanza, R., A. Voinov, R. Boumans, T. Maxwell, F. Villa, L. Wainger, and H. Voinov.
2002. Integrated ecological economic modeling of the Patuxent River watershed,
Maryland. Ecological Monographs 72:203-231.
Cowling, R.M. (ed.). 1992. The ecology of fynbos. Nutrients, fire and diversity. Oxford
University Press, Cape Town.
Cowling, R. and R. Costanza (eds). 1997. Valuation and Management of Fynbos
Ecosystems. Special section of Ecological Economics vol 22, pp 103-155.
Ford, A. 1999. Modeling the Environment: An Introduction to System Dynamics Models
of Environmental Systems. Island Press, Washington, DC
Gunderson, L. C. S. Holling, and S. Light (eds). 1995. Barriers and bridges to the renewal
of ecosystems and institutions. Columbia University Press, New York. 593 pp.
Hannon, B. and M. Ruth. 1994. Dynamic Modeling, Springer-Verlag, New York.
Hannon, B. and M. Ruth. 1997. Modeling Dynamic Biological Systems, Springer-
Verlag, New York.
Higgins, S. I., J. K. Turpie, R. Costanza, R. M. Cowling, D. C. le Maitre, C. Marais, and
G. Midgley. 1997. An ecological economic simulation model of mountain fynbos
ecosystems: dynamics, valuation, and management. Ecological Economics
22:155-169.
Hobbs, R.J., Richardson, D.M. and Davis, G.W. 1995. Mediterranean-type ecosystems:
opportunities and constraints for studying the function of biodiversity. In:
Mediterranean-type ecosystems. The function of biodiversity. G.W. Davis and
D.M. Richardson (eds), pp 1-42. Springer, Berlin.
Kahnemann, D. and A. Tversky. 1974. Judgment Under Uncertainty, Science, Vol. 185,
pp. 1124 - 1131.
Kahnemann, D., P. Slovic, and A. Tversky. 1982. Judgment Under Uncertainty:
Heuristics and Biasis, Cambridge University Press, Cambridge.
Lyneis, J.M. 1980. Corporate Planning and Policy Design: A System Dynamics
Approach, Pugh-Roberts Associates, Cambridge, Massachusetts.
Morecroft, J.D.W. 1994. Executive Knowledge, Models, and Learning, in J.D.W.
Morecroft, and J.D. Sterman (eds.) Modeling for Learning Organizations,
Productivity Press, Portland, Oregon, pp. 3 - 28.
Morecroft, J.D.W., D.C. Lane and P.S. Viita. 1991. Modelling Growth Strategy in a
Biotechnology Startup Firm, System Dynamics Review, No. 7, pp. 93-116.
Morecroft, and J.D. Sterman (eds.) 1994. Modeling for Learning Organizations,
Productivity Press, Portland, Oregon
Myers, N. 1990. The biodiversity challenge: expanded hot-spots analysis. The
Environmentalist 10: 243-255.
Oster, G. 1996. Madonna, http://nature.berkeley.edu/~goster/madonna.html.
Peterson, S. 1994. Software for Model Building and Simulation: An Illustration of
Design Philosophy, in J.D.W. Morecroft, and J.D. Sterman (eds.) Modeling for
Learning Organizations, Productivity Press, Portland, Oregon, pp. 291 - 300.
Phillips, L.D. 1990. Decision Analysis for Group Decision Support, in C. Eden and J.
Radford (eds.) Tackling Strategic Problems: The Role of Group Decision
120
-------
Support, Sage Publishers, London.
Rao, S.S. 1994. Welcome to open space. Training (April): 52-55.
Richmond, B. and S. Peterson. 1994. STELLA II Documentation, High Performance
Systems, Inc., Hanover, New Hampshire.
Roberts, E.B. 1978. Managerial Applications of System Dynamics, Productivity Press,
Portland, Oregon.
Rosenhead, J. (ed.) 1989. Rational Analysis of a Problematic World, John Wiley and
Sons, Chichester, England.
Senge, P.M. 1990. The Fifth Discipline, Doubleday, New York.
Simon, H.A. 1956. Administrative Behavior, Wiley and Sons, New York.
Simon, H.A. 1979. Rational Decision-Making in Business Organizations, American
Economic Review, Vol. 69, pp. 493 - 513.
Van den Belt, M. 2004. Mediated Modeling: a systems dynamics approach to
environmental consensus building. Island Press, Washington, DC.
Vennix, J. A. M. 1996. Group Model Building : Facilitating Team Learning Using
System Dynamics. Wiley, NY.
Vennix, J.A.M. and J.W. Gubbels. 1994. Knowledge Elicitation in Conceptual Model
Building: A Case Study in Modeling a Regional Dutch Health Care System, in
J.D.W. Morecroft, and J.D. Sterman (eds.) Modeling for Learning Organizations,
Productivity Press, Portland, Oregon, pp. 121 - 146.
Westenholme, E.F. 1990. System Inquiry: A System Dynamics Approach, John Wiley
and Sons, Chichester, England.
Westenholme, E.F. 1994. A Systematic Approach to Model Creation, in J.D.W.
Morecroft, and J.D. Sterman (eds.) Modeling for Learning Organizations,
Productivity Press, Portland, Oregon, pp. 175 - 194.
121
-------
12.4. Social/Psychological Methods for Ecosystem Values Assessments
Session Leaders:
Dr. Terry Daniel, Professor of Psychology and Natural Resources, Department of
Psychology, Environmental Perception Laboratory, University of Arizona
Dr. Kathleen Segerson, Professor, Department of Economics, University of Connecticut
Contents:
Brief overview of social-psychological methods prepared by a sub-group
of the C-VPESS that represents initial ideas about what roles these
methods might play in ecosystem values assessments. Material is
intended to stimulate discussion among members of the Committee and
participants at the workshop.
Outline of session contents
For the purposes of EPA policy and decision making the values of ecosystems
and ecosystem services are based at least in part on the judgments of stakeholders and
citizens. Social/psychological methods are proven scientific means for determining
people's value-relevant perceptions and judgments about a wide array of objects, events
and conditions. Valuations and benefit assessments based on judgments by relevant
samples of stakeholders and/or citizens provide an appropriate basis for EPA policy and
decision making, along with economic (monetary) and bio-ecological assessments.
Social/psychological methods are characterized by:
An emphasis on descriptive rather than prescriptive models and reliance
on empirically based theories of human values, judgments and decision
making;
Acknowledgment of the important effects of the assessment contexts (e.g.,
representation/framing of assessment targets, mode of preference
expression, perceived intentions/goals of the assessors) and the associated
constraints on validity and generalizability of any assessment results;
Recognition of the effects of human predispositions, interpretations and
cognitive limitations (e.g., bounded rationality, mental models,
emotional/affective responses) on the outcome of any value assessment;
Use of a wide range of overt expressions of value (narratives,
lexicographic scales, ratings, choices, actions);
Assessments over multiple value dimensions (e.g., biocentric, utilitarian,
aesthetic, ethical) expressed in qualitative (lexical) or quantitative metrics
that need not be commensurate;
Segregation of different value proponents into coherent sub-sets based on
a priori social-demographic characteristics (e.g., young-old, rural-urban,
eastern-western) or on observed patterns of expressed values (e.g., current
122
-------
versus future, utilization versus preservation, biocentric versus
anthropocentric orientations);
Resolution of conflicts between different value dimensions and/or value
proponents by explicit communication and negotiation among decision
makers and stakeholders.
Candidate methods for ecosystem values assessments
Surveys. Standardized, formal questionnaires may be conducted by mail, telephone,
internet or face-to-face interview. Assessment targets are most often represented by
verbal descriptions or labels, but photographs, videos or computer visualizations can be
used where appropriate. Questions may be presented as multiple distinct items each
focused on one aspect of an assessment target or as multi-dimensional scenarios
conjoining several aspects. Response formats range from binary choices to rankings or
ratings on various value scales to open-ended narratives.
Example: Sheilds et al (2002): multi-item questionnaire, USD A Forest Service, GPRA
Example: Kneeshaw et al (2004): conjoint survey, wildfire risk management options
Example: Ribe et al (2002): perceptual survey, forest management options
Focus groups . Small groups of relevant stakeholders are engaged in facilitated
discussion and deliberation on selected/focused topics relevant to the assessment target.
Typically open-ended narratives are collected and subjected to qualitative analyses to
identify and possibly to ascertain levels of consensus on relevant issues, perspectives and
positions represented by the participants.
Example: Winter et al (2002): wildfire risk management options
Narrative interviews. Individuals nominally representing possible stakeholder
perspectives are asked to comment on broadly defined topics with little direction from the
interviewer/assessor. Open-ended narratives are collected and subjected to qualitative
analyses to explore and articulate the breadth and depth of expressed understandings and
concerns relevant to the assessment target. Included in this category are various
ethnographic methods.
Example: Brandenburg & Carroll (1995): forest management in a local watershed
Behavioral observation/behavior trace. Changes in the patterns of movements and
activities of users or visitors are observed and correlated with changes in aspects of an
environmental setting that are relevant to the assessment target. Behavior may be
observed directly or recorded by cameras, counters or other automated surveillance
technology. Alternatively, persisting traces of visitation or use, such as written
registration lists, vegetation disturbance, soil compaction or erosion, or campfire rings
may be inventoried and analyzed to indicate patterns of behavior.
Example: Daniel & Gimblett (2000): travel patterns in a National Park
Interactive games. Patterns of responses are observed in interactions with simulated
(hypothetical) environments and analyzed to infer preferences and values relevant to
changing features of the environments. Environmental changes may be programmed by
123
-------
the investigator and/or selected or initiated by the respondent. Applications of interactive
games to environmental values assessment are still in the experimental stage.
Example: Bishop & Rohrmann (2003): responses sub-urban park designs
Example References
Bishop, I. D. & Rohrmann, B. (2003) Subjective responses to simulated and real
environments: a comparison. Landscape and Urban Planning, 65: 261-267.
Brandenburg, A.M. & Carroll, M.S. (1995). Your place or mine? The effect of place
creation on environmental values and landscape meanings. Society & Natural
Resources 8(5): 381-398.
Daniel, T.C. & Gimblett, H.R. (2000) Autonomous agents in the park: an introduction to
the Grand Canyon River Trip Simulation Model. International Journal of
Wilderness, 6: 39-43.
Kneeshaw, K., Vaske, J.J., Bright, A.D. & Absher, J.D. (2004) Situational influences of
acceptable wildland fire management actions. Society and Natural Resources,
17:477-489.
Ribe, R.G., Armstrong, E.T., Gobster, P.H. (2002) Scenic vistas and the changing policy
landscape: visualizing and testing the role of visual resources in ecosystem
management. Landscape Journal, 21: 42-66.
Shields, D. J., Martin, I.M., Martin, W.E., Haefele M.A. (2002) Survey results of the
American public's values, objectives, beliefs, and attitudes regarding forests and
grasslands: A technical document supporting the 2000 USDA Forest Service RPA
Assessment. General Technical Report, RMRS-GTR-95. Fort Collins, CO: U.S.
Department of Agriculture, Forest Service, Rocky Mountain Research Station.
Ill p.
Winter, G. & Fried, J.S. (2000) Homeowner perspectives on fire hazard, responsibility,
and management strategies at the wildland-urban interface. Society & Natural
Resources 13: 33-50.
124
-------
12.5. Spatial Representation of Biodiversity and Conservation Values and
Ecological Services
Session Leaders:
Dr. Dennis Grossman, Vice President for Science, Science Division, NatureServe
Dr. James Boyd, Senior Fellow, Director, Energy & Natural Resources Division,
Resources for the Future
Contents:
Spatial Representation of Biodiversity and Conservation Value
Ecosystem Benefit Indicators
Spatial Representation of Biodiversity and Conservation Value
Description of method: This method results in the spatial representation of the
uniqueness and irreplaceability of biological and ecological diversity in a regional
context. This is a scientifically based approach to assign a conservation value to select
species and ecological systems that are representative of an ecological region.
The values are represented as a numeric representation of the uniqueness,
irreplaceability and level of impediment for plant and animal species, vegetation, habitats
and ecological systems.
Key assumptions:
Representative biological and ecological diversity can be elaborated
spatially across any region.
The conservation value (status and quality) of each occurrence can be
ascribed to each element of biodiversity as a repeatable and consistent
procedure.
The cumulative biological and ecological diversity and conservation
values can be practically applied to inform and direct critical resource
management and conservation decisions.
125
-------
Key steps in the method include:
a) Define the biological and ecological targets for valuation
b) Define occurrence standards for each target
c) Define standards for valuing the quality of each occurrence
d) Define standards for measuring range wide status of each target
e) Create a 'conservation value layer' for each target that represents values and goals
of the stakeholder
f) Create 'conservation value summary' of all targets that represents values and
goals of the stakeholder
g) Modify the conservation value through incorporation of threats and opportunities.
Decision contexts where this method could be used:
Enumeration of biodiversity protection implications that result from policy
changes (i.e., change of protection status for isolated wetlands).
Identification of critical riparian habitat
Prioritization of remediation action on superfund sites
Due diligence reviews and EIS as a prerequisite for permitting.
Identification of reference conditions for establishment of baseline quality
metrics for wetland and aquatic habitats.
Assessment of the status of target species and ecosystems.
The method can be applied to a broad range of local to regional to national scales.
The types of data and the spatial representation of this data change relative to the
questions that are being addressed.
Resource inputs and limitations:
The assumption is that there is a sufficient coverage of standardized
biodiversity data required to run these models. The standards have been
developed, and the data required changes associated with the application
questions. Where there is a paucity of required data, it is readily
'developable', but can require the resources complete the required
databases to run the models. The method is useless without good
appropriate data.
This method requires local scientific data, knowledgeable scientific
interpretation and conservation planning expertise. The magnitude of the
need is contingent upon the application and the current state of data and
knowledge.
Lack of data, currency and confidence of data, and data sharing issues
associated with 'sensitive' data, training, and tools are the most important
obstacles to the use of this method. However, there are many ways to
create surrogate datasets that will allow users to adapt to different types of
'barriers'.
126
-------
Uncertainty. There are confidence measures built into the methodology that can
be brought into the decision making process or displayed separately for analysis. The
most significant sources of uncertainty in the use of this method include:
The variability in the quantity and quality of the data.
The limitations of scientific understanding of distribution and quality criteria for some
elements of biodiversity.
Other important dimensions:
The method is adaptable: it can be run repeatedly to represent temporal
change or different landscape scenarios.
Results are commonly aggregated to derive a single benefits number, but
all of the native data is constantly maintained in the system and can be
presented separately.
The output is both understandable and communicable to the interested
audience.
The results are repeatable, and the process and algorithms are very
transparent.
Detailed Description of Method
1. Define the biological and ecological targets for valuation
Biological diversity is often characterized by different levels of biological and
ecological organization, from genes to populations to species to natural communities to
ecosystems and sometimes to ecoregions and biogeographic provinces. All of these
levels can be used for characterization and valuation, but certain levels are most
appropriate to address specific types of assessments. For regional scale valuation,
species, natural communities and ecosystems are generally used for purposes of
conservation assessment and biodiversity valuation.
Within these categories, it is helpful to use the concept of coarse filter and fine
filter conservation elements. The fine filter elements are important biodiversity resources
that often are sparsely distributed across the landscape. These would include imperiled,
declining, endemic, vulnerable, "umbrella" species and subspecies, as well as Focal
Communities such as unique environments, rare plant communities, rare aquatic habitats,
vulnerable species aggregations, migratory stopover points, and others. These fine filter
elements represent those components of biodiversity that can become extinct due to lack
of knowledge or attention. The coarse filter elements are comprised of the broad
vegetation types, habitats and ecological systems that represent aggregations of
communities and natural landscape patterns and processes at scales useful for
management and monitoring. It is by looking at the combination of these fine and coarse
filter element that one can portray the biological and ecological valuation of the
landscape based on well developed and applied standards.
The valuation of fine and coarse filter elements across the landscape required a
defined level for the currency and level of standardization of the knowledge. For
127
-------
example, there needs to be a defined taxonomy for all species and standard classification
approach for all ecological units. NatureServe and the network of state Heritage Program
currently maintain this level of currency and standardization for over 30,000 animal taxa,
56,000 plant taxa, 7,000 vegetation types and 1,500 ecological systems.
2. Define occurrence standards for each target
This methodology then applies the concept of recognizing an area of land and/or
water in which a species or natural community is, or was present. These Element
Occurrences (EOs) have practical conservation value for the Element as evidenced by
potential continued presence and/or regular recurrence at a given location. Biologists and
ecologists have developed criteria and have been conducting inventories for many
decades to document the best occurrences of these elements of conservation across the
landscape. NatureServe databases alone manage and distribute information on the
occurrences of over 500,000 imperiled species across the United States. This number
grows dramatically when adding freshwater and coastal habitats, vegetation types and
ecological systems.
3. Define standards for valuing the quality of each occurrence
Each of the element occurrences defined above must be given a relative quality
rank to allow planners, managers and conservations to prioritize their actions relative to
management of the landscape. Biologists and ecologists have developed an approach to
designate A, B, C, and D quality ranks to these fine and coarse filter occurrences of
conservation elements.
These methods incorporate factors of occurrence size, condition and landscape
integrity. Size factors that are used in this assessment include a quantitative measure of
area of occupancy, population abundance, population density, and population fluctuation.
Condition looks at biotic/abiotic factors, structures, processes within the occurrence as
measured by population reproduction and health, development and maturity, ecological
processes, species composition and biological structure, along with abiotic physical and
chemical factors. Landscape integrity compiles a qualitative measure of biotic factors,
abiotic factors, and processes surrounding the EO. These factors include landscape
structure and extent, community development and maturity, intactness of ecological
processes, species composition and biological structure, and additional abiotic physical
and chemical factors.
Many of coarse and fine filter occurrence quality metrics have been developed
and used to provide a quality/integrity attribute to all occurrences. The quality ranks
portray what experts determine to be within acceptable ranges of variation. These ranges
are developed through the characterization of multiple, apparently undisturbed examples,
examination of impact and response to human-induced alterations, review of literature
and historical records, and the development and testing of ecological simulation models.
"A" ranked occurrences are within the preferred ecological integrity threshold. "B"
ranked occurrences have one key factor within its acceptable range of variation. "C"
ranked occurrences do not have any key factors with their acceptable range of variation,
128
-------
but they are still considered to be 'restorable'. "D" ranked occurrences are no longer
restorable. In some cases these factors can be directly measured, while in other they may
be inferred/estimated indirectly.
4. Define standards for measuring range-wide status of each target
The next step in this approach is to assign a range-wide conservation status rank
to each of the conservation elements. This is primarily completed and is most useful as
an element attribute at the global scale, but the standards can also be applied at the
national, sub-national and local scales. The conservation rank factors differ as they are
applied to species as compared to ecological communities and habitats.
For species, the factors that are considered in assessing conservation status
include total number and condition of occurrences (e.g., populations); population size;
range extent and area of occupancy; short- and long-term trends in the above factors;
scope, severity, and immediacy of threats; number of protected and managed
occurrences; intrinsic vulnerability and environmental specificity.
For ecological communities, there are primary and secondary factors used in
assessing conservation status. The primary factors for assessing community status are the
total number of occurrences (e.g., forest stands) and the total acreage occupied by the
community. The secondary factors for assessing community status are the geographic
range over which the community occurs, long-term trends across this range, short-term
trend (i.e., threats), degree of site/environmental specificity exhibited by the community,
and the impediment or rarity across the range as indicated by sub-national ranks assigned
by local natural heritage programs.
The definitions for each of the Global (G) Ranks are:
G1 - Critically imperiled: At very high risk of extinction due to extreme
rarity (often 5 or fewer populations), very steep declines, or other factors.
G2 - Imperiled: At high risk of extinction due to very restricted range,
very few populations (often 20 or fewer), steep declines, or other factors
G3 - Vulnerable: At moderate risk of extinction due to a restricted range,
relatively few populations (often 80 or fewer), recent and widespread
declines, or other factors
G4 - Uncommon but apparently secure: Uncommon but not rare; some
cause for long-term concern due to declines or other factors
G5 - Widespread, abundant and secure: Common; widespread and
abundant
129
-------
All fine and coarse filter conservation elements across North America have been
evaluated and given a conservation status rank.
5. Create a 'conservation value layer' for each target that represents values and goals
of the stakeholder
The biodiversity value attributes that have been created for the global range-wide
conservation status and the quality of viable occurrences now allows the development of
a conservation value surface layer for each individual conservation element. The creation
of this layer requires the ability to spatially portray each of the occurrences as well as the
quality and confidence of each occurrence. The spatial portrayal of element occurrences
is derived from imagery, maps and field points, along with modeled distributions of
specific elements. The element quality attributes are imported directly as available, and
generated from landscape integrity models when necessary.
6. Create 'conservation value summary' of all targets that represents values and
goals of the stakeholder
The combination of 'conservation value layers' for selected elements across a
planning or assessment jurisdiction creates an aggregated 'conservation value summary'
that provides a spatially explicit representation of the biodiversity and conservation
values that are important to the conservation and resource management community.
Different user groups can select the types of elements that there need to assess across the
jurisdiction, and they can also modify the relative conservation weight of each fine and
coarse filter conservation element. This will provide a customized conservation surface
that portrays the values that they will need to incorporate into their planning and
assessment work. This also becomes a baseline for monitoring the effects of their
programs to manage for biodiversity value over time.
7. Modify the conservation value through incorporation of threats and opportunities
in order to prioritize conservation and resource management activities.
The conservation values that are generated through processes 1-6 can be modified
to reflect values that are relevant to a specific assessment. Zoning policies, growth
models, economic values, ecological services and other values help to identify the effect
of different or future scenarios relative to the current or desired future condition of the
landscape.
Key Citations
Stoms, D. M., P. J. Comer, P. J. Crist and D. H. Grossman. 2005. Choosing surrogates for
biodiversity conservation in complex planning environments. Journal of
Conservation Planning 1: 44-63.
Grossman, D.H. and P.J. Comer. 2004. Setting Priorities for Biodiversity Conservation in
Puerto Rico. NatureServe Technical Report.
Brown, N., L. Master, D. Faber-Langendoen, P. Comer, K. Maybury, M. Robles, J.
Nichols, and T. B. Wigley. 2004. Managing Elements of Biodiversity in
130
-------
Sustainable Forestry Programs: Status and Utility of Nature Serve's Information
Resources to Forest Managers. National Council for Air and Stream Improvement
Technical Bulletin Number 0885.
Riordan, R. and K. Barker. 2003. Cultivating biodiversity in Napa. Geospatial
Solutions.
Ecosystem Benefit Indicators (Boyd And Banzhaf, 2005, Banzhaf and Boyd, 2005, Boyd
and Wainger 2002; Boyd 2004)
Because many ecosystem services are public goods, markets are not available to
provide clear units of account. Cost-benefit analysis and national accounting for
marketed goods is made easier by clear units of account: namely, the end-products
consumers enjoy. While environmental economics has grappled for decades with the
challenge of missing prices for environmental goods and services, it has neglected
another central issue: the consistent definition of the environmental units to which prices
are attached. An argument for standard units of account is that they can facilitate the
transfer of valuations across the landscape and across time.
Possible use by EPA
Clear units of account are also desirable from the standpoint of environmental
programs that police gains and losses in environmental quality or economic value.
Consider wetland banking, water permit trading, land swaps, and natural resource
damage assessment. All such activities trade compound, bundled environmental goods.
Ideally, however, what should be traded - and accounted for - are the individual
environmental goods and services provided by the bundle. In practice, however, trade
and compensation programs use blunter proxies, such as "acres of wetland" or "pounds of
nitrogen." What is lost in this kind of accounting system is gains and losses in individual
ecosystem services.
Standardized units of account are also important to the measurement of
performance. If the nation's environmental status is to be characterized and tracked over
time units must be clearly defined, defensible ecologically and economically, and
consistently measured. At present, the government and the public are presented with an
over-abundance of units of measurement and often those units are poorly defined, unclear
in their origin, and exacerbate the divide between economic and ecological analysis.
Often within a single agency there are multiple competing paradigms for what should be
measured.
At the national level, in the evaluation of new rules as part of the RIA process,
government performance reviews, strategic planning, budget justification, and priority
setting. They are also applicable at more local scales as a tool to improve regional and
local planning, such as watershed planning in the context of TMDLs.
Description of the Method: Units of Account
131
-------
There are two principle activities associated with the method. First, the definition
and measurement of ecosystem service units (quantity measures of services). Second, the
use of benefit indicators (or "willingness to pay indicators" to facilitate transfer of benefit
estimates across the landscape or to empower tradeoff analysis by regulators, planners,
and conservancies.
Analysis of the benefits of natural resources requires a distinction between
ecosystem components, processes, functions and services. The term services originates
in economics, but has been adopted within ecology as well to signify the connection
between ecosystems and human wellbeing.2 Ecosystem components include resources
such as surface water, oceans, vegetation types, and species. Ecosystem processes and
functions are the biological, chemical, and physical interactions associated with
ecosystems. These functions are the things described by biology, atmospheric science,
hydrology, and so on.
Ecosystem services arise from these components and functions but are different:
Ecosystem services are the end products of nature that yield human wellbeing. Part of
this definition is particularly important: namely, that ecosystem services are "end
products." End products are the environmental components about which people make
choices. It is important to emphasize that many aspects of nature are valuable, but are
not capable of being valued in an economic sense - because they are not associated with
social or individual choices.3
This definition restricts the units of account, relative to many ways in which
ecosystem services are commonly used. For example, nutrient cycling is often termed n
ecosystem service. This is not a service, however, but rather an ecological function. To
be sure, it is a valuable function, but it an intermediate aspect of the ecosystem and not an
end product. Being valuable is not the same thing as being a service.
Consider another example. Reference is often made to recreation being an
ecosystem service. It is not. Recreation is a benefit that relies on ecosystem services as
inputs. Recreation is the joint product of ecosystem services including surface waters and
fish populations and other goods and services including tackle, boats, time allocation,
and access. From an economic standpoint, units of ecosystem account will exclude many
things that are called ecosystem services.
Note that the above examples of economically defined units of account lead to
units that are in fact biophysical, rather than "economic" in nature. An economic
definition therefore leads naturally and necessarily to a bridge between economic and
biophysical analysis. No ecologist should think that the economic definition of services
leads away from biophysical analysis. In fact, the opposite is true.
The relationship of units of account to "environmental indicators" is as follows.
2 See Gretchen Daily, Nature's Services.
3 Many components of an ecosystem can be thought of as "intermediate products" in that they are
necessary to the production of services, but are not services themselves.
132
-------
First, the units of account described above are themselves indicators of performance or
environmental conditions. These units are countable, spatially explicit indicators of
certain biophysical characteristics. They can be expressed both numerically and spatially
via geospatial information systems. Thus, our units of account, or ecosystem service
indicators, are related to certain "ecological indicators" emanating from the biophysical
sciences.
Description of the Method: Willingness to Pay Indicators
However, we will also relate units of account to a different type of indicator:
indicators of willingness to pay. In accounting for conventional, market goods, market
prices are used to "weight" units of account. Because many ecosystem services lack
these prices, how are units of account to be weighted? This question is central to benefit-
cost analysis and welfare accounting. It should also, arguably, be central to government
performance assessment and the evaluation of environmental trades, though preservation
or enhancement of economic value is not always the aim of such programs. The
aspiration of economic analysis is willingness to pay-based weights. For this reason, the
workshop will also address the derivation of weights that can be assigned to ecosystem
units of account.
The principal observation here is that the value of ecosystem services is highly
dependent upon location in the biophysical and social landscape. In conventional
accounting, arbitrage allows us to assume a single market price. For many ecosystem
services there is no arbitrage. Also, many ecological services are best thought of as
differentiated goods with important place-based quality differences. Ecosystem services'
scarcity, substitutes, and complements are likewise spatially differentiated.
There are several implications. First, units of account should be spatially explicit.
Second, the weights assigned to units - if units are to be aggregated into summary
measures - should be spatially explicit. This can mean several things, depending upon
the valuation method being applied. For example, stated preference techniques can be
used to place value on units of account using place-specific scenarios. In other words,
the scenarios presented in stated preference surveys could rely on standardized units and
ways of measuring place-based quality, substitution, and complementary asset landscape
factors. Alternatively, meta-analysis of existing value estimates can be used to calibrate
benefit transfers. Standardized service units and location-specific factors affecting
willingness to pay would provide a consistent architecture for such an exercise.4 An
alternative approach is a reduced-form regression of willingness to pay on various
factors, including landscape-dependent indicators of the contribution of ecosystems to
final goods and services and landscape-dependent indicators of substitutes and
complements, population, and other socio-demographic characteristics.5
4 This topic was raised at NCEE's workshop on benefit transfer in Spring, 2005.
5 Willingness to pay, while not directly observable, is a function of various characteristics that are
observable. WTP weights p, can be thought of as a function of landscape indicators I. In principle, this
function, on a service-by-service basis, can be calibrated by relating observable indicators / to existing
WTP estimates of service value. Were this possible in practice, location and ecosystem-specific indicators
/ could be used to transfer monetary WTP estimates to locations where they are not available.
133
-------
Finally, there is relevance to less econometrically formal weighting procedures.
Examples here include stakeholder-driven decisions, citizen juries, and mediated
modeling exercises. In these examples weights are not derived by economic analysis, but
rather are debated and concluded via some kind of institutional process. Here too,
standardized units of account and landscape willingness to pay indicators could help
educate and discipline benefit assessment.
Strengths and Weaknesses
Both ecosystem service measurements (indicators) and benefit indicators are a
quantitative and visual, but not monetary, approach to the assessment of services. Unless
married to an econometric benefit transfer exercise, conjoint analysis, citizen jury, or
other weighting approach, the indicators will not themselves yield a single dollar-based
"answer." Rather, they should be though of as an accounting tool to measure and track
over time, in a consistent manner, changes in service levels and factors related to
willingness to pay for those services. The monetization of benefits, which is clearly
important in certain regulatory applications, demands additional methods.
Service and benefit indicators are simple, countable aspects of the biophysical and
social environment. They are transparent and easily replicable. Because indicators are
cheaper to generate than econometric value estimates they better allow for landscape
assessment of multiple services at large scales.
EBIs are drawn mainly from geospatial data, including satellite imagery. Data can
come from state, county, and regional growth, land-use, or transportation plans; federal
and state environmental agencies; private conservancies and nonprofits; and the U.S.
Census. Benefit indicators can capture the landscape, or spatial, factors that contribute to
social well-being. This is in fact a virtue of indicator methods. Indicators can be derived
from and mapped within a GIS context. Spatial analysis is important because the
ecological production function is a function of spatial interdependencies. From an
economic standpoint, the social determinants of service benefits depend upon the
landscape context in which those services arise. The consumption of services often
occurs over a wide scale. Habitat support for recreational and commercial species, water
purification, flood damage reduction, crop pollination, and aesthetic enjoyment are all
services typically enjoyed in a larger area surrounding the ecosystem in question.
The method is applicable to the full range of ecological services. In practice,
applicability may be limited by data gaps.
The principle disadvantage of indicators alone, is that they do not directly yield
dollar-based ecological benefit estimates. They also do not in themselves weight or
estimate the tradeoffs associated with different factors relating to benefits (though as
noted above they can be married to more formal methods designed to do such weighting).
This is not really a weakness to indicators themselves, but rather an acknowledgement
that more must occur than simple indicator measurement if the goal is dollar-based end-
results.
134
-------
Uncertainties associated with the method and how they would be addressed: A
core rationale for the use of a benefit indicator approach is to explicitly convey the
sources of complexity - and hence uncertainty - characterizing biophysical systems and
the service flows arising from them. The visual depiction of willingness to pay
indicators, for example, can mimic sensitivity analysis by presenting a range of benefit
scenarios in GIS form. However, the visual depiction of quantitative information
introduces uncertainties of its own. In particular, visual depictions can strongly influence
perceptions. Uncertainty with regard to how indicators are perceived, particularly when
presented visually should be acknowledged.
Key Citations
Boyd, James, and Spencer Banzhaf. 2005. What are Ecosystem Services? Resources for
the Future.
Banzhaf, Spencer, and James Boyd. 2005. The Architecture and Measurement of an
Ecosystem Services Index. Resources for the Future Discussion Paper 05-22.
Boyd, James, and Spencer Banzhaf. 2005. Ecosystem Services and Government
Accountability. Resources magazine, summer.
Boyd, James. 2004. What's Nature Worth? Using Indicators to Open the Black Box of
Ecological Valuation. Resources magazine, summer.
Boyd, James, and Lisa Wainger. 2002. Landscape Indicators of Ecosystem Service
Benefits. American Journal of Agricultural Economics.
135
------- |