Ecosystems and Human Well-Being: A Manual for - Unep-WCMC

a • n • d
Human Well-Being
A Manual for Assessment Practitioners
A Manual for
and Human
A Manual for
Assessment Practitioners
Neville Ash
Hernán Blanco
Claire Brown
Keisha Garcia
Thomas Henrichs
Nicolas Lucas
Washington | Covelo | London
Ciara Raudsepp-Hearne
R. David Simpson
Robert Scholes
Thomas P. Tomich
Bhaskar Vira
Monika Zurek
Copyright © 2010 Neville Ash, Hernán Blanco, Claire Brown,
Keisha Garcia, Thomas Henrichs, Nicolas Lucas, Ciara Raudsepp-Hearne,
R. David Simpson, Robert Scholes, Thomas P. Tomich, Bhaskar Vira,
and Monika Zurek
All rights reserved under International and Pan-American Copyright
Conventions. No part of this book may be reproduced in any form or
by any means without permission in writing from the publisher: Island
Press, 1718 Connecticut Avenue NW, Suite 300, Washington, DC
20009, USA.
Island Press is a trademark of The Center for Resource Economics.
Library of Congress Cataloging-in-Publication Data
Ecosystems and human well-being : a manual for assessment
practitioners / Neville Ash ... [et al.].
p. cm.
Includes bibliographical references and index.
ISBN 978-1-59726-711-3 (pbk. : alk. paper) —
ISBN 978-1-59726-710-6 (cloth : alk. paper)
1. Human ecology. 2. Ecosystem management. 3. Biodiversity. I. Ash,
GF50.E2617 2010
333.71'4—dc22 2009043012
Printed on recycled, acid-free paper
Manufactured in the United States of America
10 9 8 7 6 5 4 3 2 1
ix Foreword
xi Preface
xiv Acknowledgments
xvi Acronyms and Abbreviations
1 Chapter 1 Assessing Ecosystems, Ecosystem Services, and
Human Well-being
Neville Ash, Karen Bennett, Walter Reid, Frances Irwin,
Janet Ranganathan, Robert Scholes, Thomas P. Tomich, Claire Brown,
Habiba Gitay, Ciara Raudsepp-Hearne, and Marcus Lee
33 Chapter 2 Stakeholder Participation, Governance, Communication,
and Outreach
Nicolas Lucas, Ciara Raudsepp-Hearne, and Hernán Blanco
71 Chapter 3 Conceptual Frameworks for Ecosystem Assessment:
Their Development, Ownership, and Use
Thomas P. Tomich, Alejandro Argumedo, Ivar Baste, Esther Camac,
Colin Filer, Keisha Garcia, Kelly Garbach, Helmut Geist,
Anne-Marie Izac, Louis Lebel, Marcus Lee, Maiko Nishi,
Lennart Olsson, Ciara Raudsepp-Hearne, Maurice Rawlins, Robert
Scholes, and Meine van Noordwijk
115 Chapter 4 Assessing State and Trends in Ecosystem Services and
Human Well-being
Robert Scholes, Reinette Biggs, Cheryl Palm, and Anantha Duraiappah
151 Chapter 5 Scenario Development and Analysis for Forward-looking
Ecosystem Assessments
Thomas Henrichs, Monika Zurek, Bas Eickhout, Kasper Kok,
Ciara Raudsepp-Hearne, Teresa Ribeiro, Detlef van Vuuren, and
Axel Volkery
221 Chapter 6 Assessing Intervention Strategies
R. David Simpson and Bhaskar Vira
255 Index
Commissioned by the United Nations Secretary–General in 2000, and completed
in 2005, the Millennium Ecosystem Assessment (MA), based on the findings of 34
“sub-global” assessments carried out in a diverse set of ecosystems in sites around
the world, provides a state-of-the-art appraisal of the condition and trends in the
world’s ecosystems and the services they provide.
The MA presents compelling evidence that underlines the urgency and necessity
of restoring, conserving, and sustainably managing our ecosystems. Most important,
the assessment shows that, with appropriate actions, it is possible to reverse the degradation of many ecosystem services over the next 50 years. By providing invaluable
information to policy makers, the MA seeks to help ensure that the required changes
in current policy and practice undertaken will be evidence based and informed by
the best available scientific analysis.
This manual, Ecosystems and Human Well-being: A Manual for Assessment
Practitioners, allows for the wider adoption of the MA conceptual framework and
methods. The manual, which contains numerous case studies of best practice, offers
a practical guide for undertaking ecosystem assessments and includes tools and approaches that can assess options for better managing ecosystems.
UNEP and UNDP, working together with other partners, are committed to
promoting sustainable development and ensuring the protection of our planet. By
stimulating future ecosystem assessments, based on the proven methodologies of the
MA, it is our hope that this manual will provide the knowledge needed to develop
appropriate and effective policies and strategies to ensure that the earth’s ecosystems
and their vital services are restored and preserved. Our very livelihoods depend on
Achim Steiner, Executive Director
United Nations Environment
Programme (UNEP)
Helen Clark, Administrator
United Nation Development Programme
The Millennium Ecosystem Assessment
The Millennium Ecosystem Assessment (MA) was called for by the United Nations
Secretary-General Kofi Annan in 2000 in his report to the UN General Assembly,
We the Peoples: The Role of the United Nations in the 21st Century. The MA was
carried out between 2001 and 2005 to assess the consequence of ecosystem change
for human well-being, by attempting to bring the best available information and
knowledge on ecosystem services to bear on policy and management decisions. The
MA established the scientific basis for action needed to enhance the conservation
and sustainable use of ecosystems and their contribution to human well-being. The
MA was in part a global assessment, but to facilitate better decision making at all
scales, 34 regional, national and local scale assessments (or sub-global assessments)
were included as core project components. Since the release of the MA, further subglobal assessments have started.
What are ecosystems and ecosystem services?
An ecosystem is a dynamic complex of plant, animal and micro-organism communities and the nonliving environment interacting as a functional unit. The conceptual
framework for the MA assumes that people are integral parts of ecosystems. The
MA Report itself focuses on linkages between ecosystems and human well-being, in
particular on “ecosystem services,” which are the benefits that people obtain from
Ecosystem services include the following:
• Provisioning services, such as providing food, water, timber and fibre;
• Regulating services, such as the regulation of climate, floods, disease, wastes
and water quality;
• Cultural services, such as offering recreational, aesthetic, and spiritual benefits; and
• Supporting services, such as soil formation, photosynthesis, and nutrient cycling.
What is ecosystem assessment?
An ecosystem assessment provides the connection between environmental issues
and people. An assessment of ecosystem services needs to consider both the ecosystems from which the services are derived and also the people who depend on and
are affected by changes in the supply of services, thereby connecting environmental
and development sectors. Assessments play numerous roles in the decision-making
process, including responding to decision makers’ needs for information, highlighting trade-offs between decision options, and modeling future prospects to avoid
xii | Preface
unforeseen long-term consequences. They inform decisions by providing critical
judgment of options and uncertainty and through synthesizing and communicating
complex information on relevant issues. They are also of value through the process
they involve, engaging and informing decision makers long before final assessment
products are available.
Thus, decision makers—including those whose goals and actions are focused
on people, society, and economics—can benefit from examining the extent to which
achieving their goals depends on ecosystem services. Assessments can provide credible and robust information on the links between ecosystems and the attainment of
economic and social goals.
Why is this Manual needed?
This Manual makes the methods of the MA and associated sub-global (local and
regional) assessments widely accessible. While the MA is the most comprehensive
assessment of ecosystems carried out to date, there are other related assessment processes such as Global Environment Outlook (GEO), Global International Waters
Assessment (GIWA), Intergovernmental Panel on Climate Change (IPCC), Land
Degradation Assessment in Drylands (LADA), International Assessment of Agricultural Knowledge, Science and Technology for Development (IAASTD) and World
Water Assessment. Lessons learned from these assessments supplement the best
practice of ecosystem assessment identified through the MA. The publication of this
Manual aims to encourage more assessments at scales which are relevant to policy
and decision makers.
Why use this Manual?
The Manual is intended to be a “how to” guide for undertaking ecosystem assessments. The Manual contains detailed guidance on conceptual frameworks, assessing
status and trends of ecosystems, developing and using scenarios, assessing policy
options, and the process for establishing, designing and running an ecosystem assessment, including communications and outreach.
The priority audience for the Manual are individuals who are responsible for
designing and carrying out environmental or developmental assessments, and individuals responsible for building capacity for ecosystem assessments, either through
structured training (such as through developing curricula relating to ecosystem services and development) or assistance in conducting assessments on the ground.
New and emerging ecosystem assessment practitioners should use this Manual
• Familiarize themselves with the concept of ecosystem assessment;
• Understand how and why an ecosystem assessment can benefit decision making
at their scale of interest and what steps are involved;
• Improve capacity to undertake an assessment where the need for one has already been identified; and
• Act as a guide for practitioners who are undertaking an assessment to obtain
more background information and identify sources of potential assistance with
challenging areas.
Preface | xiii
Experienced ecosystem assessment practitioners should use this Manual to:
• Update and complement their knowledge and skills in ecosystem assessment;
• Serve as a basis for dialogue on methods for ecosystem assessment to improve
the shared knowledge base on this approach; and
• Train new and emerging ecosystem assessment practitioners in an applied or
classroom setting.
This Manual complements other related resources such as the Ecosystem Services:
A Guide for Decision-makers, prepared by WRI and others (which focuses on how
the findings of ecosystem assessments can be used), and the UNEP-GEO assessment
training modules (which cover a much broader institutional State of the Environment reporting process).
Authors of the Millennium Ecosystem Assessment (MA) were the recipients of the
Zayed International Prize for the Environment prize in honor of His Highness Sheik
Zayed Bin Sultan Al Nahyan, President of the United Arab Emirates and Governor
of Abu Dhabi. On agreement of the Board of the MA, a significant portion of these
funds was allocated to the production of this Manual. Financial support for the
Manual was also provided by the UK Department for Environment, Food and Rural
The support of the following institutions enabled the participation of lead authors in compiling this Manual: IUCN, Switzerland; RIDES, Chile; The Cropper
Foundation, Trinidad and Tobago; National Environmental Research Institute, Denmark; European Environment Agency (EEA), Denmark; Secretariat of Environment
and Sustainable Development, Argentina; McGill University, Canada; Council for
Scientific and Industrial Research, South Africa; National Center for Environmental
Economics, USA; University of California, Davis, USA; University of Cambridge,
UK; FAO, Italy; and United Nations Environment Programme World Conservation
Monitoring Centre (UNEP-WCMC), UK.
In addition to the lead authors, contributions to the Manual were provided by
Alejandro Argumedo (Asociacion Andes), Ivar Baste (UNEP), Karen Bennett (WRI),
Esther Camac (Association IXACAVAA for Indigenous Development and Information), Anantha Duraiappah (UNEP), Bas Eickhout (Netherlands Environment
Assessment Agency), Colin Filer (Australian National University), Kelly Garbach
(University of California, Davis), Helmut Geist (University of Aberdeen), Habiba
Gitay (World Bank Institute), Kasper Kok (University of Wageningen), Frances Irwin (WRI), Anne-Marie Izac (Consultative Group for International Agricultural
Research), Louis Lebel (Chiang Mai University), Marcus Lee (The World Bank),
Walter Reid (David and Lucile Packard Foundation), Maiko Nishi (United Nations
University), Meine van Noordwijk (World Agroforestry Centre), Lennart Olsson
(Lund University), Cheryl Palm (Columbia University), Janet Ranganathan (WRI),
Maurice Rawlins (The Cropper Foundation), Teresa Ribeiro (European Environment Agency), Axel Volkery (European Environment Agency), and Detlef van
Vuuren (Netherlands Environment Assessment Agency).
All of the MA authors and review editors who contributed to this manual through
their contributions to the MA itself, upon which this manual is based, are acknowledged. The MA process included a total of 34 sub-global assessments (SGAs) from
around the world. These assessments analyzed the importance of ecosystem services
for human well-being at local, national, and regional scales. The experience of the
team involved in carrying out or coordinating SGAs also contributed to the content
reflected in the manual.
Acknowledgments | xv
The Manual Chapter Review Editors provided valuable guidance and input:
John Agard (University of West Indies); Yolanda Kakabadze (Fundación Futuro
Latinoamericano), Richard Norgaard (University of California, Berkeley), David
Stanners (European Environment Agency), Dan Tunstall (WRI) and Tom Wilbanks
(Oak Ridge National Laboratory).
Many individuals reviewed drafts of the Manual and provided critical review
comments: Heidi Albers (Oregon State University), Salvatore Arico (UNESCO), Jan
Bakkes (Netherlands Environmental Assessment Agency), Karen Bennett (WRI),
Reinette Biggs (Stockholm Resilience Institute), Traci Birge (UNEP DEWA), Steve
Carpenter (University of Wisconsin), David Cooper (Secretariat for the Convention
on Biological Diversity), Steven Cork (EcoInsights), Anantha Duraiappah (UNEP),
Paul Ferraro (Georgia State University), Max Finlayson (Charles Sturt University),
Inger Heldal (Norad—Norwegian Agency for Development Cooperation), Bente
Herstad (Norad—Norwegian Agency for Development Cooperation), Robert Höft
(Secretariat for the Convention on Biological Diversity), Frances Irwin (WRI), Annekathrin Jaeger (EEA), Christian Layke (WRI), Marcus Lee (UNEP, now World
Bank), Markus Lehmann (Secretariat for the Convention on Biological Diversity),
Jock Martin (EEA), Connie Musvoto (CSIR), Gerald Nelson (University of Illinois at
Urbana-Champaign), David Niemeijer, Cheryl Palm (The Earth Institute, Columbia
University), Bely Pires, Janet Ranganathan (WRI), Walter Reid (David and Lucile
Packard Foundation), Belinda Reyers (CSIR), Dale Rothman (International Institute for Sustainable Development), Frederik Schutyser (EEA), Albert van Jaarsveld
(National Research Foundation, South Africa), Hans Vos (EEA), Rodrigo Victor,
Ernesto Viglizzo, Angela Wilkinson (James Martin Institute, Oxford University),
Sven Wunder (CIFOR), Kakuko Nagatani Yoshida (UNEP Regional Office for Latin
America and the Caribbean) and the participants in the first meeting of sub-global
assessment practitioners held in Kuala Lumpur, April 2008.
Translation of the manual into French was sponsored by the UNDP/UNEP Poverty and Environment Initiative. The Spanish translation was sponsored by the European Environment Agency.
UNEP-WCMC coordinated the preparation and publication of the manual.
Thanks go to Claire Brown, Matt Walpole, Philip Bubb, and Jessica Jones for
their efforts. Keisha Garcia from the Cropper Foundation was hosted as a fellow
at UNEP-WCMC and advanced the manual during her tenure and beyond. Linda
Starke’s expert eye to the text as copyeditor has been an invaluable contribution.
Acronyms and Abbreviations
Advisory Committee
agricultural knowledge, science and technology
Fourth Assessment Report (for the IPPC)
the Alternatives to Slash-and-Burn Program
Advanced Terrestrial Ecosystem Analysis and Modelling
the Business Biodiversity Offset Program
California Agroecosystem Assessment
cost-benefit analysis
Convention on Biological Diversity
cost-effectiveness analysis
Cation exchange capacity
chief executive officer
Conference of the Parties
Conservation Reserve Program
Development Assistance Committee
Dahlem Desertification Paradigm
U.K. Department for International Development
the European Commission
European Environment Agency
Gross Domestic Product
Global Environment Outlook
Geographic Information System
Global International Waters Assessment
Gross National Income
Heavily Indebted Poor Countries
Human Well-Being
International Assessment of Agricultural Science and Technology for
International Energy Agency
International Labour Organization
International Monetary Fund
Integrated natural resource management
intergovernmental panel on climate change
Inter-Parliamentary Union
Acronyms and Abbreviations | xvii
International Telecommunication Union
International Union for Conservation of Nature
Joint Inspection Unit
Land Degradation Assessment in Drylands
Least Developed Countries
Laguna Lake Development Authority
Land-Use/Cover Change Programme
Millennium Ecosystem Assessment
market-based incentives
multi-criteria analysis
Millennium Development Goals
nongovernmental organization
Official Development assistance
Ozone depletion potential
Organisation for Economic Co-operation and Development
payments for ecosystem services
Papua New Guinea
purchasing power parity
PRospective Environmental analysis of Land Use Development in Europe
Pressure State Response
Southern Africa Millennium Ecosystem Assessment
summary for decision makers
Self-Employed Women’s Association
summary for policy makers
Special Report on Emissions Scenarios
social-cultural, technological, economic, environmental, and political
driving forces
terms of reference
United Nations
The Joint United Nations Programme on HIV/AIDS
United Nations Convention to Combat Desertification
United Nations Conference on Trade and Development
United Nations Development Programme
United Nations Environment Programme
United Nations Educational, Scientific and Cultural Organization
United Nations Framework Convention on Climate Change
United Nations Children’s Fund
United Nations Statistic Division
University of the Philippines Los Baños
United States of America
World Health Organisation
World Resources Institute
World Trade Organisation
Assessing Ecosystems, Ecosystem
Services, and Human Well-being
Neville Ash, Karen Bennett, Walter Reid, Frances Irwin,
Janet Ranganathan, Robert Scholes, Thomas P. Tomich, Claire
Brown, Habiba Gitay, Ciara Raudsepp-Hearne, and Marcus Lee
What is this chapter about?
This chapter provides an overview of the process and components of scientific assessments
that have as their focus or include within their scope the connections between ecosystems
and people. It introduces ecosystem services as the link between ecosystems and human
well-being and therefore as the focus of assessing the consequences of ecosystem changes
for people. The chapter introduces and highlights the relationship between the various components of assessment. In doing so, it provides an introduction and roadmap to the subsequent
chapters of the manual.
1.1 Introduction
Section’s take-home messages
• This manual can be used as a whole document, or individual chapters can help
assessment practitioners who are looking for guidance on particular aspects of the
• Assessments are not just about the findings. Getting the process right, from the early
stages of design through to the communication of findings, is essential in order to have an
This manual is a stand-alone “how-to” guide about conducting an assessment of
the consequences of ecosystem change for people. However, the manual also relates
closely to other recent publications, particularly Ecosystem Services: A Guide for
Decision Makers (WRI 2008), which presents methods for public-sector decision
makers to use information on ecosystem services to strengthen economic and social development policies and strategies. This manual can be used as a whole document, or individual chapters can help assessment practitioners who are looking for
2 | A Manual for Assessment Practitioners
guidance on particular aspects of the process. The manual builds on the experiences
and lessons learned from the Millennium Ecosystem Assessment (MA) global assessment and from over 30 ongoing or completed sub-global assessment initiatives
at a range of scales, including local, national, and regional assessments. (See www. for further details on the MA and the various follow-up activities currently under way.) It also includes insight and experiences gained from a wider range
of assessment activities focused on ecosystem services.
The chapter begins with an overview of such assessments—what they are and
why they are useful—and then provides a summary of the step-by-step process for
conducting an assessment. Drawing on both theory and best practice from the field
and on a range of global and sub-global assessments, the chapter highlights the importance not just of the findings of an assessment but also of the process itself. Getting the process right, from the early stages of design through to the communication
of findings, is essential in order to have an impact on the intended audience.
This manual has been written to support integrated ecosystem assessment practitioners. However, it is essential that the assessment practitioner also understand
the decision-making context in which the study is being conducted and into which
the findings may be taken on board. As such, the chapter concludes with a short
section on how assessments can be considered in the context of the decision-making
process and how the focus and impact of an assessment will depend on what stage
an issue is in its policy life cycle.
Subsequent chapters in the manual elaborate on the material presented here and
address key aspects of the assessment process: engaging stakeholders; developing
and using a conceptual framework; conducting assessments of conditions and trends
in ecosystems, their services, and human well-being; developing scenarios of change
for ecosystems, their services, and human well-being; and assessing responses or
interventions that aim to improve the management of ecosystems for people. Figure
1.1 outlines the main contents and layout of this manual, and shows how key sections of the manual relate.
1.2 How to improve decision making using ecosystem assessments
Section’s take-home messages
• An ecosystem services assessment can help build a bridge between the development
and environmental communities by providing credible and robust information on the links
between ecosystem management and the attainment of economic and social goals.
• As improvements are made in describing and valuing the benefits of ecosystem services,
decision makers can better understand how their actions might change these services,
consider the trade-offs among options, and choose policies that sustain the appropriate
mix of services.
• Successful assessments share three basic features: they are credible, legitimate, and
relevant to decision makers’ needs.
People everywhere depend on ecosystems for their well-being. Ecosystems are the
source of obvious necessities such as food and fresh water, but they also provide
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 3
Figure 1.1. Contents and layout of the manual.
less obvious services such as flood protection, pollination, and the decomposition of
organic waste. The natural world provides spiritual and recreational benefits as well.
These and other benefits of the world’s ecosystems have supported the extraordinary growth and progress of human societies. Yet the Millennium Ecosystem Assessment found that the majority of ecosystem services are in a state of decline and can
no longer be taken for granted. Ignoring the links between ecosystems and human
well-being in public and private decision making puts at risk our ability to achieve
long-term development goals. An assessment of ecosystem services provides the connection between environmental issues and people. Thus, decision makers—including
those whose goals and actions are focused on people, society, and economics—can
benefit from examining the extent to which achieving their goals depends on ecosystem services (see Table 1.1).
Reconciling economic development and nature is challenging because they have
traditionally been viewed in isolation or even in opposition, and the full extent of
humanity’s dependence on nature’s benefits, or ecosystem services, is seldom taken
into account by development or environmental communities. An ecosystem services
Table 1.1. Linking development goals and ecosystem services
Dependence on ecosystem services
Ecosystem services such as food production, water purification, and
disease regulation are vital in reducing child mortality, improving
maternal health, and combating diseases. In addition, changes in
ecosystems can influence the abundance of human pathogens,
resulting in outbreaks of diseases such as malaria and cholera
and the emergence of new diseases.
Increasingly, people live in areas that are vulnerable to extreme
events such as floods, severe storms, fires, and droughts (MA
2005:443). The condition of ecosystems affects the likelihood
and the severity of extreme events by, for example, regulating
global and regional climates. Healthy ecosystems can also lessen
the impact of extreme events by regulating floods or protecting
coastal communities from storms and hurricanes.
Natural hazard
Adaptation to
Climate change alters the quantity, quality, and timing of ecosystem
service flows such as fresh water and food. These changes
create vulnerabilities for those individuals, communities, and
sectors that depend on the services. Healthy ecosystems can
reduce climate change impacts. Vegetation provides climateregulating services by capturing carbon dioxide from the
atmosphere. Ecosystem services such as water and erosion
regulation, natural hazard protection, and pest control can help
protect communities from climate-induced events such as
increased floods, droughts, and pest outbreaks.
Ecosystems help meet peoples’ need for water by regulating the
water cycle, filtering impurities from water, and regulating the
erosion of soil into water. Population growth and economic
development have led to rapid water resource development,
however, and many naturally occurring and functioning systems
have been replaced with highly modified and human-engineered
systems. Needs for irrigation, domestic water, power, and
transport are met at the expense of rivers, lakes, and wetlands
that offer recreation, scenic values, and the maintenance of
fisheries, biodiversity, and long-term water cycling.
Conservation projects often only consider a few benefits of nature’s
preservation. An ecosystem services framework can help
build support for these projects by clarifying that their success
provides multiple ecosystem services and therefore is linked to
the achievement of other development goals. If a protected area,
for example, can be shown to have additional benefits such as
providing biochemicals for pharmaceuticals, its creation is more
likely to be supported.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 5
Table 1.1. continued
Dependence on ecosystem services
Food production
Ecosystems are vital to food production, yet there is pressure to
increase agricultural outputs in the short term at the expense
of ecosystems’ long-term capacity for food production.
Intensive use of ecosystems to satisfy needs for food can
erode ecosystems through soil degradation, water depletion,
contamination, collapse of fisheries, or biodiversity loss.
The majority of the world’s 1 billion poorest people live in rural
areas. They depend directly on nature for their livelihoods and
well-being: food production, freshwater availability, and hazard
protection from storms, among other services. Degradation of
these services can mean starvation and death. Investments in
ecosystem service maintenance and restoration can enhance
rural livelihoods and be a stepping stone out of poverty.
Many renewable energy sources, such as biofuels or hydroelectric
power, are derived from ecosystems and depend on nature’s
ability to maintain them. Hydropower, for example, relies on
regular water flow as well as erosion control, both of which
depend on intact ecosystems.
Poverty reduction
Energy security
Source: WRI 2008.
assessment can help build a bridge between the development and environmental
communities by providing credible and robust information on the links between
ecosystem management and the attainment of economic and social goals. This can
mean the difference between a successful strategy and one that fails because of an
unexamined consequence, for example for a freshwater supply, an agricultural product, a sacred site, or another ecosystem service (see Box 1.1).
Undertaking an ecosystem services assessment and taking the findings into account in policies and action can improve the long-term outcome of decisions. As
improvements are made in describing and valuing the benefits of ecosystem services,
decision makers can better understand how their actions might change these services, consider the trade-offs among options, and choose policies that sustain the
appropriate mix of services. A range of assessment initiatives in recent years have
focused on various aspects of ecosystem services. Box 1.2 provides an overview of
the main recent and ongoing global assessment initiatives; further resources and
background information on ecosystem services can be found in the “Additional Resources” section at the end of this chapter.
An assessment of ecosystem services needs to consider both the ecosystems
from which the services are derived and also the people who depend on and are affected by changes in the supply of services, thereby connecting environmental and
development sectors. Assessments play numerous roles in the decision-making process, including responding to decision makers’ needs for information, highlighting
trade-offs between decision options, and analyzing ecosystems to avoid unforeseen
long-term consequences. They inform decisions through providing critical judgment
6 | A Manual for Assessment Practitioners
Box 1.1. The trade-off between food and fuel
Global food prices have been on the rise since 2000; they rose nearly 50 percent in 2007 alone.
The price of basic staples, such as corn, oilseed, wheat, and cassava, is predicted to increase
26–135 percent by 2020. The recent increase in the cost of grain-based staples, such as tortillas in Mexico, beef noodles in western China, and bread in the United States, has several
causes, including the emerging consequence of the increase in bioenergy production.
Promoted as a clean, sustainable alternative to fossil fuels, industrial countries have set
increasingly higher mandates for the use of bioenergy to combat global climate change. Efforts
to meet the rapid increase in demand for bioenergy have led to a global competition for limited
natural resources such as land and water. Experts predicted that 30 million extra tons of corn—
half of the global grain stock—would be dedicated to ethanol production in 2008 in the United
States. On average, the grain required to make enough ethanol to fill a large car is enough to
feed a person for a whole year. Crops can be used as food or fuel; both are important ecosystem services provided by nature. As countries continue to target corn and other agricultural
products as the future supply of fuel, however, less food becomes available and food prices
increase worldwide.
While the potential benefits of bioenergy can range from lower greenhouse gas emissions
(in some cases) to renewability and energy independence, there are often trade-offs across
other ecosystem services as a result of increased biofuel production. In addition to decreased
food supply, the possible trade-offs include water quality impacts associated with increases in
aggregate fertilizer use, nutrient runoff and erosion, and in some cases an increase in greenhouse gas emissions. Although the economic, social, and environmental effects of the recent
biofuel push are not yet fully understood, many countries are rapidly expanding the area dedicated to these crops. It is in situations such as this that policy makers can benefit enormously
from thorough assessments to determine options and better understand the consequences of
their decisions.
of options and uncertainty and through synthesizing and communicating complex
information on relevant issues. They are also of value through the process they
involve, which engages and informs decision makers long before final assessment
products are available.
Successful assessments share three basic features:
• First, they are credible. Involving eminent and numerous scientists as authors
and expert reviewers and ensuring that all reports undergo expert peer review
will help to ensure credibility. Assessments should focus not only on what is
known with certainty by the scientific community but also on what remains
uncertain. The clarity that assessments have given to areas of real scientific
uncertainty (such as climate change in the 1990s) has been just as important in
guiding policy as the clarity they have provided where there is broad scientific
agreement. Moreover, by identifying areas of scientific uncertainty that
matter for policy decisions (e.g., the ability to predict thresholds of change in
socioecological systems), assessments can also help stimulate more support for
scientific research.
Box 1.2. Recent and ongoing international assessments that focus on ecosystem
Millennium Ecosystem Assessment
The Millennium Ecosystem Assessment, released in 2005, assessed the consequences of ecosystem change for human well-being. The MA consisted of a global assessment and 34 subglobal assessments to assess current knowledge on the consequences of ecosystem change
for people. The MA brought about a new approach to assessment of ecosystems: a consensus
of a large body of social and natural scientists, the focus on ecosystem services and their link to
human well-being and development, and identification of emergent findings. The MA findings
highlight the strain that human actions are placing on the rapidly depleting ecosystem services
but also that appropriate action through policy and practice is possible. (
International Assessment of Agricultural Science and Technology for Development
The International Assessment of Agriculture Science and Technology for Development
(IAASTD), released in 2008, was an intergovernmental process that evaluated the relevance,
quality, and effectiveness of agricultural knowledge, science, and technology (AKST) and the
effectiveness of public- and private-sector policies as well as institutional arrangements in relation to AKST. The IAASTD consisted of a global assessment and five subglobal assessments
using the same assessment framework, focusing on how hunger and poverty can be reduced
while improving rural livelihoods and facilitating equitable, environmental, social, and economical sustainable development through different generations and increasing access to and use of
agricultural knowledge, science, and technology. (
Intergovernmental Panel on Climate Change
The Intergovernmental Panel on Climate Change (IPCC) released its fourth report (AR4) in 2007.
The IPCC was established to provide decision makers with an objective source about climate
change. Similar to the MA, the IPCC does not conduct any research or monitor specific data
and parameters; it assesses the latest scientific, technical, and socioeconomic literature in an
objective, open, and transparent manner. Ecosystem services are addressed in the fourth report of the IPCC by the reports of Working Group II (Impacts, Adaptation and Vulnerability) and
Working Group III (Mitigation of Climate Change). The findings of AR4 highlighted a number
of overarching key issues in relation to ecosystems and the services they provide for climate
mitigation and adaptation. Specifically, the report drew links between the loss of ecosystem
services and the reduction of societal option for adaptation responses. (
Land Degradation Assessment of Drylands
The Land Degradation Assessment of Drylands (LADA) is an ongoing assessment that aims to
assess causes, status, and impact of land degradation in drylands in order to improve decision
making for sustainable development at local, national, subregional, and global levels. Currently
the LADA is focusing on developing tools and identifying available data that will be required to
discover status and trends, hotspots of degradation, and bright spots (where degradation has
been slowed or reversed). (
Global Environment Outlook
The Global Environment Outlook (GEO) is the United Nations Environment Programme’s ongoing assessment of the environment globally. The fourth GEO was released in 2007 and consists
of a global assessment and subglobal assessments. GEO-4 provides information for decision
makers on environment, development, and human well-being. (
8 | A Manual for Assessment Practitioners
• Second, they are legitimate. It is relatively easy for the administration of one
province or country to ignore an assessment and report done by experts in another province or country, or for the CEO of a private company to ignore the
findings of a report by a nongovernmental organization (NGO). What possible
leverage would such a report have, no matter how thorough the science in it?
Thousands of assessments and studies are published every year; what gives an
assessment more weight with decision makers than others? Partly it is the authoritative status and credibility of the assessment through the organizations
and individuals involved. But, equally important, the involvement of users of
the assessment in the process itself ensures greater impact with decision makers
through instilling a sense of “ownership” of the findings. A successful assessment is one that is legitimate in the eyes of the users, where decision makers use
it as their own product.
• Third, they are relevant (or salient) to decision makers’ needs. This is not to say
that scientists do not have an opportunity to introduce new issues and findings
that decision makers need to be aware of. They certainly do. But the priority for
the assessment is to inform decisions that are being faced or soon will be faced
by decision makers, at a particular scale, and in a particular context. (Section
1.4.2 provides further details on ensuring an assessment is policy relevant.)
Early on, assessments should evaluate whether they meet these three criteria and,
if not, take the steps necessary to incorporate them. Chapter 2 explores in greater
detail the various approaches to implementing these criteria and elaborates on the
importance of stakeholder involvement at all stages in the assessment process.
1.3 How to conduct an ecosystem assessment—an overview
Section’s take-home messages
• The assessment process has three key stages that are generally sequential but usually
overlapping and iterative: the exploratory stage, the design stage, and the implementation
of the assessment workplan.
• In the design stage, assessment organizers need to consider governance of the process,
conceptual frameworks, how to link the different scales to be addressed, how to bridge
different knowledge systems, capacity-building needs, and how to evaluate the process.
• The implementation stage requires the greatest need for flexibility. This stage generally
assesses conditions and trends in ecosystems and their services, scenarios for the future,
and past and current responses taken to enhance the contribution of ecosystems to
human well-being.
An assessment will have the greatest impact where consideration is given to both
process and products, where stakeholders are fully engaged, and where assessment
design follows scoping of user needs. The assessment process has three key stages
that are generally sequential but usually overlapping and iterative: the exploratory
stage, the design stage, and the implementation of the assessment workplan (see
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 9
Figure 1.2. An overview of the assessment process.
Figure 1.2). User engagement, communication, and capacity building occur throughout the entire assessment process. A review process is essential for the assessment
and provides both credibility and an opportunity for further engagement of users.
All assessments are likely to need to be flexible and adaptive—to changing circumstances, user requirements, and process (and funding) constraints.
1.3.1 Exploratory stage
Determining the need for an assessment
The concept of an authorizing environment is a useful way to ensure that an assessment has the necessary level of buy in from key stakeholders. The authorizing
environment is the set of institutions and individuals who see an assessment as being
10 | A Manual for Assessment Practitioners
undertaken on their behalf and with their endorsement and engagement. Examples
might be village elders, land managers, agricultural cooperatives, or local or national
governments. In practice, whether or not the members of the authorizing environment have provided formal authorization, the true test of whether the authorizing
environment was sufficient is whether those stakeholders have a substantial ownership in the final products and a commitment to take actions based on the findings.
In some instances it may be appropriate to stimulate or encourage demand for
an assessment. For example, in a situation where there is a lack of a consensus on
aspects of the connection between ecosystems and people, an assessment might be
proposed and communicated as being a useful tool for local decision makers to resolve particular issues, and thereby stimulate the demand for the assessment process
and its outputs. In all cases, however, the decision to proceed with the assessment
should be taken on the basis of actual rather than perceived demand, demonstrated
through the recognition of, and approval as appropriate by, an authorizing environment. See Chapter 2 for further information on the importance and approaches for
engaging users in the exploratory phase of the assessment.
Defining scope and boundaries
An assessment can be defined by its intended audience. If the primary users are national decision makers, then it will be a national assessment even though it might
be examining ecological and economic processes from local to global scales (such as
international trade in natural resources or climate change). If the primary users are
international conventions, then the assessment is global. And if the primary users are
a particular local community, then it is local.
The audience for an assessment is not the only factor involved in defining an
assessment’s scope. The scope ultimately depends on political, socioeconomic, and
environmental circumstances that might constrain its boundaries. Even if the primary audience is a particular local community, ecological or social factors in the
region might suggest that the scale should be larger than just one community. For
example, if an issue of major concern to a community is water, then a river basin
scale may ultimately be more appropriate since both ecological and social processes
throughout that basin will strongly affect water availability in that community. The
level of involvement of particular expertise or disciplines might constrain the themes
addressed by an assessment; the level of involvement of users (or conflict between
them) might also constrain the questions being asked. Constraints may arise as well
from funding limitations, data shortages, or methodological constraints. In some
cases these constraints can be overcome, such as through active recruitment of additional expertise. But in many cases the constraints will need to be acknowledged
and considered during the design and process of the assessment.
1.3.2 Design stage
Once the needs and constraints have been identified, then the governance, content,
and process for implementing the assessment can be determined. A thorough design
phase, including consideration of funding and the ongoing engagement of users, is a
key step in eventual success.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 11
The governance (including leadership) of an assessment can be a critical factor in
ensuring user engagement, raising funds, and overseeing progress in implementation
of the assessment. It is also crucial to ensure legitimacy and credibility. A model that
has been found to be effective in the MA and other assessments is having the assessment overseen by a technical Steering Committee or Assessment Panel and an associated “User” Advisory Committee or Assessment Board. In some cases, involving
both the technical experts and users in a single committee might work well; in other
instances, it may be more appropriate to establish a separate Advisory Group to
represent the various users. The governance structure for the Millennium Ecosystem
Assessment is provided in Figure 1.3.
Typical functions of a Steering Committee would be to:
• Promote coordination among the institutions and individuals carrying out the
• Develop the detailed assessment design (what information will be produced by
which individuals and institutions);
• Increase the legitimacy of the assessment, and guard against bias from particular
interest groups;
• Assure quality of assessment outputs;
• Design the outreach and communication activities; and
• Help to raise funds for the assessment.
Figure 1.3. Governance structure of the Millennium Ecosystem Assessment.
12 | A Manual for Assessment Practitioners
The size and composition of the Steering Committee will of course vary, depending
on the scope of an assessment. In the case of community efforts, it might consist of
village leaders and researchers who will be involved in the assessment. In regional
assessment activities, it is likely to include representatives of a number of different
scientific networks and institutions within the region. Depending on the region or
assessment, the Steering Committee could consist primarily of technical experts (in
which case a separate advisory committee or Board should be established to ensure
oversight of the process by users) or it could involve both a mix of technical experts
and users. In either case, a small “executive” committee is likely to provide the most
effective day-to-day oversight of activities.
If an advisory group is established, an important consideration is the extent
to which that group should be strictly advisory or should have decision-making
authority—for example, over the distribution of funding against priorities within
the assessment. It is likely to be far more effective if the users hold decision-making
authority either as part of the core steering group or as part of a formal board rather
than serving in only an advisory capacity. If the intended audiences do not have
some level of authority over the assessment, it is unlikely they will be sufficiently
engaged in the process to fully use its findings. Moreover, without a formal requirement to respond to the needs of the audience, there remains a significant risk in any
scientific assessment that the focus will begin to reflect the research interests of the
scientists more than the needs of the decision makers.
Effective governance is also essential to ensure that assessment practitioners
abide by the process of which they are part, including on agreements to share information and data used and resulting from the assessment (which helps to ensure
transparency), on peer review of materials, and on engaging sufficient author expertise to ensure credibility and legitimacy.
Conceptual frameworks for assessment
Central to the coherence of an assessment is the design or adoption and use of a
conceptual framework—a common understanding of what the assessment aims to
do. A single agreed conceptual framework guides the assessment, allowing multiple
practitioners to work within the same boundaries and understanding of what is being assessed, and therefore allows integration of the components of the assessment.
Conceptual frameworks help clarify the complex relationships between elements of
the human–ecological system, including how those relationships may be changing
over time. Chapter 3 presents further information on the design, adoption, and use
of conceptual frameworks.
Linking scales in ecosystem assessments
In many cases, assessments seek to identify important relationships between processes and phenomena at a particular scale. However, processes operating at a local
scale are affected by processes at larger scales, from human migration to air pollution. In turn, processes operating at larger scales reflect cumulative or systemic effects of what is going on in a number of localities, such as land use and agricultural
production. Assessing the condition and trends in ecosystems, their services, and
human well-being requires an understanding of these cross-scale relationships. In
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 13
many cases it will be possible to place an assessment in the context of others that
have already been conducted at larger scales (and in all cases in a global context
from the MA, the IPCC, or other global assessments). Important cross-scale linkages
that might be considered as part of an assessment include:
• Linkages between environmental systems that operate at different scales;
• Linkages between institutional systems that play roles in ecosystem management
and use; and
• Linkages with larger driving forces, including globalization, technological
change, and institutional change.
The scale of the assessment will also influence the analytical units used for assessment and communication. In many cases there will be a geographic mismatch between the most appropriate units for assessment of ecosystems and their services
(which depend on biophysical data and boundaries) and those for assessment of
human well-being (which depend on sociopolitical and economic data and boundaries). See Chapter 4 for a further discussion on scale and units of analysis.
It is essential that an assessment (and thus the methods and data it uses) is at a
scale that matches that of the biological and physical processes generating the ecosystem services being assessed and also at a scale relevant for the decision-making
processes of the target audience. Sometimes these two scale demands can be reconciled by choosing one particular scale that satisfies both. Often this will not be possible, however, either because the scale gap between them is too large (which in itself
is a warning signal that problems of management could arise) or because important
processes are simultaneously occurring at several scales and interacting across them.
In this case there is little choice but to do a multiscale assessment. In the case of
cross-scale interactions, which tend to be common (for instance, national policy
may determine a local outcome), it is far better to carry out an integrated multiscale
analysis than simply to carve the assessment into a number of discrete scales that are
independently assessed without trying to understand the linkages.
Integrated multiscale assessments are inherently more complicated and often
more expensive than assessments at single scales. However, there are a number of
ways to simplify a multiscale assessment. First, the assessment does not need to
cover every possible scale. Usually there is enough flexibility in the scales of underlying processes that selecting about three key scales is sufficient: for instance, local,
national and regional. Second, the sampling at a finer scale does not have to represent complete coverage at the coarser scale—it can simply cover what is determined
to be a representative sample (this has been called a “sparsely nested hierarchy”).
For example, the Southern African MA (SAfMA) assessed at the local scale in five
areas, at the water basin level in the Gariep and Zambezi river basins, and at the
regional level in Africa south of the equator. Third, limiting the number of issues
to be assessed at multiple levels to a core set of perhaps three to five helps to keep
the process manageable. Three ecosystem services were assessed by SAfMA, for instance, across all three scales: food, fresh water, and biodiversity. Finally, it is helpful
if the indicators chosen at the various scales are either identical or bear some clear
relationship to one another. It then becomes possible to trace how a particular issue
is expressed at the various scales. More details about aggregating and disaggregating
across scales are given in Chapter 4.
14 | A Manual for Assessment Practitioners
Bridging knowledge systems—linking formal and informal knowledge
Scientific assessments are based on a particular mainstream epistemology (way of
knowing)—one that often pays little attention to informal (local, traditional, or
indigenous—see Box 1.3) knowledge and that takes little in the way of cultural
values into account. Scientists and policy makers alike have become aware that new
assessment processes need to be robust enough to accommodate and value these
different knowledge systems and the multiscale and multistakeholder nature of environmental concerns (Reid et al. 2006). A significant challenge for a multistakeholder
assessment is to effectively bridge these traditional and scientific ways of knowing
the world. A rich body of knowledge concerning the history of ecosystem change
and appropriate responses exists within local and traditional knowledge systems. It
makes little sense to exclude such knowledge just because it is not published. Moreover, incorporation of traditional and local knowledge can greatly strengthen the
legitimacy of an assessment process in the eyes of many local communities. Finally,
particularly at the scale of a local assessment, traditional knowledge may often be
the primary source of historical information for the assessment.
It is helpful to think of information falling into one of four quadrants defined
by the axes of formal–informal and tacit–explicit, because it makes clear what steps
need to be taken to make such information usable in an assessment. Assessments fit
into the formal–explicit quadrant (see Table 1.2). Information must either be made
formal (by documenting it) or explicit (by placing it in the public domain), or both
if it is to be used in an assessment. These techniques do not only apply to local and
Box 1.3. The semantics of different forms of knowledge
Informal (local, traditional, or indigenous) knowledge can add significantly to a scientific (formal
knowledge) assessment, but the semantics that surrounds these different forms of knowledge
can lead to confusion among assessment practitioners and users. The following working definitions are provided to help incorporate different knowledge systems into assessments and
to bring some clarity and standardization to the terms used in the scientific and assessment
• Local knowledge refers to place-based experiential knowledge systems. It may include
traditional and indigenous knowledge; it is largely oral or practice-based and so is rarely
• Traditional knowledge is the body of information, practices, and beliefs that evolves
through adaptive processes and is handed down through generations through traditional
transmission practices. If such knowledge relates to local ecology, then this is traditional
ecological knowledge. Traditional knowledge is not necessarily indigenous, but it certainly
has its roots in the past.
• Indigenous knowledge is the local knowledge held by indigenous peoples or the local
knowledge unique to a particular culture or society. The term is usually only applied when
referring to knowledge held by people who identify themselves as indigenous.
• The term local and traditional knowledge is the most encompassing of the informal knowledge systems. Although it may also incorporate elements of scientific or formal knowledge, it is often the most appropriate way of referring to a worldview that is different from
mainstream science or government decision makers.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 15
Private images or photos
Unpublished models and databases
Table 1.2. A categorization of types of information.
Private beliefs and values
Ecosystem assessments
Peer-reviewed papers, chapters, or
books in the scientific literature
Peer-reviewed databases
Oral traditional knowledge
Indigenous knowledge, rules, and
Communal beliefs and values
Untested scientific databases
Tacit information is known only by individuals, whereas explicit information is shared, with some level
of agreement.
Source: Fabricious et al. 2006.
traditional knowledge. For instance, much information is contained in the private
experience of scientific experts and can be formalized and made explicit by questionnaires or interviews.
Capacity building
In the context of an assessment, capacity building is a continuous process aimed at
strengthening or developing long-term relevant human resources, institutions, and
organizational structures to carry out ecosystem assessments of relevance to decision
makers and to act on the findings. Capacity building within an assessment has two
objectives: to enhance the expertise of individual scientists to carry out ecosystem
assessments and to enhance institutional expertise, particularly the science–policy
interface, for effective adoption and use of the assessment findings.
Assessments may also provide important capacity-building opportunities
through improving research capacities of universities and other research and training institutions, establishing baseline data for further assessments in the future, fostering an appreciation for scientific knowledge on the part of decision makers, and
establishing or strengthening regional networks of experts.
Tangible ways to incorporate capacity building into the assessment implementation include developing a “fellows program” for young scientists to partner with
more senior or experienced scientists engaged in the assessment, providing training
courses on scenarios (or other assessment) methodologies, developing training materials, and forming partnerships with other institutions to expand the reach of these
activities and to address decision makers’ needs to build their own capacity to use
the assessment’s findings.
In order to learn constructively from assessments and to communicate those lessons
to future assessment activities, a formal evaluation should be undertaken. Given the
16 | A Manual for Assessment Practitioners
learning benefits both during the assessment period and afterward, two components
of an evaluation may be considered: interim evaluations to allow for midcourse corrections during the assessment process and “post-hoc” evaluations after the initial
phase to learn what worked, what did not, and where improvements could be made.
Regardless of the specific mechanisms used, any rigorous and meaningful evaluation
requires clear and early articulation of assessment goals, objectives, and Terms of
Reference to provide a primary benchmark for evaluation and measurable indicators
of success.
1.3.3 Implementing the assessment
Once an assessment has been designed around the requirements of identified users
and the work plan has been developed to deliver the assessment, engage users, and
communicate the process and findings, then the technical work of the assessment can
begin. Of all the stages of the process, it is implementation that typically entails the
greatest need for flexibility. In some cases the various components of technical work
will need to be consecutive and sequential; in other cases, iterative and interactive.
In all cases this will be determined by the constraints imposed on the assessment
from finance or capacity limitations and by the options and opportunities available
in terms of timing and resources.
The MA contained three basic components: assessment of condition and trends,
scenarios, and responses. Although each of these can be undertaken by a separate
working group of scientific experts, at many of the sub-global scales of the MA all
three components were addressed by the same group. In many cases it would be
better to undertake these three components sequentially, so as to first assess the
condition and trends of ecosystems, their services, and human well-being; second,
develop scenarios of change; and third, assess available and potential responses. In
practice, however, these elements also need to interact closely, and what is learned
from doing any one of these components can inform other technical parts of the
assessment. The most important issue is that these components remain connected—
through use of a joint conceptual framework, terminology, and approach. See also
Chapter 5 for a discussion on the use of scenarios in the various stages of the assessment process, including as a tool to help determine the questions to be addressed by
an assessment.
The condition and trends component should assess the priority ecosystem services and the associated drivers of change and impacts on human well-being that
were selected during the design phase, based on the requirements and priorities determined from the user community and authorized by the assessment governance.
Depending on data availability, it is preferable to consider the condition and geographical distribution and trends of the supply and demand for each service, as well
as the effect of changes in ecosystems on their capacity to supply these services and
the consequences of historical changes in these services for human well-being. A key
element of the assessment of services will also be to consider the trade-offs that have
been made between the supply of the various ecosystem services being assessed and,
again, the consequences in these trade-offs for people.
The condition and trends assessment will by necessity focus on historical changes
and impacts of ecosystem change on human well-being, and it should aim to use a
sufficient variety of data, information, and indicators to be as comprehensive as
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 17
possible regarding the selected ecosystem services. However, it is also feasible for
this component to consider the foreseeable future and to thereby make a direct link
into the scenarios assessment. See Chapter 4 for a full discussion on how to assess
conditions and trends.
The scenarios component should aim to develop a set of scenarios providing
descriptive storylines, supported by quantitative approaches, to illustrate the consequences of various plausible changes in drivers, ecosystems, their services, and
human well-being. Scenarios are not attempts to forecast the future but rather are
designed to provide decision makers with better understanding of the potential consequences of decisions. They also help to understand the uncertainty about the future in a creative way and can help explore new possibilities to respond to change.
See Chapter 5 for details on how to use scenarios in an ecosystem assessment.
The responses component should aim to examine past and current actions taken
to enhance the contribution of ecosystems to human well-being. In doing so, it
should provide practical observations, tools, and guidelines for the various users on
the effectiveness of actual and potential interventions, as well as promising options
for action by a variety of stakeholders. Such interventions might include policies,
practices, financial mechanisms, or communications and awareness-raising activities. See Chapter 6 for a full discussion on assessing the effectiveness of responses.
The peer review process is essential to ensure validation of the findings and to
provide credibility to the process. Involving the user community in peer review of
the findings also enables early feedback from the users on the utility of the assessment’s outputs and is a key part of the communications strategy. The time involved
to conduct a comprehensive review process—in terms of providing sufficient time
both for reviewers to provide comments and for the assessment team(s) to incorporate (or justify the exclusion of) comments can be considerable. However, given the
importance of credibility, in all cases the value of peer review will outweigh the time
and costs of the process. In addition, it can be extremely helpful for increasing the
transparency and objectivity of the assessment process if responses to review comments are made publicly available.
Assessments will need to develop review processes that are tailored to their own
circumstances and to the scale and the context in which it is undertaken. In general,
however, review processes should aim to meet the following criteria:
• The review process should be independent. An independent party not involved
in the governance or operations of the assessment must have the authority
to determine whether reviewer inputs have been sufficient and whether the
comments have been handled adequately.
• Relevant governments (for the scale at which the assessment is conducted),
NGOs, regional institutions, and other organizations as appropriate should be
contacted in advance to identify appropriate reviewers, and reviews should be
requested from all these sectors.
• Reviewers should be requested with the aim of obtaining a balanced representation of views within the scope of the assessment and among scientific, technical,
and socioeconomic perspectives.
• Reviewers should include experts involved in the larger- and smaller-scale assessments within which the assessment is nested or that are contained within the
18 | A Manual for Assessment Practitioners
• All written review comments, and the responses to those comments, should be
archived and made publicly available.
1.3.4 Communicating ecosystem assessments
Assessments can succeed or fail depending on the communications strategy, and one
of the most important decisions made by the assessment team or advisory body will
be how to distribute limited resources between technical work and the communications component. Both the process and the outputs of the assessment are critical to
communications, and the impact of the assessment will depend as much on communicating the legitimate and credible process as it will on communicating the policyrelevant findings.
The primary purpose of an assessment is to meet the needs of decision makers
at the scale at which the assessment is conducted. In doing so assessments will almost always produce products, including written reports, audiovisual materials, and
other products tailored to the needs of the decision makers at that scale. The specific
nature of these products will depend on the circumstances and scale involved and on
the specific needs of the users. Thus, for example, a regional (multicountry) assessment would likely produce a technical report volume and a summary for decision
makers aimed at the needs of any regional governing bodies and the national policy
makers of countries within the region. At the same time, it could also produce products that could empower local communities or contribute to educational activities.
On the other hand, a local community assessment may produce only a single report,
since the users of the assessment might largely be the same as the people producing
the assessment. Documents published as components of an assessment should adhere to a set of criteria for the preparation, peer review, and approval of the documents approved by the assessment governing body.
1.4 How to link an assessment to the decision-making process
Section’s take-home messages
• Entry points for incorporating the results of an ecosystem services assessment into
decision-making processes occur at all levels of governance and are important for
both development officials and those approaching problems from an environmental
• Opportunities for mainstreaming ecosystem services can be categorized into four
intersecting entry points: national and subnational policies, economic and fiscal incentives,
sector policies, and governance.
• The information that will be most effective in eliciting constructive responses—and hence
the most appropriate focus of an assessment process—depends on the place of a
particular issue in the policy life cycle of environmental issues.
• Four key elements ensure that an assessment is policy relevant: identifying the questions
for policy or decision making that the assessment should try to answer; answering those
questions; stating the levels of certainty (or uncertainty) associated with the findings; and
presenting the answers in a separate Summary document, an Executive Summary, or a
“Main Messages” section.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 19
1.4.1 Understanding the decision-making context
Decision makers as diverse as mayors, national economists, natural resource managers, and conservation planners can use an ecosystem services assessment to explore
the links between ecosystems and economic development, gaining a better understanding of how their goals both affect and depend on ecosystems and their services.
Those working in the social and economic development community often base their
analysis on elements of human well-being such as health or food; through the participation in, or outputs of, the assessment, they can connect those to ecosystem services. The environmental conservation community, on the other hand, can start from
an environmental perspective and assess the implications of conservation actions on
development and human well-being.
Entry points for incorporating the results of an ecosystem services assessment
into decision-making processes occur at all levels of governance and are important
for both development officials and those approaching problems from an environmental perspective. Many entry points are found at the national or provincial level.
Some, such as the Millennium Development Goals or international trade and investment, are at the global level, but these usually have more detailed counterparts at
the national or local level. Opportunities for mainstreaming ecosystem services can
be helpfully categorized into four intersecting entry points: national and subnational
policies, economic and fiscal incentives, sector policies, and governance (see also
Table 1.3).
• National and subnational policies: The preparation of national and subnational
trade, economic growth, or immigration policies provides important entry
points for managing the cumulative demand and impacts on ecosystem services
from individual or multiple sectors. Ministries of the environment, treasury,
development, and planning, among others, may play a role.
• Economic and fiscal incentives: Fiscal measures such as subsidies, taxes, and
pricing influence decisions throughout the economy, from firms and farms to
factories and households. They can be designed to create incentives to sustain
and efficiently use ecosystem services, as well as to create disincentives for activities that drive ecosystem degradation.
• Sector policies: Ministries of commerce and industry, science and technology,
and agriculture and forestry, among others, can play an effective role in advancing policies and actions that sustain ecosystem services. Environment agencies
can work with other government agencies and departments to develop information, tools, and analyses that help make the connection between ecosystem services and the attainment of sector goals.
• Governance: Strong governance is at the heart of sustaining ecosystem
services. This includes public participation in decisions that affect or depend
on ecosystem services, a free press, and requirements to provide information
to the public, including regular indicators of ecosystem health. All branches of
government also have a role in providing oversight. Such mechanisms enable
citizens to hold governments and businesses accountable for their use and
management of ecosystems.
20 | A Manual for Assessment Practitioners
Table 1.3. Entry points for mainstreaming ecosystem services
Entry points
Ministry, agency,
or organization
National and
policies and
Development &
Physical planning,
planning, and
Economic and
fiscal incentives
Sector policies
and plans
Commerce and
Science and
natural resources
Prime minister’s or
president’s office,
justice ministries,
legislature, local
Examples of decision processes
Poverty reduction strategies, land use planning, water supply, and sanitation
Protected area creation, climate adaptation
National budgets, public expenditure reviews,
Integrated ecosystem management of coasts,
river basins, forest landscapes, and watersheds
Subsidies, tax credits, payments for ecosystem services, import duties, and tariffs
Tax policies to support easements or promote
alternative energy technology, pricing regulations for water
Corporate codes of conduct/standards,
assessment of new technologies
Applied research, technology transfer, business
capacity building
Extension services, best management practices
Forest sector action programs, mapping initiatives, concession management
State of the environment reports, strategic
environment assessments, environmental
impact assessments, information/tools,
legal instruments
Decentralization policies, free press, civil society,
accountability of government through elections, access to information and decisions,
judicial review, performance indicators
The examples provided for each entry point are not intended to be exhaustive but rather to illustrate
the variety of ways that ecosystem service considerations can be incorporated into decision
Source: WRI 2008.
Decision making is complex and multidimensional and occurs at all scales, from
individual to international. Decision makers can be individuals or groups of people,
in private or public sectors. Assessments take place in the context of this political
economy. Thus, asking the right questions to inform decision making requires close
interaction with the decision makers who are the target audience of the particular assessment. It also requires sufficient knowledge of the decision-making process to be
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 21
able to tailor the assessment, and particularly the communications from the assessment, appropriately. The effectiveness of an assessment in terms of influencing policy
and actions can also be constrained by how open and willing the political arena is to
the concept of bringing science to bear on decisions. Better science can certainly lead
to better decision making, but it does not guarantee it.
Decision makers have at least three strategies to choose among in responding to
information presented in an ecosystem assessment:
• Do nothing.
• Compensate people who are affected by ecosystem change.
• Prevent (or reduce) ecosystem degradation through policy and regulations,
market-based instruments (taxes and/or subsidies), other means of social
control, or some combination of these approaches.
Even with overwhelming evidence of the impacts of ecosystem change on people,
considerable pressure may be required before decision makers take action. The information that will be most effective in eliciting constructive responses—and hence
the most appropriate focus of an assessment process—depends on the place of a
particular issue in the policy life cycle of environmental issues (see Figure 1.4):
• Stage 1: Identification of a particular environmental or development issue by a
small group of “pioneers” but no broader awareness either by society at large or
by decision makers.
• Stage 2: Lobbying by “action groups,” often a denial of effects by some groups
of stakeholders, and incipient awareness but no action by decision makers.
Figure 1.4. The environmental
issues “life cycle,” showing how
interest and public perception
of environmental issues changes
through time.
Source: Tomich et al. 2004
22 | A Manual for Assessment Practitioners
• Stage 3: Widening acceptance of the potential or actual environmental or development issue, with mounting awareness and pressure from civil society for decision makers to take action.
• Stage 4: Debate on evidence for understanding and causes of the issue, which
has now become a policy issue.
• Stage 5: Inventory and assessment of prevention and mitigation options and
their environmental, socioeconomic, and administrative costs and benefits.
• Stage 6: Negotiations on prevention or mitigation of impacts.
• Stage 7: Implementation, monitoring, and enforcement of prevention or
mitigation actions.
Although the details will differ between various environmental and development issues, the course of events in local, national, and regional issues could follow similar
paths, at least if actions somehow respond to both the numbers of people concerned
and the intensity of their concern. For an issue in Stages 2 and 3 of this cycle, an
assessment would be most effective if it focused on testing the validity of Stage 1
“suspicions” about a link between an undesirable ecosystem change and an impact
of human well-being. Establishing a probable cause-and-effect chain as opposed to
“mere coincidence” or “spurious correlation” is important at this point as a basis
for sound policy intervention. This information also could help build broader support for action and undermine resistance from vested interests. In these stages there
also is a need to estimate the likely magnitude of impacts (are the effects big or
small?) since initial uncertainty may range over several orders of magnitude.
Once awareness and support is formed for action on a particular issue, the debate may shift focus to understanding the causality of the problem. Perceived gaps
in the quantification of impacts or in a causal explanation of the phenomenon are
major obstacles in Stage 4. Various stakeholders may agree on the need for an inventory of prevention and mitigation options (Stage 5), or the process may be more
adversarial, with each group applying evidence selectively and advocating a position
serving its own interests. An assessment focusing on the effectiveness of response
options and scenarios will therefore be particularly relevant and useful for issues in
Stage 5 and 6.
Beyond stage 6, the information needs tend toward those of monitoring rather
than assessment, although it is likely that a future assessment will allow for more informed adaptive management and a review of successful interventions in the future,
and some problems that have apparently been understood and solved, or at least
brought under control, may re-emerge in a new cycle if situations change or if the
initial understanding proves to be incorrect or responses ineffective.
1.4.2 The importance of being policy relevant: moving beyond
scientific reviews
A scientific assessment is fundamentally different from a scientific research project or
review. An assessment leads to a different product and has a different audience and
communications strategy. It is not a research initiative and does not generally seek
to undertake new research, but rather applies the judgment of experts to existing
knowledge generated from the scientific community (and other forms of knowledge)
to provide credible answers to policy-relevant questions.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 23
The audience of an assessment is decision makers, typically in government (national and local), business, or civil society, and particularly those operating at the
scale of, and within the scope of, issues addressed in the assessment—although there
may also be other key audiences. For example, an assessment of water resources and
associated ecosystem services in a particular river basin will be of direct interest to
water and land mangers in the river basin, but it also might generate useful insight
for managers in adjacent and other basins and for decision makers in other sectors,
such as health professionals interested in water-related health issues.
Four key elements enhance the policy relevance of an assessment: identifying the
questions for policy or decision making that the assessment should try to answer;
answering those questions; stating the levels of certainty (or uncertainty) associated
with the findings; and presenting the answers in a separate Summary document, an
Executive Summary, or a “Main Messages” section.
First, it is important to identify the relevant questions for policy or decision
making that the assessment should try to answer. A “policy-relevant” question is one
that is asked by a user group, audience, or decision maker or one where the answer
to the question can be used to justify or support a decision or action that directly or
indirectly affects the allocation of public or private resources (money, time, land use
rights, etc.). A question is also policy relevant if it informs conceptual understanding
of how the world works or otherwise informs decision makers about the context for
their decision making and is linked to the decision-making process. The most effective way to identify policy-relevant questions is to survey users directly to determine
their “user needs.” Given the scope of a particular assessment, it might also be useful to identify other policy-relevant questions that the assessment could answer even
if they have not yet been formally posed by users (i.e., other policy-relevant issues
may emerge from the assessment that are not yet on the agenda of the assessment
users but might be useful to consider or be aware of). Chapter 2 provides further
guidance on ensuring that an assessment is responding to user needs.
Second, to be policy relevant, the assessment will need to answer the questions
identified. Certainly it will need to be designed around answering the user needs.
But also it will often make sense to structure the assessment outputs around the
policy-relevant questions that were identified in the planning stages. This is a different approach than the one common in the scientific community, which would allow
for a more comprehensive analysis of the underlying science. Such a “science-led”
approach runs the risk of losing the policy-relevant assessment outputs among the
presentation of other materials.
A critical issue that each assessment must confront is whether or not to provide
specific recommendations for action. In general, experience shows that scientific assessments are far more effective if they do not make specific recommendations for
policy or other actions. The experience of global assessments to date is that it is
most effective to present findings that are policy relevant, as described here, but not
policy prescriptive. The rationale for this approach is that the choice among policy
options is never a strictly scientific one. Science can legitimately inform policy options, but actually selecting the most desirable option is a political role. The transition in an assessment to selecting preferred options can undermine the credibility
and legitimacy of the science itself.
While the approach of providing policy-relevant but not policy-prescriptive findings is by far the best approach for global, regional, and even national assessments,
24 | A Manual for Assessment Practitioners
it may not apply in all cases. For example, community assessments will involve as
experts most of the individuals in certain communities (since they will have local
knowledge of the historical changes in the system, which can complement the scientific information being contributed from outside scientists.) In such a situation,
the experts are also the users of the assessment. It thus would become somewhat
artificial for that assessment to draw a boundary between the presentation of policyrelevant findings and the selection of the most promising options.
The following checklist of guidelines around presentation and content can help
ensure that key questions are appropriately addressed by an assessment:
• Present and synthesize information that is needed to answer the question: no
more and no less.
• If an authoritative source for information already exists from other assessments
at the same or other scales, then cite it directly (e.g., IPCC, World Water Development Report, Global Biodiversity Assessment). In this way new information
can be added to the existing assessment base, rather than being re-created.
• Do not just list the existing data or information—provide the assessment team’s
judgment concerning best estimate or range, likelihood, probabilities, uncertainties, and reliability if the quantity or value is relevant to policy decisions (i.e.,
rate of forest loss, rate of local species extinction, economic contribution of pollinators to agricultural production, etc.).
• Present all credible points of view or scientific results and identify areas of scientific disagreement. A “credible point of view” is one present in the peer-reviewed
literature unless there is a body of evidence that clearly negates it. Different perspectives should be included not only because they might be right, but also because by not citing different viewpoints or providing judgment on them directly,
you open the assessment up to charges of being biased and not rigorous, thereby
affecting the credibility of the findings.
• Do not make specific policy recommendations. However, it is perfectly feasible
to explore the likely consequences of different options using an “If . . . then
. . . ” approach—so, for example, a policy-relevant, nonprescriptive finding
might be that “If fertilizer use is reduced by 20% in a particular landscape,
then freshwater quality in the area is very likely to be improved, leading to
measurable increases in local fish harvests.”
• Generally, research findings should not be presented for the first time in the
assessment. All information in an assessment should have already been subject
to a peer review process (or equivalent for nonformal knowledge systems—see
Annex 2 for guidance on procedures for incorporating nonpublished, nonpeerreviewed material into assessments). However, it is fine to use existing peerreviewed models with different datasets—so, for example, a scenarios exercise
might run new estimates of changes in drivers through existing models.
• New datasets generated through assessments should undergo peer review as part
of the assessment process.
• If data or information do not exist to answer a policy-relevant question that
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 25
has been identified in user consultations, do not just ignore the question—
conclusions such as “there is insufficient scientific information to determine
whether . . .” are still relevant to decision makers.
The third way to enhance the policy relevance of the assessment is to state the levels
of certainty (or uncertainty) associated with the assessment findings. The treatment
of scientific uncertainty within assessments is also a particularly important factor
influencing the credibility of the process. Clear statements of what is unknown are
often as influential for policy makers as are statements of what is known with relative certainty. This is particularly important in the main messages at the highest
level of reporting, such as in any Executive Summary. The assessment of the state of
knowledge should reflect both the type and amount of evidence (e.g., observations,
interpretation of model results, or expert judgment) and the level of peer acceptance
or consensus. Further information on the treatment of uncertainty can be found in
Chapters 3, 4, and 5.
Any agreed language for stating levels of certainty should be appropriately and
consistently used. Some summaries may not need to use them at all or may use them
only once or twice. Specific uncertainty language is most appropriately used for
statements of most direct relevance to policy decisions. So the statement “alkaline
lakes are characterized by lower species richness than freshwater lakes” does not
need a statement of certainty, for example. The statement “climate change will cause
the most significant changes to ecosystem services in the following ecosystems in
the region: xxx and xxx (medium certainty)” should be accompanied by a statement of certainty, however, based on the evidence found in the scientific literature.
Uncertainty language should also be used when the finding reported depends on a
judgment of the group of experts who wrote the chapter rather than a reporting of
“fact.” So, for instance, “although the magnitude of the positive feedback to global
warming due to changes in each individual ecosystem feedbacks has low certainty,
the balance of evidence indicates a high certainty that together these ecosystem
changes have contributed to the recent increases in Arctic air temperature.” However, there is rarely a need to include uncertainty language when reporting on trends
or data unless the interpretation of the data required a judgment call by the experts.
So “the rate of forest loss over the past 20 years in the district has been 4 ha/year”
does not require a statement of certainty, although typically the level of certainty
would be reflected when a range is presented—that is, “152 hectares” implies a different level of certainty than “100–200 hectares.”
The final way to improve the policy relevance of the assessment is to present the
answers to the questions identified at the outset in a Summary document, an Executive Summary, or a “Main Messages” section. Summaries can contain more than
this, but they should be sure to contain the material most relevant to policy makers.
A good summary
• Highlights key policy-relevant findings supported by arguments/evidence;
• Frames the information presented in relation to the questions asked of the assessment by users—this often follows the assessment outline and structure but
does not need to have key findings for all the sections;
• Is brief but not cryptic and avoids ambiguity;
26 | A Manual for Assessment Practitioners
• Has a tone consistent with that of the underlying assessment;
• States the degree of certainty when necessary, highlighting robust findings and
key policy-relevant uncertainties;
• Contains no literature references but can have internal references to the underlying assessment;
• Identifies gaps in knowledge and so is clear about what is known and what is
not known; and
• Is reader-friendly, written in an accessible and not too technical style (see Box
Box 1.4. Some useful writing suggestions for assessment reports
These suggestions are based on comments received during the MA peer review process.
• Avoid writing assessments to read like textbooks aimed at giving a general appreciation
for the systems discussed and possible problems they may be experiencing. Actual assessment information about problems, threats, and actionable items becomes buried in
these sections. Generally speaking, putting the educational material first is a turnoff for
decision makers. The problems and actions should be discussed first. Necessary educational material should be intermingled with that discussion. General descriptions of systems could be appended or referenced to other available information. The assessment is
what needs to be up front, not the background.
• Avoid passive voice and focus on definable measures and actions. For example, “there
are reasons to believe some trends can be slowed or even reversed” is a totally ignorable
statement in policy circles. If there are some opportunities for reversal, state precisely
what we believe they are, as best we know.
• Statements like “ . . . might have enormous ramifications for health and productivity . . . ,”
while they seem to the scientist to be strong because of the word “enormous” are actually politically impotent because of the word “might.” If data were used in the assessment,
what does they say about what “is” happening? What can we recommend, based on best
knowledge, about what actions would be effective?
• Statements like “There is a long history of concern over the environmental effects of
fishing in coastal habitats, but the vast scope of ecological degradation is only recently
becoming apparent (citation)” is a case where something strong could be said, but it is
weakened by putting the emphasis on the late arrival of this information and knowledge
“becoming apparent.” It does not matter so much when the degradation was discovered,
what matters is that it was. Cite the source and say “fishing practices are causing widespread destruction.”
• Do not use value-laden, flowery, or colloquial language (e.g., “sleeping dragon,” “elephant
in the room,” etc.).
• Statements like “we do not yet have clear guidelines for achieving responsible, effective
management of natural resources” could result in a legitimate policy response of “OK,
so we’ll wait until we do.” Instead, the statement could be changed to recommend what
needs to be done, such as “if clear guidelines were developed, then . . . ”.
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 27
Additional Resources
Chapter 6: Assessment Process, in Millennium Ecosystem Assessment, Ecosystems and Human
Well-being, Volume 4: Multiscale Assessments (Washington, DC: Island Press, 2005). This chapter of the Millennium Ecosystem Assessment sub-global working group provides an analysis of
the lessons learned in implementing the sub-global assessments of the MA. Available from www.
Ecosystem Services: A Guide for Decision Makers, (Washington, DC: World Resources Institute,
2008). This guide demonstrates how a city mayor, a local planning commission member, a provincial governor, an international development agency official, or a national minister of finance,
energy, water, or environment can use information from an integrated ecosystem assessment to
strengthen decisions. It forms a companion volume to this practitioner’s manual. Available from
MA Sub-global assessment network. The network of sub-global assessments established during or
since the MA provides a forum for exchange of ideas and best practices. Representatives from
the various assessments meet periodically to share experiences. (
The Global Environmental Assessment Project is an international, interdisciplinary effort directed
at understanding the role of organized efforts to bring scientific information to bear in shaping
social responses to large-scale environmental change. The focus of the Project was the growing number of assessments—ranging from the periodic reports of the Intergovernmental Panel
on Climate Change to the Global Biodiversity Assessment to the MA—that had recently been
conducted in support of international policy making. Its focus was to understand the impacts
of environmental assessments on large-scale interactions between nature and society and how
changes in the conduct of those assessments could alter their impacts. The Project attempted
to advance a common understanding of what it might mean to say that one effort to mobilize
scientific information is more “effective” than another. It tried to view such issues from the
perspectives of the scientific experts involved in producing assessments, the decision makers at
multiple scales who use those assessments, and the societies affected by the assessments. (www.
28 | A Manual for Assessment Practitioners
annex 1
A step-by-step guide to the assessment process
Exploratory Stage (steps to be taken simultaneously)
1. Establish the need or demand for an assessment of ecosystem services and the
consequences of ecosystem change for human well-being at a particular scale.
2. Convene a technical and user planning group to determine the feasibility, scope,
and extent of the assessment. Draft a scoping study.
3. Interact extensively with the intended users of the assessment—for instance, recruit an advisory board consisting of end users and stakeholders and modify the
scoping document based on their subsequent review.
4. Explore funding support and generate funding proposals.
Design Stage
5. Establish the governance and institutional structure: technical executive group,
advisory board, contracts and memoranda of understanding between participants and funding agencies.
6. Communicate the objectives, agenda, and opportunities for involvement to the
stakeholders through Internet, newsletters, and presentations.
7. Assemble an initial assessment team based on competence in the topic areas selected, experience in the geographical area, availability to participate, and credibility with stakeholders.
8. Through an assessment team workshop, draft an Implementation Plan detailing
who is responsible for what actions, by when. Review it in the light of comments from the advisory board.
9.Agree on a shared conceptual framework among all assessment practitioners
and users.
Assessment Implementation and Communications Stage
10. Perform preliminary assessments of each of the focus areas identified in the
scoping study, focusing on the three elements of the assessment: condition and
trends, scenarios, and responses.
11. Continue interactions with the intended users of the assessment through workshops and briefings. This will help ensure the assessment is still of value and aid
in development of a communications strategy that considers how best to deliver
the preliminary findings and assessment reports to different user groups.
12. Circulate the assessments among the assessment team and affiliated technical
experts for “internal” peer review.
13. Key issues for which there is no credible (preferably published, peer reviewed)
supporting material may require a focused, high priority research action. Commission or encourage such studies as early as possible and submit the results to
a peer-reviewed, open literature publication process.
14. Revise the preliminary assessments in the light of the review comments. Draft an
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 29
Executive Summary (typically 2–5% of the length of the full text) to accompany
the full assessment.
15. Circulate the updated draft to at least one independent technical expert per topic
16. Revise the draft in the light of the expert reviewers’ comments. This is the last
opportunity to incorporate new material.
17. Finalize communications and outreach strategy and begin implementation.
18. Circulate the updated assessment draft to the experts who previously reviewed
it, additional experts, the users and other stakeholders identified during scoping, and the advisory board.
19. Revise the draft in the light of the comments received to create the penultimate
draft assessment report. Document the revisions made or the reasons for not
making them.
20. Submit the penultimate draft to the advisory board for approval. This board or
a separate review board should ensure that due diligence has been performed in
addressing reviewer comments.
21. Incorporate any revisions required by the board and perform final formatting
and graphics editing. No substantive additions or deletions not mandated by the
board should be introduced at this stage.
22. Publish and distribute the assessment through appropriate media: in a special
issue of a journal or as a stand-alone volume. Publication on the Internet is recommended as a secondary source.
23. Extract and distribute communication products from the assessment, taking
care not to deviate from the spirit of the accepted text. These may include policy
summaries, graphic-rich brochures, media releases, posters, maps, CD-ROMs,
videos, and radio interviews.
Scoping for Subsequent Assessment
24. Convene a workshop consisting of technical participants, stakeholders, users,
and donors to analyze the strengths and weaknesses of the assessment and to
document the lessons learned.
25. Explore the possibilities for and scope of a follow-up assessment.
26. Find an institutional home for the records and archives of the project and as a
point of contact until such time as a possible future assessment.
30 | A Manual for Assessment Practitioners
annex 2
Guidance for using nonpublished/
nonpeer-reviewed sources
Guidance for using nonpublished and nonpeer-reviewed sources into the MA was
provided to all authors involved in the process. The following text is taken from
this guidance and may serve as a useful reference for assessment initiatives that are
seeking to benefit from the incorporation of knowledge from sources beyond the
peer-reviewed scientific literature.
Because considerable materials relevant to MA Reports, in particular, information
based on indigenous, traditional, or local knowledge or information about the experience and practice of the private sector, are found in sources that have not been published or peer-reviewed (e.g., industry journals, internal organizational publications,
nonpeer-reviewed reports or working papers of research institutions, proceedings of
workshops, personal communication, etc.) the following additional procedures are
provided. These have been designed to make all references used in MA Reports easily accessible and to ensure that the MA process remains open and transparent.
1. Responsibilities of Coordinating, Lead and Contributing Authors
Authors who wish to include information from a nonpublished/nonpeer-reviewed
source are requested to:
a. Critically assess any source that they wish to include. This option may be used for
instance to obtain case study materials from private sector sources for assessment
of adaptation and mitigation options. Each chapter team should review the quality and validity of each source before incorporating results from the source into
an MA Report.
b.Send the following materials to the Working Group Co-Chairs who are coordinating the Report:
• One copy of each unpublished source to be used in the MA Report
• The following information for each source:
– Title
– Author(s)
– Name of journal or other publication in which it appears, if applicable
– Information on the availability of underlying data to the public
– English-language executive summary or abstract, if the source is written in
a non-English language
– Names and contact information for 1–2 people who can be contacted for
more information about the source.
c. Information based on personal communication from individuals with indigenous,
traditional, or local knowledge, or direct input as a member of a working group
by an individual with indigenous, traditional, or local knowledge should be handled in the following way:
i. In situations such as local assessments where extensive use of local and traditional knowledge will be involved, the assessment must establish a process
Assessing Ecosystems, Ecosystem Services, and Human Well-being | 31
of validation for the findings as part of the application by the assessment to
become a component of the MA. The features of such a validation process are
described in Section 4.2.
ii.Metadata concerning the personal communication (e.g., names of people interviewed, dates and types of notes recorded, presence or absence of self-critical
review notes by the researcher, sources of ‘triangulation’, etc.) should be made
available to the Co-Chairs of the Working Group.
iii.Where an individual provides direct input of indigenous, traditional, or local knowledge as a member of a working group, the individual should provide the Working Group Co-Chairs coordinating the report the following
– Basis for knowledge of the particular issue (length of time living in the area,
individuals from whom historical information was obtained, etc.)
– Names and contact information for 1–2 people who can be contacted for
more information about the source.
2. Responsibilities of the Review Editors
The Review Editors will ensure that these sources are selected and used in a consistent manner across the Report.
3. Responsibilities of the Working Group Co-Chairs
The Working Group Co-Chairs coordinating the Report will (a) collect and index
the sources received from authors, as well as the accompanying information received about each source and (b) send copies of unpublished sources to reviewers
who request them during the review process.
4. Responsibilities of the MA Secretariat
The MA Secretariat will (a) store the complete sets of indexed, nonpublished sources
for each MA Report not prepared by a working group and (b) send copies of nonpublished sources to reviewers who request them.
5. Treatment in MA Reports
Nonpeer-reviewed sources will be listed in the reference sections of MA Reports.
These will be integrated with references to the peer-reviewed sources stating how
the material can be accessed, but will be followed by a statement that they are not
32 | A Manual for Assessment Practitioners
Fabricious, C., R. Scholes, and G. Cundill. 2006. Mobilizing knowledge for integrated ecosystem
assessments. In Bridging scales and knowledge systems: Concepts and applications in ecosystem
assessment, ed. W. V. Reid, F. Berkes, T. Wilbanks, and D. Capistrano, 165–82. Washington, DC:
Island Press, for World Resources Institute.
MA (Millennium Ecosystem Assessment). 2005. Ecosystems and human well-being: Current state
and trends. Washington, DC: Island Press.
Reid, W. V., F. Berkes, T. Wilbanks, and D. Capistrano, eds. 2006. Bridging scales and knowledge
systems: Concepts and applications in ecosystem assessment. Washington, DC: Island Press, for
World Resources Institute.
Tomich, T. P., K. Chomitz, H. Francisco, A. M. N. Izac, D. Murdiyarso, B. D. Ratner, D. E. Thomas,
and M. van Noordwijk. 2004. Policy analysis and environmental problems at different scales:
Asking the right questions. Agriculture, Ecosystems and Environment 104 (1): 5–18.
WRI (World Resources Institute). 2008. Ecosystem services: A guide for decision makers. Washington, DC: WRI.
Stakeholder Participation, Governance,
Communication, and Outreach
Nicolas Lucas, Ciara Raudsepp-Hearne, and Hernán Blanco
What is this chapter about?
In this chapter the case is made for developing an assessment process that is relevant, credible, and legitimate to stakeholders and end users, and guidance is offered on how to implement these principles. The assessment process is as important as the reports it produces, and
therefore it is crucial to be strategic when managing stakeholder engagement and participation, organizing the assessment process, and communicating with stakeholders. Section 2.1
describes the principles of relevance, credibility, and legitimacy and offers suggestions on how
to achieve these principles within the assessment process. Section 2.2 presents approaches
for exploring the need for an assessment and then initiating the process. Sections 2.3, 2.4,
and 2.5 provide a blueprint for organizing and managing the assessment process, based on
past experiences with integrated ecosystem assessments at the sub-global scale. And, finally,
section 2.6 offers information on how to develop and communicate messages based on assessment findings.
2.1 How to ensure relevance, credibility, and legitimacy
Section’s take-home messages
• The three fundamental qualities of a sound assessment are relevance, credibility, and
legitimacy. Is assessment information significant in relation to an actor’s priorities or
decision-making issues? Does the assessment meet standards of scientific rigor and
technical adequacy? And do participants perceive the assessment process as unbiased
and meeting standards of political and procedural fairness?
• Assessments need to strike a balance between relevance, credibility, and legitimacy
and, if possible, promote synergies between them. At the local level, these qualities are
sometimes at odds with those at the national or global level.
• The core values of relevance, credibility, and legitimacy are best achieved through
strategic and effective participation in the assessment process.
34 | A Manual for Assessment Practitioners
How an assessment process is organized is just as important as the information it
produces. A sound assessment design helps ensure that audiences will be interested
in the assessment findings and able to make use of the information in their decision
making. Normally, scientists, managers, and policy makers focus their efforts on
ensuring the credibility of technical information by attempting to produce authoritative, believable, trusted technical reports. However, recent research and practice has
shown that underemphasizing the relevance and legitimacy of ecosystem information, which depend on how an ecosystem assessment is conducted, can lead to ineffective and ultimately futile exercises (see Box 2.1) (Mitchell et al. 2006, Farrell and
Jager 2005, Eckley 2001).
Information-gathering processes are more likely to be seen as legitimate and to
be used more constructively in policy making when they are transparent and involve
those who will be affected by any decisions influenced by them. This section explains the three fundamental qualities of a sound assessment—relevance, credibility,
and legitimacy—and gives examples of how they might be achieved.
2.1.1 Relevance, credibility, and legitimacy
Relevance, or salience, refers to the significance of assessment information in relation to an actor’s priorities or decision-making issues (Cash et al. 2002). Relevance
is measured both in terms of contents (the questions that the assessment answers,
the issues it addresses) and timing (information that feeds too late or too early into a
decision process is likely to be considered irrelevant). Issues of interest to researchers
may not be the same ones that interest decision makers, and, ultimately, information
will be relevant if judged useful by decision makers rather than by researchers, as
this is the goal of an assessment.
Assessment information must help decision makers address relatively precise
questions. For example, if the policy goal is to reduce rural poverty, the information
must answer what ecosystem services are critical to rural livelihoods, what is their
status of degradation or enhancement and how they might change in the future,
what are the actual and expected consequences for human well-being, and how different decisions or policies will modify the supply of ecosystem services to the poor.
Designing an assessment in order to produce relevant information begins with
framing the assessment around questions of importance to target audiences (see
Box 2.1. Lessons from international processes
In 1995 some 1,500 scientists released the Global Biodiversity Assessment. This 1,152-page
high-quality report was produced largely independently from its primary intended audience—
namely, governments who were parties to the Convention on Biological Diversity. As a result
of this legitimacy deficit, the information was not welcome and ultimately the parties ignored
the report for the most part (Raustiala and Victor 1996). Similarly, the early stages of the Intergovernmental Panel on Climate Change (IPCC) lacked legitimacy in the eyes of participants
from developing countries, who saw that greater participation of their scientists was needed
(Agrawala 1998). This was addressed in later stages, and the IPCC in time became enormously
influential globally.
Source: Philippines and Chile SGA teams, personal communication.
Stakeholder Participation, Governance, Communication, and Outreach | 35
section 2.4). A “user needs assessment” can be conducted to find out what these
questions are. Both the assessment design and audience needs may evolve over time,
and it is therefore important to have regular communication with targeted audiences
throughout the assessment to maintain a fit between information and needs. Finally,
how the information is conveyed to different audiences has implications for how
relevant they find it (see section 2.6).
Temporal relevance is also important. If information is being developed to fit
directly into a decision-making process that will take place during a certain time
period, the assessment must be concluded before that period. This may mean making decisions to drop elements of the assessment in order to meet the deadline, as
a slightly more compact assessment delivered on time is more useful than a longer
assessment report that is obsolete. Firm dates for when the assessment findings will
be released and ready for use should be negotiated early on in the process.
Credibility refers to whether the assessment meets standards of scientific rigor
and technical adequacy. Sources of knowledge must be considered trustworthy and/
or believable, along with the facts, theories, and causal explanations invoked by
these sources (Cash et al. 2002). Technical information intended to feed into decision making is always subject to criticism, especially from those who stand to lose
from the resulting decisions, so it is important that the assessment process builds the
credibility of the information generated (Ranganathan et al. 2008).
What is judged to be believable will differ between communities, and therefore
it is important to understand key audiences well (see section 2.6). While some audiences may consider published scientific papers to be a credible source of information, some local actors may find information to be more credible if it fits into their
cultural context and knowledge system. Local and traditional knowledge can be
included in a scientific assessment, but it must be validated and reviewed in a manner that is credible to all stakeholders. Assessments are powerful policy tools because they aim to include information and expertise from multiple disciplines, and
sometimes also different forms of knowledge. Integrating these different sources of
information results in more broadly credible findings.
The process by which knowledge and information are collected, sorted, and
made sense of needs to be highly transparent in order to achieve broad credibility.
Before and during the assessment process, transparency can be emphasized by developing and publicizing process documents that describe:
The assessment process;
The conceptual framework;
Guidelines for participation;
Guidelines for how to validate unpublished knowledge, including local and
traditional knowledge;
• Guidelines for dealing with uncertainty; and
• Guidelines for the review process.
Legitimacy refers to whether an actor perceives the assessment process as unbiased
and meeting standards of political and procedural fairness, and whether the process
considers appropriate values, concerns, and perspectives of different actors. Audiences judge legitimacy based on who participates and who does not, the process for
making those choices, and how information is produced, vetted, and disseminated
36 | A Manual for Assessment Practitioners
(Cash et al. 2002). Legitimacy refers to the politics of the information-gathering process. It is a function of the political context into which the information is released. In
essence, the goal of a strategy to establish legitimacy is to involve users to the point
where they invest sufficiently in the process that they will adopt and use the information produced, or at least not reject it. This requires keeping users informed and giving them a chance to influence the process so that, at the very least, they do not feel
threatened by it. And in an ideal situation, they “own” the process and its results.
The extent of user engagement can range from reporting on progress to information
users to the users being centrally involved in the production of the information. In
any event, at the very least users need to be comfortable with the motivation, origins,
and general orientation of the technical work.
During the exploratory and design stages, the global Millennium Ecosystem Assessment (MA) proponents spent a lot of time discussing the project with the targeted
stakeholders to build the legitimacy base. Meetings were held with leading countries
and figures within the three main ecosystem-related international conventions (the
Convention on Biological Diversity, the Convention to Combat Desertification, and
the Ramsar Convention on Wetlands), and presentations made to their governing
bodies, which, early on in the MA process, formally expressed their interest and
expectations. In addition, multiple contacts were established with 22 representative
institutions that were eventually invited to form the Board of Directors for the project. Throughout the assessment, efforts were made to open the process to the participation of national and local stakeholders through a variety of means (from user
forums to the circulation of drafts for review). In the end, this huge effort delivered
a very sound legitimization, as well as a very valuable platform for outreach.
Many of the MA sub-global assessments observed that navigating the politics of
stakeholder participation often results in a trade-off between advancing with technical work and ensuring the legitimacy of the process. Broadening and deepening
participation may conspire against expediency in technical tasks and also subject
scientists and experts to policy discussions that they are not prepared for or comfortable with. A balance between technical work and participation must be found
without compromising either. Stakeholders can sense when their participation is
merely lip service or considered a “distraction from the real work.” Even minimal
stakeholder participation can be made more effective by paying attention to the
• Ensure participation is built into actually relevant stages of the assessment
• Invite key stakeholders to participate in the assessment’s governance structure.
• Establish communication channels between stakeholders and technical experts
involved in the assessment to clarify uncertainties and verify assumptions.
• When stakeholders are invited to contribute based on their experience, make
sure they can recognize their inputs in the analysis and reports, and inform
participants that their contribution and participation will be properly acknowledged in outputs.
• Where possible, ensure stakeholder inputs are recorded and made available to
• When the assessment is meant to serve the needs of a large organization, ensure
Stakeholder Participation, Governance, Communication, and Outreach | 37
that the assessment is formally recognized and supported by the governing bodies of the organization.
2.1.2 Synergies and trade-offs between relevance, credibility,
and legitimacy
Relevance, credibility, and legitimacy are interconnected, positively and negatively.
In the previous section they were presented separately for the sake of clarity, but in
reality actions to strengthen one may have both positive and negative consequences
on the others (Ranganathan et al. 2008). For example, the involvement of government officials in the review process may be undertaken initially to ensure the relevance of the drafts being produced, but it is likely to enhance the legitimacy for the
assessment and add to its technical credibility with some audiences as well.
On the other hand, the enhancement of one quality can come at the expense of
another. Traditionally, the scientific community has favored credibility over legitimacy and relevance as its main concern when generating information, the assumption being that technical and scientific information needed to be walled off from
the potentially biasing influence of politics, prepared with quality control of peer
review, and delivered to decision makers in the form of reports (Cash et al. 2002).
This strategy does protect the technical credibility of the information, but often at
the expense of making it less relevant to policy makers. Similarly, when policy makers are allowed to develop the substantive contents of the assessment reports (as opposed to only deciding on the questions that frame it), the legitimacy of the reports
may in some cases be enhanced, at the cost of becoming less technically credible.
A good design for an assessment needs to strike a balance and, where possible,
promote synergies between relevance, credibility, and legitimacy. This balance needs
to be found across scales as well. The experience of many sub-global assessments
was that local relevance, credibility, and legitimacy were sometimes at odds with
achieving these qualities at national or global scales. In these cases, the assessment
team might choose to focus on meeting the expectations of the primary audience,
but they might also develop strategies to work with audiences at other scales. For
example, an extensive and inclusive review process is a strategy for achieving credibility and legitimacy across scales. Most of the MA sub-global reports were reviewed by global experts in addition to experts and stakeholders from the most
relevant scales.
2.1.3 Participation as a strategy for achieving relevance, credibility,
and legitimacy
The core values of relevance, credibility, and legitimacy are best achieved through
strategic and effective participation in the assessment process. To participate is to
share, and a participatory process involves having different stakeholders engaged in
an interactive process that promotes knowledge and information exchange and that
allows them to express their positions and interests on issues.
An integrated ecosystem assessment requires blending knowledge and perspectives from many different points of view. It usually aims to influence audiences with
different interests and information needs. In order to maximize impact, it is essential
38 | A Manual for Assessment Practitioners
that a wide range of actors participate throughout the process, either as contributors
or audiences—or both. This helps identify key issues that matter most within a given
context, strengthens the analysis of perceived changes in ecological and social systems, and builds ownership of the assessment’s findings among audiences who are
supposed to follow up with action. Participation can take the following forms:
• Being consulted on the need for an assessment;
• Being consulted on key questions framing the assessment;
• Receiving information about assessment progress, findings, and opportunities to
• Contributing knowledge to the assessment report;
• Contributing contextual information about an ecological or social system;
• Being consulted on the condition and trends of ecosystem services and human
well-being in a region (practitioners and holders of local knowledge);
• Attending public hearings about assessment process and findings;
• Attending education or capacity-building workshops on assessment process and
• Participating in the assessment process as student interns or fellows of the
• Becoming a member of the advisory committee;
• Being a formal end user of the assessment products;
• Reviewing assessment materials; and
• Acting as a partner for the dissemination of assessment findings.
Some forms of participation are considered necessary for an assessment (e.g., consulting potential end users on key questions to frame the assessment), while other
forms may or may not be necessary or possible (e.g., consultations on the conceptual
framework or methodology to use, or parliamentary intervention to obtain formal
endorsement). The assessment team generally decides on participation strategies
based on the desired impact of the assessment on different stakeholder groups. The
team will generally emphasize spending energy on encouraging the most important
stakeholder groups to take ownership of the process and outcomes. Communication with target stakeholders should be sustained throughout the entire assessment
Improving information through participation
In addition to building relevance, credibility, and legitimacy, encouraging broad
participation in the writing of the assessment report can improve the quality of
the findings. To assess ecosystem services in relation to human well-being, technical
teams must first of all be multidisciplinary and incorporate understanding of natural
and social sciences. (See Chapter 4.) At smaller scales, the incorporation of local
stakeholder knowledge into integrated assessment has recently been established as
critical (Pahl-Wostl 2003). This is because local stakeholders have particular and
unique knowledge about the ecosystems where they live and work, as well as about
their own associated well-being. Often the links between ecosystem services and
human well-being are complex and obscured and must be teased out with contextual knowledge. In addition, published data about ecosystems and societies in many
Stakeholder Participation, Governance, Communication, and Outreach | 39
parts of the world is scarce and can be much enhanced with local knowledge. In the
Peruvian sub-global assessment, communities gathered to assess the condition and
trends of water and soil by consensus in an area with little data availability.
The MA encouraged the use of local, traditional, and practitioner knowledge in
assessments. However, in order to be credible and useful to decision makers, the MA
Conceptual Framework stated that “all sources of information, whether scientific,
traditional, or practitioner knowledge, must be critically assessed and validated as
part of the assessment process through procedures relevant to the form of knowledge” (MA 2003). Chapter 4 describes methods for validating different types of
Beyond sound information: Participation in assessments as a moral issue
To build relevance, credibility, and legitimacy, an assessment process needs to be
inclusive and open to different stakeholders. Openness, transparency, and participation are therefore justified on practical grounds—that is, the need to ensure that assessment information is effectively used by decision makers. But the social dimension
of an assessment is not fully captured by this justification. Assessments, especially
when they involve governments or publicly held information, and precisely because
they are intended to influence decisions with a public impact, are part of the cultural
life of a community and the advancement of science in it. The right to participate in
such a process has been recognized by the Universal Declaration of Human Rights in
Article 27 (1): “Everyone has the right freely to participate in the cultural life of the
community, to enjoy the arts and to share in scientific advancement and its benefits.”
Moreover, an open assessment is a process of democratization of both information
and governance.
In practice, the composition of stakeholder groups invited to join the assessment process is under control of those proposing the assessment. An assessment of
relevant stakeholders and their information needs is one way to establish the target
audiences (see section 2.3), but keeping the process open to interested parties promotes participation as a right, even when the degree to which all stakeholders can
participate will be limited by time, funding, and language constraints.
2.2 How to establish the need for and scope of an assessment—
the exploratory stage
Section’s take-home messages
• It is of strategic importance to develop the specific objectives of an assessment according
to what is possible—in order to keep expectations realistic—and what is needed and
sensible—in order to get as much support as possible from potential stakeholders. The
subtle, early work of an assessment involves discussing its rationale with a diversity of
stakeholders, which requires understanding and integrating their diverse perspectives and
• The first and most obvious scope to be defined is the geographic extent or spatial scale of
the assessment. There are trade-offs to consider for both large- and small-scale efforts.
• In the exploratory phase, a user needs assessment can explore who should be included
40 | A Manual for Assessment Practitioners
in the assessment, and in what capacity. This would identify the information that different
stakeholders need about ecosystem services. Ideally it would also produce a database
listing all potential stakeholder groups, plus information about their relation to specific
ecosystem services and their potential and capacity for engaging in the assessment.
An ecosystem assessment is a technical, social, and political process. It is technical as
it entails the use of sound science, it is social as it engages a diversity of actors in a
collective endeavor, and it is political as it involves different stakeholder interests. An
assessment may be mandated by law, but it still requires leaders who first sense the
need for an assessment and then work with others to make it happen. Assessment
initiators might be direct users of ecosystem services, scientists, decision makers, or
any other actor interested in the environment and its link to human well-being. This
section aims to help these early assessment promoters; it also highlights some of the
main challenges that might be encountered in the early stages of the process and suggests ways to deal with them.
2.2.1 How to define the need for an assessment
Sometimes the need for an assessment will be evident to users or decision makers.
Sufficient and valid conditions to consider initiating an assessment process include:
• A lack of information about trends in ecosystem services;
• Particular events (such as unusual floods) that highlight the lack of knowledge
about ecosystems in a region;
• Observed decreases in the quality of ecosystem services;
• Latent or overt conflicts among users over the use of ecosystem services;
• Poor or unsatisfactory levels of governance of natural resources;
• Impacts on human well-being; and
• Negative or uncertain prospects for the future of local ecosystems.
These conditions are common in most developing and industrial countries. At this
early point in the assessment, it is of strategic importance to develop the specific
objectives of the work according to what is possible—in order to keep expectations
realistic—and what is needed and sensible—in order to get as much support as possible from potential stakeholders. Promoters of an assessment are responsible for the
subtle, early work of discussing the rationale for an assessment with a diversity of
stakeholders, which requires understanding and integrating their diverse perspectives
and interests. A number of MA sub-global assessments held exploratory workshops,
for example, to discuss the need for an ecosystem assessment with local people.
The promoters of an assessment then have to exercise the art of balancing expectations and possible results of the assessment. In particular, promoters will have to
prepare and present the project in such a way that the principal stakeholders—users
of the ecosystem and users of the assessment results—can find value in it. (See Boxes
2.2 and 2.3.) This is a challenging task, particularly considering that the results of
such an assessment are, in general, medium to long term and not as tangible as most
typical development projects. The task is made easier if the promoters are themselves stakeholders in the system and well integrated into the local decision-making context. In many cases, however, promoters are scientists or nongovernmental
Stakeholder Participation, Governance, Communication, and Outreach | 41
Box 2.2. Initiating an ecosystem assessment
Defining the need for an assessment in the Laguna Lake Basin, Philippines
Prior to the Philippines MA subglobal assessment, the scientific community of Los Baños was
heavily involved in different research programs that focused on the Laguna Lake Basin and its
management, because of its location next to important urban centers. The assessment was
viewed by the University of the Philippines Los Baños (UPLB) as an opportunity to connect to
the ongoing efforts of many agencies for collaborative management of the lake. The assessment initiators (one forestry scientist and one social scientist at UPLB) then invited key people
from the Laguna Lake Development Authority (LLDA), the agency with the mandate to manage
the region, to form the core assessment team. With this group formed, other agencies engaged
in the research, use, and management of economic activities in the lake basin were invited
to participate. Agencies like the Philippine Marine and Aquatic Resources Management and
Development of the Department of Science and Technology and the Bureau of Fisheries and
Aquatic Resources of the Department of Agriculture sent their own scientists to the workshops
and meetings. Together with the core team, a scientific committee was loosely organized and
contributed to the assessment of the different ecosystem services.
To validate the objectives of the assessment, the core assessment team met with LLDA
executives, who facilitated a meeting with the Secretary of the Department of Environment and
Natural Resources and with government executives of the two main provinces involved. This was
when the need for such an assessment was fully established and legitimized, which ensured the
acceptability of the process to various stakeholders. The proposed assessment objectives and
process were finally presented to a wider forum in Manila, where representatives of other sectors
were invited in order to get inputs from the wider audience. In retrospect, the assessment team
felt that it should have included voices from the local communities in the region.
Recognizing the potential of assessment work in the Salar de Atacama, Chile
In the Salar de Atacama, Chile, a lack of access to information on water quality and quantity
in one of the driest areas in the world made it difficult for the users of that resource to design
an acceptable water management plan. Latent conflict existed among these resource users
(including mining companies, tour operators, and the indigenous local communities), in part
due to the lack of information. Because the MA was soliciting assessment projects and offering
some funding, capacity building, and a credible international network, researchers in Chile developed a proposal for a subglobal assessment. They saw the assessment as an opportunity to
bring stakeholders together to discuss different points of view and to establish credible baseline information on water resources that were at the center of stakeholder conflict. Indigenous
people were a central stakeholder in this assessment.
Source: Maria Victoria O. Espaldon, University of the Philippines Los Banos, personal
organizations (NGOs) who must work closely with local leaders in order for them
to eventually take ownership of the process and outcomes.
Assessment promoters may also be faced with unsupportive contexts. For instance, local authorities might have a record of negative experiences with environmental research projects that have been extractive as opposed to collaborative. If this
is the case, promoters will have to make every effort to differentiate the assessment
42 | A Manual for Assessment Practitioners
Box 2.3. Getting an assessment started in Western China
In October 2006, the Ministry of Science and Technology of China funded a follow-up assessment of China at the national level, based on the experience of the MA subglobal assessment
of Western China, the results of which are currently being used to inform policy and action
in that area. The municipality of Qing-Yang was selected as one of the case studies for the
China assessment. In November 2007 the assessment team organized a workshop in QingYang, which was attended by the mayor of the municipality and officials from several public
After the workshop, the mayor expressed a strong interest in exploring trends in local
ecosystem services, and future scenarios for the region. The Qing-Yang municipality asked the
national assessment team to help them develop tools for ecological management and to develop a decision-support system for development and conservation of the Dong-Zhi Tableland,
which is the economic and political center of Qing-Yang Province and includes four counties.
In 2008, Qing-Yang municipality collected data and maps and provided them to the national assessment team for analysis. At the same time, the Qing-Yang municipality organized
six officials and 70 technicians from different divisions to investigate trends in ecosystems and
driving forces. These officials and technicians were trained by scientists from the national assessment team in methods of data collection and processing. The setting up of data collection
procedures and training has established conditions that will be amenable to future ecological
management and planning by the local government.
Source: Prof. Dr. Tian Xiang YUE, Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences, personal communication.
process from previous research endeavors. Spending enough time explaining how
and why the assessment can be useful to local stakeholders, and discussing the perceived and potential links between ecosystem services and local health, well-being,
and economic development, is key. One great value of an assessment process is the
chance to raise awareness about human dependence on ecosystem services. An assessment process can contribute to developing knowledge about the state of ecosystems so that decision makers can make smart and sustainable decisions to benefit
human well-being. When referring to outcomes, promoters can thus emphasize the
following benefits of conducting an assessment:
• The generation of and access to relevant information on the condition and
trends of ecosystem services that support local populations in different ways;
• The opportunity for constructive stakeholder participation and collective
learning opportunities;
• The opportunity for directed capacity-building processes;
• The integration and coordination of diverse private and public initiatives; and
• State-of-the-art information for better ecosystem management and increased
human well-being.
2.2.2 How to define the scope of an assessment
The most obvious scope to be defined is the geographic extent or spatial scale of the
assessment. Sub-global assessments within the MA demonstrated that an assessment
Stakeholder Participation, Governance, Communication, and Outreach | 43
approach centered on the MA conceptual framework can be implemented from
very local scales to large, regional scales. The reasons (or motivations) behind the
need for an assessment will dictate the appropriate scale. For example, if assessment
stakeholders are interested primarily in how land use is affecting water quality in the
region, the watershed scale may be the appropriate one to consider. The choice for
the preferred spatial scale is also determined by the ecosystems themselves, how they
function, and the location of the ecosystem users.
The MA advocates a multiscale assessment process, which integrates assessments at diverse scales into one overarching assessment. (See Chapters 3 and 4.) The
benefits of conducting an assessment at multiple scales are many (see MA 2003);
however, this type of process will inevitably require more time and resources.
There are trade-offs to consider when selecting the appropriate spatial scale. A
large scale might allow for a comprehensive analysis of relevant social and ecological systems and interactions between them. But it will demand more resources, and
the findings may be more difficult to integrate into a focused policy-making process.
Larger scales will also pose a number of challenges to the participatory process, the
governance structure, and the communication and outreach initiatives. For example,
stakeholders might not recognize the relevance (or urgency) of the assessment if it is
not entirely focused on issues of importance to their local context.
A small-scale assessment, on the other hand, might facilitate coordination and
integration with the policy- and decision-making process, but it might lose a wider
perspective on important issues occurring outside the chosen scale. A smaller-scale
process might also facilitate the organization of the participation, governance, communication, and outreach aspects of the assessment, although this is not always the
case. Because assessments at smaller scales are focused on issues of direct relevance
to local stakeholders, they tend to be more interested in the outcomes, as well as
more divided on issues that are controversial, and the participatory process can thus
become an extremely intensive part of the assessment process.
In practical terms, there is usually a demand generated for an assessment at a
specific scale, generally coming from the most enthusiastic stakeholders. They will
be able to determine (sometimes with the help of scientists) what spatial scale is
most relevant to their decision making and to the ecosystem services of interest to
them. When seeking to define the boundaries of the assessment area, there are two
aspects to consider:
• The natural (geographic) boundaries: For example, river basins are natural
boundaries of ecosystems; they also allow for an integrated analysis of the
relationship between water resources, soil, flora, and fauna. The exception
might be the case of groundwater resources (and related ecosystems) that in
certain cases do not follow the superficial boundaries of basins. Additionally,
some situations might not follow the rationale of geographic boundaries; one
such case would be when environmental services are imported from a different
country or region.
• The political–administrative boundaries: Most administrative divisions within
countries (and between them) do not follow natural limits such as river basins
or ecosystems. Institutions, governments, and decision-making processes, which
often have a large influence on the way ecosystems are managed, are crucial to
the assessment process and hence are a logical way to define spatial scale.
44 | A Manual for Assessment Practitioners
Other factors that need to be determined when deciding on the scope of the assessment include how many and what ecosystem services will be assessed, over
what timeframe the ecosystem services and human well-being will be assessed, and
whether sophisticated technical approaches such as modeling will be used. These are
all dictated by the goal of the assessment, and while the initiators will have an idea in
advance about these factors, they will need to be discussed further as the governance
structure and technical team are assembled (see section 2.3). Chapter 4 goes into
further detail about how to develop the substantive elements of the assessment.
2.2.3 How to identify and engage relevant and diverse actors
A user needs assessment at the beginning of the assessment process is a good way to
begin a stakeholder engagement strategy. This will include the communities, institutions, organizations, groups, and individuals that may be interested in the assessment process and findings, either because they affect ecosystem services or because
they themselves are affected by changes in such services. A user needs assessment
would ideally produce a database listing all potential stakeholder groups, plus information about their relation to specific ecosystem services and their potential and capacity for engaging in the assessment. More important, the needs assessment would
include a survey to identify the information that different stakeholders need about
ecosystem services.
This database could be added to and improved on as the assessment progresses.
One way to initiate the user needs assessment is to have a social scientist, preferably
with field experience, conduct desk research and then visit the diverse communities, organizations, and individual users. After several interviews, the social scientist
will be able to start consolidating a users database. However, there are many other
ways to conduct a user needs assessment, depending on the context and scale of the
A systematic understanding of the main ecosystem and assessment users, as well
as the main social, economic, and political components of the system, is central
to the success of the assessment process. This may sound like common sense, but
a thorough assessment of user needs is not often done. This exercise is not only
about getting access to information but, most important, about getting to know the
relevant stakeholders and initiating or strengthening relationships with them. It is
through discussing a potential assessment with stakeholder groups that a vision of
the assessment is formed and the more engaged stakeholders get on board formally.
In addition to the focus on users, a database identifying relevant initiatives related to ecosystem services (present and future) is also useful. This would include
initiatives such as new legislation, conservation movements, or health projects that
the assessment process might influence or contribute to. The database of potential
users and relevant initiatives can be complemented with a simple background study
of the region that would set the stage for developing the assessment. This might
• A local history;
• Important economic activities and basic human well-being indicators;
• Recent issues and conflicts and the way they have been dealt with;
Stakeholder Participation, Governance, Communication, and Outreach | 45
• Channels of communication used for communication amongst stakeholders; and
• The identification of leaders (formal and informal).
A brief report synthesizing the user needs assessment and assessment context study,
and highlighting the implications for the assessment process, will define how the assessment process is developed. For instance, when it comes to communicating and
involving stakeholders, knowledge about channels of communication and key local
leaders will be crucial. In and of itself, the user needs assessment—if done respectfully—will contribute to building a relationship with users.
Once the user needs assessment is completed, further work is required to
strengthen the involvement of stakeholders. As mentioned in section 2.2.1, the exploratory stage requires the patient work of meeting with a diversity of relevant actors, getting to know them, communicating the essentials of the assessment process,
and listening to their interests and positions. The following issues will need to be
discussed with stakeholders in order to define their roles within the developing assessment process:
The spatial scale of interest;
Important ecosystem services;
Availability of information about ecosystem services and human well-being;
Governance structure for the assessment;
Participation and communication activities;
Coordination with existing initiatives;
Consideration of traditional knowledge; and
Possible uses of the assessment results.
At this stage, the assessment initiators will have to decide who the key users are. (See
Box 2.4.) These might be the groups that are most affected by changes in ecosystem
services and/or groups that have the most influence over managing the services. In
some cases, it may be the groups that are most interested in the process, regardless
of their relationship with ecosystem services. It may be useful to formalize the relationship with these users by producing a memorandum of agreement or a formal
definition of user roles. Section 2.3.2 describes the process of organizing a governance structure that will include key stakeholders. Despite this formalization, the
process can remain open to all additional interested stakeholders. The role of each
stakeholder may be different, however, and will need to be defined (see section 2.1.3
for a list of possible roles of stakeholders). The resulting group of key stakeholders
who now have a formal relationship to the assessment will be those who are then
convened to formally launch the assessment process (see section 2.4.1 on convening
assessment participants).
Promoters of an assessment will have to be cautious about participants’ expectations. Public participation experiences tend to fail because of early unrealistic expectations that are not managed in a timely manner and are eventually not met.
The best way to manage expectations and keep them realistic is by making ongoing
efforts to be extremely clear and honest in terms of what an assessment can and
cannot achieve and about what is uncertain. In more practical terms, though this
exploratory stage will require bilateral meetings, it is recommended that multiactor
46 | A Manual for Assessment Practitioners
Box 2.4. User groups in the MA subglobal assessments
Almost all subglobal assessments involved national, regional, or local government agencies as
users. A large number of them identified local communities, NGOs, universities, and research
institutes as important users in addition to the government agencies. The private sector (for
example, the tourism industry, mining companies, and logging companies) was involved in
only five assessments, despite the MA goal to support a greater role for the private sector in
environmental decisions. This may be because the assessment teams did not have experience
in working with the private sector. One lesson from this experience is to dedicate some time or
team member to working on a relationship with important private-sector stakeholders.
Indigenous communities were involved in six assessments. Assessments that conducted
user needs assessment during the exploratory stage of the assessment found them to be very
valuable. Assessments also found it useful to have team members with good networking skills
in order to involve key decision makers who might make use of assessment findings.
meetings be carried out after a first round of meetings. These tend to be an easier
way to check unrealistic expectations, as all actors are confronted with one another’s
interests and positions. In addition, written agreements and terms of reference for
all activities should be produced, clearly establishing what the assessment process
does and does not intend to do.
2.3 How to organize an assessment process
Section’s take-home messages
• The first step in organizing an assessment process is to identify what resources,
particularly expertise and financing, are needed. The roles within an assessment team that
need to be filled include scientists, communication specialists, a facilitator, local experts
with contextual knowledge, social and political analysts, and a project coordinator.
• The organizational structure of an assessment includes a governance structure that will
keep the assessment work on track and ensure the process is credible, legitimate, and
relevant, along with work teams established around different themes or stages of the
• One way to organize a governance structure is to convene an advisory group—up to 20
representatives from diverse stakeholder groups. The assessment may need a separate
technical advisory group competent in specific subjects that are the focus of the assessment.
2.3.1 How to identify the resources needed for an assessment
Who to include in an assessment team
Promoters of an assessment are not necessarily the people best prepared to coordinate and carry out all the assessment work. An assessment process might require a
range of capacities not found within one organization. In this exploratory stage, one
crucial task is therefore to define the capacities that will be required and to identify
Stakeholder Participation, Governance, Communication, and Outreach | 47
the possible assessment team. Critical criteria to consider when consolidating the
assessment team include the following:
• Capacities to deal with the relevant assessment components (ecosystem goods
and services, human well-being, conditions and trends, scenarios, responses,
communication and participation).
• Credibility of the work team. Respected scientists and practitioners will add
weight to the assessment findings.
• Politically relevant organizations/individuals that may help ensure that the process and results are effectively considered and integrated into decision making.
• Local organizations/individuals as part of the team to ensure legitimacy and
relevance. Local experts (practitioners, leaders, or simply individuals with good
knowledge of the history of the area) can also provide contextual information
and knowledge that is critical to developing assessment findings.
• Ethnic/cultural balance, to be coherent with the reality in the assessment area.
For instance, in an area with a significant indigenous population, the project
team might include relevant indigenous organizations or individuals, either directly in the project team or indirectly in the assessment governance structure.
• Gender balance. The legitimacy of the process is strengthened by a balanced
gender ratio.
The assessment team is often a partnership of organizations from diverse sectors.
Organizations from the public and private sectors might increase the likelihood that
the results will be considered and implemented; scientific organizations (e.g., universities) might have access to relevant information and capacities; NGOs might have
networks and links to key users. A partnership of diverse organizations may make
the assessment process (and its results) more transparent and accountable and will
facilitate communication and outreach activities. On the other hand, partnerships
will add complexity to the governance structure. Therefore a clear and transparent
way of working is required, including a structure defining responsibilities and persons in charge.
In some cases, the capacities required for the assessment are not available in
the project area, and the assessment team may be brought in from a different area.
In these cases, care should be taken to discuss the reasons for this with local stakeholders and seek their support. And it will probably be necessary to integrate local
people with diverse capacities and knowledge into the assessment team in order to
contribute knowledge about local trends and relationships between ecosystem services and human well-being.
The concrete capacities required by an assessment team will vary according to
the ecosystem services being considered and the human well-being indicators of interest. In general, there will be a need for professionals from the natural and social
sciences. Ideally, the assessment team should also include people with expertise in
participation, communication, and outreach.
The following is a short list of roles to fill within an assessment team:
• Scientists and/or practitioners with expertise in measuring the condition and
trends in the ecosystem services of interest and relevant human well-being
48 | A Manual for Assessment Practitioners
• Communication and outreach specialists;
• An experienced facilitator for managing participatory processes;
• Local experts with contextual knowledge of trends in ecosystem services and
human well-being;
• Geographic Information System (GIS) specialists for mapping conditions and
• Social and political analysts with local knowledge and experience to work on
relevant responses;
• A project coordinator to manage the process; and
• Designated persons to manage the governance body and review process.
How to estimate the necessary funding
The MA experience shows that the provision of seed funding was an appropriate
mechanism for facilitating the generation of assessment processes; however, this type
of funding is not always available. In fact, MA sub-global assessments found it difficult to obtain outside funding, as assessments were not seen as activities generating
immediate “results.”
One important result of the exploratory stage is the estimation of the whole
project budget. Among the aspects that will define the extent of funds needed are the
spatial scale of the assessment; the size and nature of the technical effort (e.g., the
specific ecosystem services to be assessed and the way they will be studied); the size
and nature of the participatory, communication, and outreach processes; and the
availability of information and local capacities. Budgets will vary widely depending
on all these details and therefore it is impossible to offer more concrete guidance on
the size of budget needed. MA sub-global assessment budgets ranged from $15,000
to several million dollars (see Table 2.1). For many of the sub-global assessments,
in-kind contributions were a significant way to add needed resources.
Despite the fact that core funding might come from one specific source (e.g., international donor or a central government), it is highly desirable that local potential
donors—if available—be approached and invited to contribute. This will not only
add resources but, more important, might provide an opportunity to gain the trust
and commitment of relevant stakeholders. For instance, private companies that directly benefit from ecosystem services (e.g., forestry, mining, or fisheries) might be
good candidates to contribute project funding, provided the funds do not affect the
outcome or legitimacy of the process. Companies’ dependence and impact on ecosystem services can be assessed through tools such as the “Ecosystem Services Review” recently developed by the World Resources Institute. Assessment promoters
can approach local potential donors at the beginning of the process and learn about
their interests and needs. Their inclusion in a user needs assessment will facilitate
The participatory, governance, communications, and outreach activities within
the assessment might require a nonnegligible proportion of the total budget. In the
MA, project leaders often did not budget significantly for these kinds of activities,
which were considered to be “add-ons” at the end of the core work. Yet as already
mentioned, the usefulness of the assessment hinges on a successful and well-planned
participatory process, which thus needs to be budgeted from the beginning.
Stakeholder Participation, Governance, Communication, and Outreach | 49
Table 2.1. Budget for the San Pedro de Atacama Subglobal Assessment, Chile
Salaries and consultants
Meetings, workshop and
travel expenses
Basic information
Materials and products
GIS and satellite imaging
Tourism assessment
Water assessment
Administration costs
2.3.2 How to design a governance structure
The ultimate goal of setting up a governance structure for the assessment is to ensure
the relevance, credibility, and legitimacy of the process and findings, as described
earlier. This means ecosystem (and assessment) users must, to some extent, own the
process. If this is not achieved, the assessment will become an academic exercise with
little real impact.
One way to organize a governance structure is to convene an advisory group,
a set of representatives from diverse stakeholder groups associated with the assessment (see section 2.2.3). Advisory groups were used in a number of sub-global
assessments in different capacities. (See Box 2.5.) To be a manageable forum for discussion and decision making, ideally it should include a limited number of persons
(fewer than 20) who represent the community at large.
There are no fixed rules on how an advisory group should function. In terms
of power to make decisions, its role may range from being solely advisory (decisions are made by the project team) to having final decision-making responsibility (with decisions executed by the project team). In general, the objective of the
advisory group is to accompany the project development, providing information
about user needs and advice to the project team. This ensures the relevance of the
ultimate assessment outcomes. The advisory group is also a key mechanism through
which important stakeholders (some of whom are key decision makers) can learn
to trust the credibility of the assessment work and take ownership of the process
and outcomes. Building the relevance, credibility, and legitimacy of the process and
50 | A Manual for Assessment Practitioners
Box 2.5. User engagement and governance structure in the Southern African
The Southern African Millennium Ecosystem Assessment (SAfMA) was designed and implemented in a way that encouraged the participation of multiple stakeholders and users of the
assessment information. SAfMA set out to be user-driven, and stakeholders played an important role in its governance. At the regional scale, an Advisory Committee (AC) with 10 members
of different groups was responsible for representing the interests of the different stakeholders,
balancing the various interests within the region, and creating a receptive policy environment
for the assessment’s outputs. The AC directed the work of the teams conducting the assessment and endorsed SAfMA outputs at each stage of the process. At the other scales of assessment, User Advisory Groups played this role. In this way, users had ownership of the process,
and they endorsed and signed off on outputs that they considered to be credible. The AC also
ensured that the different assessment teams were adhering to agreed-upon schedules and
timelines, as this was imperative for the integration of the assessment findings across scales.
The AC played a role as well in steering the project through difficult phases and provided leadership and guidance.
Due to its multiscale nature, SAfMA stakeholders were varied. The different categories
of SAfMA users were engaged in a variety of ways. In addition to their being in the AC and
User Advisory Groups, other stakeholders were appointed to review panels and were involved
in intensive meetings and workshops at the various scales. Stakeholders were also engaged
through a SAfMA Fellowship Programme, where individuals from stakeholder organizations
were invited to become SAfMA Fellows. This involved participating in SAfMA activities, reviewing SAfMA documents, and assisting with outreach and dissemination of SAfMA materials.
SAfMA Fellows also acted as bridges between SAfMA and other programs in the region.
Users expressed their needs in the meetings and workshops held. Prior to the start of the
assessment, the need for information had been stressed at numerous national workshops, in
various State of the Environment Reports, and in Strategy Documents of the Southern African
Development Community (a regional grouping of Southern African countries). The needs of users were also ascertained through direct consultation in workshops and meetings and through
the participation of user groups in the review of various reports and documents.
outcomes ensures that the assessment is considered and integrated into targeted decision-making processes.
The role of the advisory group can be both political (as described above) and
technical. The assessment process may require a technical advisory group competent in specific subjects that are the focus of the assessment. Such a group can act
as a sounding board to deal with complex and contentious issues. The social and
ecological complexity of the assessment questions, as well as the spatial scale of the
assessment, will determine the need for a more or less technical advisory group. In
either case, the advisory group will play a role in managing the review, ensuring a
balanced and fair assessment process, and making decisions on how to present contentious results.
Once a decision has been made in terms of setting up an advisory group, it needs
to be formally implemented. Ideally, during the user needs assessment (see section
Stakeholder Participation, Governance, Communication, and Outreach | 51
2.2.3) promoters will have already identified key organizations and individuals that
could be part of an advisory group. The rights and responsibilities of the advisory
group, as well as the working style, are usually defined in writing as terms of reference (ToR). The very first meeting of the advisory group should be devoted to discussing and approving the ToR. The ToR can define:
• Whether members are invited as at-large members (e.g., as individuals) or as
institutional representatives;
• How final decisions will be negotiated and who will have the final say;
• How conflicts will be resolved;
• Whether there is financial compensation involved for advisory group work;
• How many meetings must be attended; and
• The structure of the advisory group.
The group’s structure will depend on the size and the scope of the assessment. Larger
assessments may require members to assume oversight of particular aspects of the
assessment process, such as the budget, the review process, or the communication
process. Advisory groups will almost always require a chairperson.
The decision to include political leaders in the advisory group is not straightforward. Political leaders can raise the political profile of the assessment, but they can
also subvert the credibility and legitimacy of the process in the eyes of some stakeholders. It may be more effective to involve technical representatives of the political
leaders, ensuring that the leaders are informed about the assessment activities and
invited to participate in some of them.
2.3.3 How to organize the work team
During the exploratory stage, assessment team members are proposed and discussed
(see section 2.3.1). Once the work team has been assembled, dividing the assessment
work and formalizing teams responsible for different components of the assessment
may be necessary, depending on the size and scope of the assessment. In the global
MA, researchers were divided into three working groups focused on condition and
trends of ecosystem services, scenarios, and responses. A fourth working group
worked on assembling the sub-global assessments. Sub-global assessments were not
usually divided along these lines, and their full assessment teams usually worked
together on condition and trends, scenarios, and responses in order to facilitate the
integration of these components of the assessment.
Assessment work can be divided by scale (in the case of multiscale assessments),
by ecosystem service, or by focal question. In all cases it is important that the members of the assessment team meet frequently in order to ensure that all components
of the assessment can eventually be integrated. A lesson learned from the MA subglobal assessments is that it is difficult to integrate findings at the end, and therefore
it might be useful to designate a team member to focus on integration issues and to
meet regularly to discuss integration of findings. Other roles to be designated within
the work team are listed in section 2.3.1. The most effective way to manage the
work of the assessment team is described in the next section.
52 | A Manual for Assessment Practitioners
2.4 How to manage the assessment process
Section’s take-home messages
• The way participants are convened is a first opportunity to build the legitimacy and
credibility of the process, but also a way to undermine it. The convening process is also
an early opportunity to start communicating about the assessment.
• At the first meetings, participants need to make some fundamental decisions: What are
the precise goals of the assessment? What conceptual approach will be used? What rules
will govern decision making?
• A clear workplan will help minimize problems and address issues that may arise during an
assessment. Management issues that should be considered include periodic meetings
of the technical team and any governing body, stakeholder consultations, conflict
management, sources of information, and processes for peer review.
• Capacity-building activities are an integral component of any assessment. The focus of
capacity building can differ, depending on identified needs. And the capacities of both
those conducting the assessment and those using it might require development.
2.4.1 Convening assessment participants
As noted, assessments may be undertaken upon the initiative of a group of individuals or organizations or even by legal mandate (as State of the Environment Reports
are in many countries). However, as mentioned previously, there is always an actor
with initiative or leadership who starts identifying the members of the governance
structure and the technical teams and brings them together.
Convening is not as simple as it sounds, even if the assessment is undertaken
by mandate of a higher authority. The way participants are convened is a first opportunity to build the legitimacy and credibility of the process, but also a way to
undermine it. Some important aspects to consider when convening are:
• Leadership—Conveners need to be respected by peers.
• Formality—Participants must perceive that the process they are being invited to is
appropriately managed. The exploratory stage of the assessment may be iterative
and flexible, but once the assessment is formally launched, a degree of formality is
a strategic way to convince stakeholders of the importance of the work.
• Representation—In the case of the governing structure, participants must ideally
represent the major stakeholders.
• Transparency—The criteria and mechanism for the selection of participants
must be clear to all and avoid arbitrary decisions.
A good way for proponents to proceed is to engage first those who will be the leading figures of the governance and technical structures and then to work with them in
the gradual identification of the other members. That is, if the governance structure
involves an advisory group, a leading figure might be invited to chair it and participate in identifying the other members, preferably with the aid of a completed user
needs assessment (see section 2.2.3). As nominations proceed, for reasons of balance
or the need to engage a key stakeholder, a co-chair may be appointed.
Stakeholder Participation, Governance, Communication, and Outreach | 53
The convening process is also an early opportunity to start communicating
about the assessment. Ideally the process will be perceived as of such high quality
and with such potential real impact that stakeholders and experts will find it attractive to participate.
The question of who gets involved in an assessment is a difficult one to deal
with, as interest should be the most important factor deciding participation. Genuine interest is what is likely to sustain participation and determine its quality. Representatives of different user groups should be invited to participate, but whether they
do decide to participate or not will depend on other issues. Incentives can facilitate
the participation process and can help to keep people engaged. The type of incentive
needs to be considered carefully in order not to raise expectations unrealistically
and to ensure that the right people are attracted. The sustainability of the incentive
system also has to be considered.
2.4.2 Early governance decisions: goals, approach, and rules of the process
Once participants have accepted the invitation to participate and their roles have
been defined (Chairs of the governance structure, at-large or institutional members
of the advisory group, lead scientist on the technical team, etc.), the group needs to
make its first fundamental decisions: What are the precise goals of the assessment?
What conceptual approach will be used? What rules will govern decision making?
The assessment must have a clearly stated purpose that will guide the whole effort,
from information generation to communication and engagement activities. Because
the process involves people with very different backgrounds and understandings, it
is important that participants take sufficient time to discuss and define what they are
trying to achieve. The goals of various MA sub-global assessments have included:
Informing a development plan for one region in a country;
Building a rationale for local management of ecosystem services and landscapes;
Improving national management of ecosystem services;
Building local appreciation for ecosystem services;
Informing a fisheries management plan;
Understanding trade-offs between agricultural development and other ecosystem
services; and
• Developing baseline data on ecosystem services and their relation to human
Any focused goal is acceptable, but ideally a mechanism should be in place for actors and decision makers to use assessment information to improve the management
of ecosystem services. The identification of policies, initiatives, and projects that the
assessment can feed into can help in this regard (see section 2.2.3).
Assessment approach and the conceptual framework
Equally important as the goal is a common understanding of the conceptual and
methodological approach of the assessment. Not only does this determine the
54 | A Manual for Assessment Practitioners
assessment work, it also facilitates the task of communicating the goal and rationale
of the initiative to its diverse users. During the exploratory stage and when convening participants, promoters will have presented and discussed the rationale for an
assessment exercise. They would have offered the framework as a robust and desirable way to approach the relationship between ecosystems and human well-being.
Clearly, however, there might be other views—for instance, indigenous people’s approaches to nature and ecosystem services—that might call for changes, either subtle
or substantial, to the approach. So the whole approach needs to be discussed and
validated by the governance and technical structures. Chapter 3 explains the benefits
and approaches to participatory processes aimed at building a common understanding of a particular system through the use of a conceptual framework.
Rules of the process
To ensure credibility and legitimacy, it is important to have clear rules about who
will decide things like how information will be generated (e.g., what sources are acceptable or how reviews will proceed), how the reports will be structured, what language to use, what the appropriate communication strategy is, how to set priorities
on the use of resources, and when the reports will be officially considered final. It is
up to members of the governance structure to decide this as early on in the process
as possible and to develop documents that outline these rules. Formalizing the rules
will help the assessment team stay on track.
Establishing process rules builds legitimacy because they promote transparency
and mechanisms of accountability in the process and because participants will then
also be responsible for the correct implementation of the assessment. Credibility, in
turn, results from the determination of adequate standards for quality of the scientific or technical work. (See Box 2.6.)
2.4.3 Governance and management issues during the assessment
Being a social process, an infinite number of issues and problems will emerge in the
course of any assessment. Some elements can be built into the design of the assessment, however, that will help minimize problems and address the main issues that
may arise.
A clear workplan
To be able to monitor progress, it is useful to have a workplan with clearly defined
timelines and milestones. (Figure 2.1 presents the MA workplan and timeline.) This
is especially important in assessments with several components, where it is necessary
to integrate the work of different teams into a single product. Integration is better
undertaken as the assessment progresses instead of waiting until the end. The experience of the Southern African Millennium Ecosystem Assessment was that integration
needs to be planned from the outset of the assessment, as it is difficult to achieve
afterward. In the case of SAfMA, regular meetings of the assessment teams and the
Advisory Committee were held to review progress against agreed timelines and to
address problems and keep the assessment on course.
Box 2.6. Data quality assurance in the MA
Quality assurance of data is needed in any assessment, and there are different ways of achieving this. The MA adopted the following rules to assure data quality:
• Most data used or cited must be from peer-reviewed scientific publications.
• Most data sets used are from large national or international organizations that have internal procedures for maintaining quality control.
• Datasets developed by the MA will aim at a high level of quality control and archive all
• Archiving of metadata is designed to help assure the quality of information coming from
traditional knowledge and undocumented experience.
• Local and traditional knowledge needs to be critically assessed (e.g., cross-checked or
triangulated) before being used.
These rules were applied at the global scale in the MA but were more challenging to apply at
subglobal scales where fewer published data were available. However, extremely successful
uses of local and traditional knowledge produced comprehensive and credible data, collected
through a variety of rigorous approaches. In Peru, communities developed data on conditions
and trends in soil and water resources by consensus. In communities in South Africa, researchers and community members used Participatory Rapid Appraisal techniques to develop and
validate data on several ecosystem services. Chapter 4 presents different approaches for validating information from a variety of sources.
Source: Georgina Cundill, SAfMA.
Figure 2.1. Schematic presentation of MA workplan.
56 | A Manual for Assessment Practitioners
Interaction between the technical team and the governance structure
Throughout the process it is critical to have a fluid and positive interaction between
those gathering the information and putting together the reports and the group of
people guiding the process, who will ultimately “approve” the work. Periodic meetings between the technical team and the advisory group or governing body of the
assessment serve various purposes, mostly related to the relevance and legitimacy of
the process, such as:
• Checking on the fit between the policy questions posed in the assessment and
the technical answers provided;
• Checking that the format of the assessment is useful and friendly for the target
• Checking that the language being used is appropriate for targeted users;
• Identifying the main findings that need to be communicated, and defining how
they should be stated; and
• Managing the expectations of stakeholders.
Stakeholder consultations
Consulting with a broader range of stakeholders than those represented in the governing bodies is a good way of validating the latter’s decisions, obtaining further
input on the technical work, and disseminating findings in advance of the final products. Formal stakeholder consultations occur during the review process, but other
workshops and meetings give stakeholders the opportunity to ask questions and
voice any concerns they may have.
Consultations can be made at every stage of the assessment—from the development of the conceptual framework to different parts of the technical teams’ work.
Some parts of the assessment, however, will be more malleable to stakeholder consultation than others, and this should be explained in advance. Scenario development is a particularly useful part of the assessment to engage stakeholders. Chapter
5 presents the stakeholder engagement process in the context of scenarios.
Review as an engagement process
Validation through a review process is key to ensuring the quality, and thus the credibility, of an assessment. Peer review is a standard way of approving the quality of
information in the scientific community. However, the review process should not be
restricted to only scientists. The involvement of different users in the review process
is desirable as it can provide a much broader range of comments, form part of the
communication strategy, and contribute to ongoing user engagement in the process.
It is important to note that if comments on drafts are requested from stakeholders,
these need to be explicitly addressed by authors whether they accept them or not.
It is damaging to the process to invite comments and then disregard participants’
The peer review process needs to be transparent and should be agreed to by the
governing body. Review need not focus only on complete reports, as even the way
the assessment is progressing and the interim products can be reviewed. In SAfMA,
Stakeholder Participation, Governance, Communication, and Outreach | 57
the Advisory Committee reviewed work in progress in addition to completed draft
Conflict management
As in any participatory process, conflicts of interest among stakeholders are very
likely to come up, and the advisory group should be well positioned to resolve them.
Plenty of tools for conflict management, constructive negotiation, and facilitation
can be applied, but it will always be critical that group meetings are adequately organized and chaired. Some ways to minimize conflict during the process are to:
Establish by consensus clear, but flexible, rules of participation;
Have an agenda and clear objectives for each meeting that is convened;
Promote communication among members in between meetings; and
If the governing body is a large one, create a committee to deal with operative
issues between meetings.
2.4.4 Issues in the technical process: sources of information, review, and
capacity building
Identifying sources of information
Chapters 4, 5, and 6 point to sources of information for assessing conditions and
trends of ecosystem services and human well-being, developing scenarios, and assessing responses. In many situations, particularly in developing countries or in remote locations, information on ecosystem goods and services (beyond provisioning
services) is scarce. The exploratory stage of the assessment will usually reveal this
condition early on. Depending on the resources, capabilities, and time frame, decisions will have to be made about the need for generating new data. If this is not possible, the advisory group and the project team will have to devise ways for dealing
with the lack of data. The systematic use of experts’ judgment might be an appropriate way to generate relevant information.
In addition to science, MA sub-global assessments have favored, to the extent
possible, the use of local or traditional knowledge and practitioners’ knowledge (a
comprehensive definition of systems of knowledge is found in MA 2005 and Reid
et al. 2006). This, however, might not be immediately accessible and will probably
require a dedicated effort to build trust and commitment to the project by the holders of local knowledge. Another challenge will be the eventual integration of local
knowledge with published information.
Peer review as a technical process to ensure credibility
As mentioned in previous sections, peer review of assessment findings is a key mechanism for building the relevance, credibility, and legitimacy of the assessment. The
credibility of the findings improves with each additional review, especially if the pool
of reviewers has a diversity of expert perspectives. The legitimacy of the findings
improves as stakeholders are given the opportunity to comment on the findings and
take part in the process. The relevance of the assessment is improved as stakeholders
58 | A Manual for Assessment Practitioners
can comment on whether their specific information needs are being met. The review process must therefore be conducted in a transparent and professional manner.
This can occur anywhere from once to several times during the assessment process.
Each additional review helps the process and outcomes, but it requires additional
resources and time.
It is common for advisory groups and assessment teams to develop a list of
potential reviewers well before the review process occurs. These individuals can be
chosen on the basis of their technical expertise, decision-making capacity, or political stake in the region and its ecosystem services. Reviewers can be identified
by the experts writing the reports, by the Board, by assessment proponents, or by
stakeholders. Reviewers thus often include local people and international experts.
Advisory groups associated with MA sub-global assessments were very useful in the
review process, as they generally included representatives from many user groups.
Review processes require advance warning for reviewers in order to prepare
them for the relatively short windows of time in which they have to read assessment
drafts and submit comments. Reviewers should be sent invitations to review either
specific assessment chapters (if the assessment is long) or relevant sections of full
reports, to prevent overburdening individual reviewers and to maximize the chance
of getting a response. These invitations could be sent out several months before the
review process begins and be followed up with correspondence when necessary. The
dates of the review should be listed from the beginning. When the review period approaches, reviewers are sent a form that they will complete with review comments.
(See Figure 2.2 for an example of such a form.)
Review comments can then be sent to the appropriate assessment team members responsible for the specific sections referred to by each comment. Each review
comment needs to be acknowledged, either by adjusting the assessment text (if the
author agrees with the comment) or by explaining in writing why no change has
been made. The transparency of the process is improved if the review comments and
author responses are made public.
Capacity building
Capacity-building activities are an integral component of any assessment, but especially a complex one in the style of the MA. In the MA, capacity building served
to overcome a variety of constraints faced by a number of sub-global assessments.
Many teams did not have the capacity to fully address the linkages between ecosystem services, human well-being, and drivers of change, were challenged in the
development of scenarios, and were not able to work at multiple scales. The focus
of capacity building can differ, depending on identified needs. The capacities of both
those conducting the assessment and those using it might require development.
Different approaches can be used to build the capacity of assessment teams. In
the MA, sub-global assessment teams met at MA sub-global working group meetings to share methodologies and lessons learned. Regular workshops can be held at
the level of an assessment to build capacity. Invited experts can hold seminars on
how to conduct specific technical work. The involvement of students and young
researchers at the beginning of their careers is another way of building capacity.
The ability of different institutions and interest groups to use assessment findings may be limited in some cases. In the Bajo Chirripo assessment, for instance, a
Stakeholder Participation, Governance, Communication, and Outreach | 59
Figure 2.2. Example of a form to request peer reviews.
local NGO attempted to develop resource management plans with local communities based on assessment findings, but the communities did not have established
institutions to implement the findings effectively (MA 2005). In some cases, cooperation is required among regional institutions in order to implement the findings of
an assessment. Building capacity to use findings can thus range from giving workshops on the conceptual framework or potential response mechanisms to convening
different user groups to encourage partnerships that will work together to make
use of findings. The capacity of users to apply the findings of an assessment will be
enhanced through their sustained involvement in the assessment and their participation in activities to discuss and analyze issues central to it.
2.5 How to bring the assessment to a close
Section’s take-home messages
• The assessment can be considered finished once it has undergone a sufficient number
of peer and stakeholder reviews and has been approved by the technical team or by the
leading authors.
• The final discussion to approve the assessment findings is extremely important for the
assessment’s legitimacy. This is also a final opportunity for any stakeholders who see their
interests threatened by the assessment findings to voice their opposition.
60 | A Manual for Assessment Practitioners
An assessment can be expected to raise more questions than answers. There will
always be a need for further research and analysis. However, at some point both the
technical team and the stakeholders involved will have to express their satisfaction,
or lack thereof, with the “final” work. So when can an assessment be considered
From the perspective of the technical work, the assessment can be considered
finished once it has:
• Undergone a sufficient number of reviews (both peer reviews and stakeholder
reviews)—Determining how many reviews are sufficient depends on each
particular assessment, but normally at least two rounds are advisable.
• Been approved by the technical team as a whole or by the leading authors—
Open, technical scrutiny by everyone of everyone else’s work will help the
quality of the work and allow team leaders to sign off the final document.
Once these two conditions, which basically consolidate the credibility of the work,
have been met, the technical team can present the final document for discussion and
approval by the governing body of the assessment, where stakeholders will decide
whether the work is officially approved or not.
The final discussion to approve the assessment findings is extremely important for the assessment’s legitimacy. If interaction between the technical team and
the governing structure has been fluid throughout the process, this final discussion
should not present major problems. But this is also a final opportunity for any stakeholders who see their interests threatened by the assessment and its findings to voice
their opposition to it.
Normally, the documents of most concern for stakeholders will be those that
will be widely distributed and used for communication: the summaries for decision
makers or the thematic syntheses. There is wide scope for interpretation between the
way findings are produced by scientists and a set of conclusions that can be effectively communicated to stakeholders and the wider public. Different scientific findings can be emphasized in different ways, spins can be put on statements, narratives
can be constructed by collecting findings from different sections of the assessment,
and so on. The discussion is thus a critically important one and must proceed with
the active participation of the technical team, who must tread a thin line between
preserving the “scientific truth” of the findings and yielding to the political priorities
of users.
This has been the case with the Intergovernmental Panel on Climate Change,
whose results are debated by an assembly of government representatives. Governments approve the work of IPCC, but they focus long and hard discussions on the
summaries for decision makers, the contents of which are scrutinized line by line
and often dramatically altered from what the technical team originally proposed.
With the formal approval of the documents by the governing bodies, the assessment findings are now ready to be widely disseminated.
Stakeholder Participation, Governance, Communication, and Outreach | 61
2.6 How to communicate assessment findings
Section’s take-home messages
• Effective dissemination of results needs to be guided by clear communication goals. To
develop a strategic communication plan, it is important to know the target audience or
audiences well. One way to do that is to identify all the people and organizations in a
position to influence the types of interventions that have been identified as desirable in the
• Deciding on the key messages of an assessment is one of the most important steps of
the communication process—messages that are a strategic culling of the points most
relevant to each audience, presented in a way that promotes the credibility of the findings.
Experience has shown that assessments may lose legitimacy and credibility in the eyes of
some audiences if they go beyond presenting objective scientific findings into suggesting
how policy makers should do their jobs.
• The strategy and extensiveness of a communication program will depend on the
assessment’s budget. The formats for presenting findings include early release of
products such as the conceptual framework, final reports tailored for specific audiences,
summaries for decision makers, Web sites and other online resources, workshops and
meetings, and coverage in various media.
2.6.1 Defining a communication goal and knowing the audience
An effective dissemination of results needs to be guided by clear communication
goals. This is the goal that will define the specific target audiences, which will determine the appropriate means of dissemination.
The dissemination or communication goal should be subservient to the general
goal of the assessment, and dissemination activities should all support the purpose
of the whole assessment. It is important to keep this in mind because it helps discriminate between the broad range of good communication ideas to focus resources
on those that are most specifically conducive to the substantive goal.
In the case of the MA, the purpose of the assessment was to establish the scientific basis for actions needed for the conservation and sustainable use of ecosystems
for human well-being. To help achieve this, the dissemination strategy focused on
two goals: ensuring that stakeholders were adequately engaged through appropriate access to the assessment process and its products and creating a demand for the
assessment reports and for technical expertise to conduct sub-global assessments. It
was understood that these two dissemination goals were key for achieving the main
goal of the assessment.
In order to develop a strategic communication plan, it is necessary to know
the target audience well. The audience will include the stakeholders that have been
involved in shaping the assessment from the beginning (see section 2.2.3), but potentially many other groups, such as international organizations, businesses, and
public officials in a diversity of sectors, as well as the larger public. In the MA
sub-global assessments, the diversity of audiences was often defined by the scale of
62 | A Manual for Assessment Practitioners
the assessment, with larger regional assessments having a higher diversity of audiences and a broader communication strategy. One way to determine the target audiences is to identify all the people and organizations in a position to influence the
types of interventions that have been identified as desirable in the assessment and
then to reduce this list based on what is a feasible strategy under time and budget
Common audiences for assessment information include:
Governments (various levels and various departments)
Researchers and analysts
Nongovernmental organizations
General public
Schools and universities
Industries and businesses
Women’s groups
Indigenous peoples’ groups
Target audiences are defined by their profession and areas of focus, which will influence the content and style of the materials used to reach them. Audiences are also
defined by differences in language and culture. This may result in increased costs of
printing materials in multiple languages or the need to have several strategies for the
dissemination of assessment findings using very different media. In some areas, for
example, newspaper articles may be effective for disseminating information to the
general public, while in other areas the radio or television may work better. In still
other cases, personalized approaches and targeted products are required.
While putting together a list of desired audiences for the assessment information, it helps to take notes on what kind of information might be most relevant or
useful to different audiences, the perceptions of each group on the issues included
in the assessment, and the type of communication method has been used to reach
each group in the past. It is also useful to distinguish between potential end users of
assessment information, who make decisions based on the information (e.g., adopt a
law or not, buy or not buy), and “broadcasters,” who recycle information for their
own communication goals and thus multiply its impact (e.g., the mass media, the
educational system, and many NGOs).
2.6.2 Developing the content and style of reports
Deciding on the key messages of an assessment is one of the most important steps
of the communication process. Full assessment reports are useful reference documents and will contain all the information produced during the assessment. But
these documents will rarely be used to disseminate information to target audiences.
At this point, the content and conclusions must be synthesized into short and specific messages that will resonate with the audience. The main messages are usually
not simply a summary of all the information produced by an assessment but rather
a more strategic culling of the points most relevant to each audience, presented in a
Stakeholder Participation, Governance, Communication, and Outreach | 63
way that promotes the credibility of the findings. This means backing up important
statements with data and examples and using easy-to-understand graphs, illustrations, and tables. Graphic figures can be very powerful tools for conveying complex
information in a way that is understandable and memorable.
The MA took care that outputs developed for communication were, like the
assessment itself, relevant to policy makers but did not tell them what to do. In
other words, main messages were policy relevant but not policy prescriptive. This
is generally because assessments may lose legitimacy and credibility in the eyes of
some audiences if they go beyond presenting objective scientific findings into suggesting how policy makers should do their jobs. Assessments can still affect policy
by making sure that the information that is most relevant to the choices being faced
by audiences is included in the communication products.
The style in which the main findings are communicated depends on the audience. Box 2.7 gives some examples of how to match the style of reporting to a
specific audience. Section 2.6.3 outlines the different formats that can be used to
communicate findings. But within each format, attention to the style of presenting
information is key to reaching the target audience.
Acknowledging uncertainty is also a strategic part of putting together messages
that preserve the credibility of the work. The assessment team must decide whether
to include information that is uncertain. If it is included, the associated level of
uncertainty needs to be clearly stated. As described in Chapter 1, the MA assigned
certainty levels to findings based on the collective judgment of the authors, who
used observational evidence, modeling results, and theory to decide which level of
certainty applied. (See Box 1.3 in Chapter 1 for further details.)
2.6.3 Communication formats
The strategy and extensiveness of a communication program will depend on the assessment’s budget. With a clear idea of the communication goal, the target audiences,
their information needs, and the available budget and communication expertise, the
next step is to decide on the format of communication. This section describes several
common formats, but many others may be suitable.
Box 2.7. Target groups and report style
• Decision makers. Content should be short, specific, fact based, and consist of the latest
• Media. Content should be short and consist of findings relevant for broad audiences, with
messages that can easily be linked to other issues in the news. There is a better chance of
media coverage if there are supporting visuals such as graphs or photographs.
• Students. Content should be well explained, and the language should be simple.
• Scientists. Content should be fact based and rely on the latest data. The language can be
scientific and include technical terms.
Source: UNEP 2007.
64 | A Manual for Assessment Practitioners
Tailored reports
To ensure that as many people as possible could obtain access to the information, the
MA produced reports tailored to different audiences. This approach involves writing
targeted publications that focus on the most relevant information for specific groups
and sometimes translating publications into multiple languages. In addition to the
full technical assessment reports, the MA produced six synthesis reports aimed at
different users. Information from the main assessment volumes was summarized and
repackaged in short, carefully designed volumes dealing specifically with biodiversity, desertification, wetlands, health, and business and industry, in addition to an
overall synthesis directed at a more general audience. Content, language, and style
were modified to suit each audience group.
When designing a report for a particular audience, it helps to ask the following
• How long a document will a person in this position typically read? Often a
decision maker will not read more than 1–5 pages, while their advisors might read
10–30 pages. Only scientists and practitioners will read longer, technical reports.
• Should the content be written in technical, popular, or formal policy language?
The document must be both appealing and easy to understand for the intended
• What kinds of issues or decisions are facing this audience? Is there a specific issue
or decision that the report can address explicitly? The report will ideally not offer
guidance on a decision but rather will supply targeted, relevant information.
• What information from the assessment needs to be included? In some cases, only
a small segment of the entire report will need to be presented to a particular
• What figures and formatting will be effective in communicating to this
audience? The presentation of the information, including clear figures, graphs,
and drawings, and the layout of the document will make a big difference in how
different audiences take up the information. It is often helpful to look over other
reports and discuss with users what they find effective and appealing.
Depending on the scale and scope of the assessment, the reports may need to be
translated into several languages. Although in the end almost all the synthesis reports were translated into the five U.N. languages, this proved to be one of the main
dissemination difficulties in the MA, with translations taking longer than desired
and requiring multiple reviews to ensure quality. Leaving certain language groups
out by not translating reports can politically undermine the dissemination process.
Translation of reports is an opportunity to engage more stakeholders in the process,
such as universities and NGOs, which can then also assist very effectively in dissemination activities. Hence it would be useful to begin engagement arrangements for
translation early on in the process.
Summaries for decision makers
A category of tailored report is the summary for decision makers (SDM or SPM
[summary for policy makers]). This is usually a very short document (one to several
Stakeholder Participation, Governance, Communication, and Outreach | 65
pages) that highlights the key messages in one or two sentences each. It does not
provide all the evidence behind the findings, but there should be a clear indication of
where further information can be found. When messages are boiled down to this degree, there is a tendency to make very general statements that may not relate directly
to the policy action agenda of the targeted decision maker. Going over the main messages with decision makers is one way to come up with precise, strong messages that
can be acted upon. But even when only general statements can be made, assessments
can be very significant sources that help policy makers reinforce an argument, confirm a widely held belief, or contest it. When the MA, in its summaries for decision
makers, said that “over the past 50 years, humans have changed ecosystems more
rapidly and extensively than in any comparable period of time in human history,
largely to meet rapidly growing demands for food, fresh water, timber, fiber, and
fuel,” it was not revealing something that many people did not intuitively know. But
the weight of this statement expressed by hundreds of the best scientists in the world
turned it into a powerful communication instrument.
Electronic communications
Establishing electronic communication mechanisms is important, as a growing number of people find and share information on the Internet. Electronic communication
mechanisms may include a Web site, a system to share data, and an intranet system for internal communications among the assessment team members. Electronic
mechanisms are becoming easier to set up, even without programming expertise,
although a well-designed Web site and intranet often require the input of a specialist. Assessments with larger budgets may choose to hire a Web designer to produce a
professional, easy-to-navigate site where all the assessment products will eventually
be located. Smaller projects can take advantage of free online resources, including
preformatted Web sites, blog spaces, and wikis, among others. Local universities are
a good resource for locating ideas, expertise, and available Internet sites.
Workshops and meetings
Explaining the findings of an assessment in person is a powerful way of disseminating information. Workshops and meetings provide audiences with the opportunity
to ask questions and understand the findings more deeply than they would through
reading a report. However, the trade-off is that these meetings are costly and only
reach a small number of people. At smaller scales, this may be the most effective
communication strategy and does not have to be too costly. For larger-scale assessments, the most relevant decision makers and stakeholders can be invited to a
workshop to discuss assessment findings, while other methods may be used to disseminate information to the broader public. Workshops can also be used to build
capacity among different stakeholders to communicate the principal messages to
other audiences.
Nontraditional communication methods
There may be an opportunity to use less traditional means of communication—theater, art, calendars, or video, for example—to capture an audience and communicate
66 | A Manual for Assessment Practitioners
important messages to them in imaginative ways. Some care should be taken to
ensure that the credibility of the findings is not compromised by the method of communication, as some audiences will be more receptive to nontraditional methods of
communication than others. In some cases, these methods will be more appropriate
or effective than reports. In addition to reports and summaries, the MA sub-global
assessments produced brochures and pamphlets, atlases, posters, calendars, theater
pieces, and videos. For example, the local assessment in Vilcanota, Peru, trained
community members to produce a video to disseminate findings to local communities. Video was considered to be culturally appropriate in this context, as it is a form
of communication that is similar to the local tradition of storytelling and visual
representation of environment and culture.
The media plays an important role in disseminating assessment findings to the general public. Assessment teams can prepare press releases and make the media aware
of people who are available to answer questions. The MA was a high-profile international assessment and wanted to achieve a certain level of press attention when the
findings were finally released. Three approaches were used to accomplish this:
• Organizing seminars for the media while the assessment was being conducted to
explain what it was, why it was being done, and what to expect from it;
• Establishing a loose working group with the media officers of partner organizations; and
• On the day the MA was released, organizing press briefings and seminars in 13
cities around the world, which ensured that appropriate angles and languages
were used to draw national media attention.
The MA also posted a list of contributing authors around the world who were available for media interviews. This allowed local press to contact authors in their areas
who could link the MA findings to local issues.
2.6.4 Conveying relevance, credibility, and legitimacy to the audience
The relevance of the final assessment products will depend on how well the stakeholder process was set up from the beginning. However, having different communication products tailored to different audiences will help maintain the relevance of
the assessment to those audiences, and they will appreciate being able to read only
the information they would be most interested in.
There are many approaches for branding assessment products as credible and
• Have enough participation and buy in from well-known organizations to put
their logo on the assessment products;
• Have highly respected scientists, politicians, or public figures introduce the assessment to the media and general public;
• Invite respected and well-connected people to join the advisory group and help
communicate the assessment findings; and
Stakeholder Participation, Governance, Communication, and Outreach | 67
• Organize a thorough review process and include the number of reviewers and
review comments on communication products.
2.6.5 Strategies to leverage communication
In many of the MA sub-global assessments (and also at the global level), specific
individuals played key roles during different stages of the assessment—for example,
as external facilitators in determining the demand for the assessment, in providing
leadership and sustaining the process, and in communicating findings. In some cases
this was a member of the assessment technical team; in other cases it was someone
who played an advisory role. Many sub-global assessments found advisory group
members to be a powerful means for communicating the findings to a diversity of
audiences. Some advisory group members might therefore be chosen specifically for
this purpose and be high-profile, respected individuals within a particular context.
Champions can also be individuals who are highly respected and considered to
be neutral within politically charged or conflicted contexts. In these cases, the individual is usually not associated with the assessment but can help build trust among
audiences in order to communicate the findings effectively.
Partner institutions
The engagement and outreach team of the MA saw its communication activities
as an instrument not just to reach out and convey an image of the MA but also to
enhance the ownership of the MA and improve the ability of third parties to understand and make better use of it. Hence, the MA sought to rely on as many partners
as possible for outreach and to encourage many third parties to undertake outreach
for the MA on their own. This resulted in several instances where volunteers approached the MA to undertake activities, which was highly beneficial in dealing with
media enquiries.
The MA found that partnering required two important elements: a minimum
level of coordination in terms of setting key dates and sharing basic strategies for
communicating assessment findings, plus the generation of materials to support
outreach by third parties. These materials were shared not only through the MA
intranet but also through an “outreach kit” distributed on CD. This contained a
collection of elements developed by the MA, including:
Guidance on how to explain the MA to the uninitiated;
Guidance on how to develop a communications strategy;
Graphic elements (posters, maps, logos, photographs, videos); and
PowerPoint slides.
While all assessments might not be able to produce as complete a kit, the same types
of products and guidance can be shared through a simple document or developed in
a workshop with the partner institutions.
68 | A Manual for Assessment Practitioners
Sustained interaction with audiences
Keeping an assessment visible for its potential users is key to building up enthusiasm
for the findings. Often assessments are launched with the participation of stakeholders and then the whole process takes up to several years to complete. For example,
the global MA was a four-year endeavor. Even after being approached and consulted
at the inception of the assessment, targeted users needed to be kept updated and reminded of upcoming work. These activities were also meant to build momentum and
expectations; they involved multiple briefings and smaller meetings in international
and national arenas. Briefing audiences on the progress of the assessment is a simple
way to maintain a positive relationship with audiences.
Early products
One way to increase the demand for assessment findings is to release some products early. The global MA did not wait until the end to start releasing outputs, for
example. In particular, releasing the conceptual framework and early findings on
sub-global assessments permitted better outreach during the process. The Portugal
sub-global assessment released a User Needs report near the beginning of its process
that showed how the assessment would meet the information needs of important national stakeholders. Early products can pique the interest of audiences and give them
a concrete example of what can be expected at the end of the assessment.
The assessment provides diverse opportunities for generating and disseminating relevant information. Most of the assessment elements, such as conditions and
trends on ecosystems and human well-being, scenarios, and responses, might be
valuable stand-alone products for users. A timely dissemination—for instance, first
as drafts for comments and then as final products—will secure the interest of users
and increase their constructive involvement. If the assessment team includes the dissemination of intermediary products in their workplan (e.g., a status report), complying with this plan will be seen as a positive signal of the project’s success.
Agrawala, S. 1998. Structural and process history of the intergovernmental panel on climate change.
Climatic Change 39(4): 621–642.
Cash, D., Clark, W., Alcock, F., Dickson, N., Eckley, N., Jager, J. 2002. Salience, credibility, legitimacy and boundaries: linking research, assessment and decision making. John F. Kennedy School
of Government, Harvard University. Faculty Research Working Papers Series. RWP02-046.
Eckley, Noelle. 2001. Designing effective assessments: the role of participation, science and governance, and focus. Report of a workshop co-organized by the Global Environmental Assessment
Project and the European Environment Agency, 1–3 March, Copenhagen, Denmark. Research
and Assessment Systems for Sustainability Program Discussion Paper 2001-16. Cambridge, MA:
Environment and Natural Resources Program, Belfer Center for Science and International Affairs, Kennedy School of Government, Harvard University.
Farrell, Alexander E., and Jill Jäger, eds. 2005. Assessments of regional and global environmental
risks: designing processes for the effective use of science in decisionmaking. Washington, D.C.:
RFF Press.
MA (Millennium Ecosystem Assessment), 2003. Ecosystems and Human Well-Being: A Framework
for Assessment. Island Press, Washington, D.C., 245 pp.
Stakeholder Participation, Governance, Communication, and Outreach | 69
MA (Millennium Ecosystem Assessment). 2005. Ecosystems and Human Well-being, Volume 4:
Multiscale Assessments: Findings of the Sub-global Assessments Working Group. Island Press,
Washington, D.C.
Mitchell, Ronald B., William C. Clark, David W. Cash, and Nancy M. Dickson, eds. 2006. Global
environmental assessments: information and influence. Cambridge: MIT Press.
Pahl-Wostl, C., 2003: Polycentric integrated assessment. In: Scaling Issues in Integrated Assessment,
J. Rotmans and D.S. Rothman (eds.), Swets & Zeitlinger, Lisse, the Netherlands, pp. 237–261.
Ranganathan, J., Raudsepp-Hearne, C., Lucas, N., Irwin, F., Zurek, M., Bennett, K., Ash, N., West,
P. 2008. Ecosystem Services: A Guide for Decision Makers. World Resources Institute.
Raustiala, K. and D. Victor. 1996. Biodiversity since Rio: The future of the Convention on Biological Diversity. Environment 38(4): 17–20, 37–45.
Reid, W. V., Berkes, F., Wilbanks, T. J. & Capistrano, D. (2006) Bridging scales and knowledge
systems: concepts and applications in ecosystem assessment (Island Press, Washington D.C.).
Conceptual Frameworks for Ecosystem
Assessment: Their Development,
Ownership, and Use
Thomas P. Tomich, Alejandro Argumedo, Ivar Baste, Esther Camac,
Colin Filer, Keisha Garcia, Kelly Garbach, Helmut Geist,
Anne-Marie Izac, Louis Lebel, Marcus Lee, Maiko Nishi,
Lennart Olsson, Ciara Raudsepp-Hearne, Maurice Rawlins,
Robert Scholes, and Meine van Noordwijk
What is this chapter about?
This chapter provides information on and lessons from experiences with conceptual frameworks that may help in adapting and developing a framework for an ecosystem assessment.
The social process to create the conceptual framework is as important as the final product.
This creative process requires interaction—and often involves tension—between users and the
assessment team. The challenge of working together to create a shared conceptual framework can play an important role in creating ownership by the users of the assessment and in
building an assessment team.
Recent experiences with global assessments, such as the Millennium Ecosystem Assessment (MA), show that conceptual frameworks can provide greater focus on key issues and
relationships and serve a useful role in synthesis and cross-site comparisons. Although the
MA framework has in some respects become a standard point of departure for ecosystem
assessment, there is no unified theory on creating conceptual frameworks. Examples from MA
sub-global assessments illustrate a range of pragmatic approaches, ranging from adaptation
of the global conceptual framework to independence from it and including the use of multiple
The people who are (or are not) informed about, consulted, and involved in creation of
the conceptual framework and the ways in which their knowledge and expertise are valued (or
not) will in many ways govern the entire assessment process. Both the groups consulted and
the components that are valued by the assessment team as well as the quality of interaction
between the assessment team and the stakeholders are important to developing a conceptual framework that effectively balances the principles of legitimacy, relevance, and credibility
discussed in Chapter 2.
The chapter begins with a simple definition of a conceptual framework and then discusses
some practical considerations of its meaning in ecosystem assessment. Section 3.2 explores
the often intertwined challenges and opportunities involved in developing a conceptual framework. Sections 3.3 and 3.4 juxtapose the dual roles of conceptual frameworks in ecosystem
72 | A Manual for Assessment Practitioners
assessments: as a means for clarity, credibility, and comparison and as a tool for engagement, usefulness, and legitimacy. Rather than adopting a conceptual framework entirely “off
the shelf,” a pragmatic approach that blends various frameworks and methods to balance
strengths and offset weaknesses seems to be the most appropriate method.
3.1 How to understand the relationship between people and nature
Section’s take-home messages
• In ecosystem assessment, a conceptual framework is a concise summary in words or
pictures of relationships between people and nature—in other words, among the key
components of interactions between humans and ecological systems.
• Conceptual frameworks can help clarify and focus thinking about complex relationships,
including how those relationships may be changing over time. They also can be a focus
for interaction to build shared understanding.
• The understanding developed when building a conceptual framework includes but is
not limited to scientific knowledge. Personal experiences, history, cultural practices and
values, political savvy, and other forms of knowledge are also important in clarifying and
enriching shared understanding.
• It is possible, perhaps even desirable, for an assessment team to use more than one
conceptual framework. What is important, though, is that at least one conceptual
framework must be embraced (that is, “owned”) by the assessment team and users alike.
This chapter proposes a simple working definition of a conceptual framework for
ecosystem assessment: a concise summary in words or pictures of the relationships
between people and nature, including how those relationships are changing over
time. These frameworks often are anthropocentric, centering on people and their
needs. This is because assessment users often are focused on the issue of how human well-being is influenced by environmental change, but also because any effort
to change the system will necessarily involve actions to change human behavior. But
there are many ways to view relationships between people and nature, and cultural
perspectives differ significantly on the centrality of nature or people. These contrasting perspectives can be enlightening in and of themselves; one reason for taking a
pluralistic approach to conceptual frameworks is that the choice of frameworks and
comparisons among them may help clarify underlying assumptions and the implicit
values being placed on different outcomes for people and nature.
Conceptual frameworks can help organize thinking and structure the work that
needs to be accomplished when assessing complex ecosystems, social arrangements,
and human–environment interactions. From a scientific perspective, a conceptual
framework can be viewed as a model to guide the assessment process. Like any
model, the framework will always be a simplification; often an extreme simplification. But simplification is a useful tool—indeed, an indispensable tool—in clarifying
and focusing an assessment process. This capacity for illuminating abstraction is
a key reason why science is indispensable to a credible ecosystem assessment. As
discussed in greater detail in section 3.3, a conceptual framework can be helpful to
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 73
focus on key issues among the myriad natural and social processes affecting ecosystems and human well-being, to frame those issues spanning multiple scales across
space and time, and to manage the interlinkages among these elements.
Highlighting underlying assumptions and gaps in understanding is an important
part of a well-designed conceptual framework. However, a framework is more than
just a list of shared assumptions. Since the assessment framework necessarily is developed by diverse stakeholders, it should ultimately be a synthesis of various ways of
knowing, explaining, and valuing ecosystems and human–environment interactions.
Developing a shared conceptual framework may generate dialogue among groups
with different assumptions, ways of understanding, and approaches to managing
dynamic natural and social systems. Therefore, a conceptual framework draws on
a variety of types of knowledge. Indeed, it typically will be a synthesis of more than
one way of knowing or understanding and hence may not be perfectly consistent or
particularly elegant.
The focus necessary for a successful assessment should reflect what people value
most about an ecosystem. Different stakeholders may emphasize different parts of
the system or different relationships. Engaging diverse groups and assessment users in development of the conceptual framework is important for ensuring that the
result is accepted or “owned” by users and by the assessment team. A conceptual
framework that is not developed through engagement with participants and stakeholders but is instead “imposed” on those who have to use it can turn out to be a
rather fruitless exercise (see Box 3.1).
Do clarity and focus require total commitment to a single framework? No, not
at all. Ecosystem assessments have been undertaken successfully with more than
one conceptual framework. In the Millennium Ecosystem Assessment, for local assessments in the Southern Africa region two additional conceptual frameworks—
on adaptive renewal (Gunderson and Holling 2002) and sustainable livelihoods
Box 3.1. The need for “ownership” by assessment teams
Assessment teams working on the International Assessment of Agricultural Science and Technology for Development (IAASTD) initially intended to use a conceptual framework inspired by
the Intergovernmental Panel on Climate Change (IPCC) and the Millennium Ecosystem Assessment (MA). Many authors pointed out what was missing from this framework and essentially
used their own approaches in early drafts. As a result, the fact that authors were expected
to produce an assessment of options became blurred, as people could not see how to conduct an assessment of agricultural knowledge, science, and technology within the proposed
Eventually the scenarios chapter in the original structure was dropped altogether, which
marked the de facto abandonment of the initial conceptual framework. It was not until the different lead authors for different chapters started discussing these challenges directly that a
new framework based on the concept of “multifunctionality of agriculture” was developed out
of an initiative that originated with the French government. This concept took on a new life over
the three years it took to produce the IAASTD, signifying that agriculture provides multiple services to people—from food, feed, and fiber to aesthetic landscapes and ecosystem services.
This framework was used (at least in passing) in most of the final chapters.
74 | A Manual for Assessment Practitioners
(Carney 1998)—were “superimposed” and used in a complementary manner with
the MA conceptual framework, in order to better capture the dynamic interplay between ecosystems and humans at the local level (MA 2005b:73). The crucial caveat
suggested by experience is that whether one, two, or several frameworks are used,
there must be at least one that is embraced by the assessment team and users alike.
3.2 How to develop a conceptual framework
Section’s take-home messages
• Rather than adopting a conceptual framework entirely “off the shelf,” a pragmatic
approach for each assessment can blend various frameworks and methods to balance
strengths and offset weaknesses.
• It is important to be aware of differing perspectives and conflicting interests within the
assessment team and among stakeholders and intended users of the assessment.
• The process of developing a conceptual framework involves predictable lines of
tension—among stakeholders, for example, or between the local assessment team and
stakeholders, within the local team, or among the local team and assessment colleagues
working at coarser scales. It is important to recognize these tensions, as they are not
dangerous if dealt with in a respectful setting.
• It is impossible to capture all of reality. A well-constructed conceptual framework should
clearly characterize the attributes of the system (conditions and interactions) that are
perceived to be most important from the standpoint of the users of the assessment as
well as the assessment team.
• To span boundaries among groups, it is important to make efforts to translate,
communicate, and mediate across participants who hold different views and have
conflicting interests. Semantics matter: the meanings of key words and concepts need to
be understood and broadly acceptable.
• The right professionals, with the right skills and experience, need to be involved in the
process. In addition to ecologists, economists, and physical scientists, skilled social and
political analysts with local knowledge and experience are essential for grounding the
process in local reality. Professional facilitators also can help with process insights and
techniques for communication and mediation to bridge lines of tension; professional
facilitation of the process of creating a conceptual framework can be an excellent
investment of time and funds.
The primary concern in this section is a local or national group that is considering
undertaking an ecosystem assessment. If they decide to proceed, what positive or
negative insights could the team draw from the experience of the MA and other
assessment processes? As emphasized in Chapter 2, meeting users’ needs is an essential element of a useful assessment process. Similarly, relevance of conceptual
frameworks to those users’ needs must be addressed as a prerequisite to articulation
of broader comparative, synthetic, or “global” aspirations of assessment processes.
This chapter—indeed, the entire manual—is intended to help new assessment teams
avoid pitfalls and enhance their effectiveness in bringing in local perspectives on
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 75
“truth” to enrich or at least balance those of “global experts” and “international
processes.” This section reviews a few lessons and insights regarding challenges of
interaction among diverse people with differing interests in order to create an (inevitably) abstract conceptual framework.
3.2.1 Striving for a workable framework that can be understood and is flexible
One of the most important functions of a conceptual framework is as a guide to
what is not being done. Since perfection is impossible, pragmatism should be the
guide. A workable, operational framework with essential elements is good enough.
Remember the saying that “if everything is important, nothing is important.” Although it is important to find the few key links among the possible relationships,
it is not helpful if the arrows indicating links show that everything is connected to
everything else.
Expect an interesting conceptual framework—one that takes some risks—but
also expect it to evolve during an assessment process. Documenting the modifications that arise during this evolution can help new entrants to an assessment process understand better why the current perspective has been adopted. It also can
allow those involved to better appreciate what has been learned. A good illustration
of the evolution of a particular family of conceptual frameworks is given in Waltner-Toews and Kay (2005) based on their work assessing ecosystem sustainability
and health.
3.2.2 Getting the process started
To get started, consider a very simple case as a point of departure: a single person
(hence a unified point of view and a single set of values, interests, and objectives) in
an isolated, island setting. The steps in developing a conceptual framework for assessment in this kind of “Robinson Crusoe” assessment would include the following:
1.Identify the key elements of well-being or “quality of life,” whether or not they
are shaped by ecosystem services.
2.Identify the ecosystem goods and services that matter the most for the elements
of well-being in (1).
3.Sketch a diagram of the factors that directly affect the supply of the ecosystem
goods and services in (2): these can be called “direct drivers.”
4.Then, move back one level and add the “indirect drivers” that influence the
direct drivers included in (3).
5.Finally, see if there are any connections from the elements of Robinson’s wellbeing in (1) back to the direct drivers (3) or indirect drivers (4) you identified.
The diagram created usually takes the form of “boxes and arrows” (see section 3.3),
where the boxes are filled with “things” such as an ecosystem service or a process
(population growth), and the arrows are influences from one thing to another. Of
course, by framing this as a “one man show,” this simple case misses a central point.
It is not the ecosystem services that you (as an assessment team member) consider
76 | A Manual for Assessment Practitioners
important that matter most. Instead, in a real application, the conceptual framework
should feature the services that your intended audience of assessment users considers
most important.
The initial diagram often is quite complex. It helps to simplify it to just the
most important links—those factors that are changing at a timescale relevant to the
assessment. Cut out the factors that are constant or change very slowly and those
that fluctuate much faster than the assessment time period. It is also safe to ignore
processes that are relatively weak.
The next step is to consider the spatial scale at which key influences occur: Are
they imposed at scales much larger than the assessment? Or do they bubble up from
much finer scales (smaller than the minimum resolution of the assessment)? The
large-scale influences are important, and they can be clarified by handling them as
indirect drivers (step 4 above). The finer-scale processes may involve too much detail
to be appropriate; they might be eliminated to help simplify the conceptual framework (and the entire assessment task).
The key distinction between this fictional island of Robinson Crusoe’s and a
more realistic assessment situation is not its ecosystem complexity (which in principle could match that of any real ecosystem) but the social, cultural, and political
complexity arising from the multiplicity of different points of view, types of knowledge, values, interests, and objectives of individuals and groups in any society (Reid
et al. 2006). The conceptual framework developed in this way is a helpful place
to start the larger process, with the understanding that it will change. Building on
the engagement strategy described in Chapter 2 to ascertain users’ needs, the communications involved in developing a conceptual framework should involve diverse
stakeholders and encompass a variety of perspectives and types of knowledge. These
connections can help build the shared understanding needed for a successful collaborative process.
3.2.3 Anticipating predictable lines of tension
If a conceptual framework is developed properly, the ecosystem services that become the focus of the framework involve high stakes and, as a result, tend to be
politicized, “hot button” issues. These often are the focus of conflict between different stakeholder groups over control and access to valuable resources and over who
bears costs and reaps benefits. But the salience of those issues for assessment users
may not fit well with the methods and approach that the assessment team needs to
ensure credibility. Moreover, because assessment teams must be multidisciplinary in
a broad sense (involving social scientists and natural scientists), an additional line of
tension typically arises within a team from interactions between scientists (regardless of their discipline) who may have little experience with (or even respect for)
colleagues trained in other disciplines. Finally, if an assessment team is engaged with
colleagues conducting parallel assessments at coarser scales, the experiences with the
global and “sub-global” components of the MA suggest that there could be tension
between the team’s broad imperative to “get on with it” and the organizer’s responsibility to deal with complex, even messy, local reality in order to produce useful
results for people with a direct stake in the assessment topics. Tomich et al. (2007)
document insights from one process of building a multiscale, multidisciplinary, multicultural team.
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 77
3.2.4 Impossibility of a ”neutral” or value-free conceptual framework
By definition, a conceptual framework seeks to draw attention to a subset of components and relationships that are believed to be most important for understanding the
system. What is “most important” to one person may be less important to others, however. From the standpoint of an environmental activist or government negotiator, the
“most important issue” may be a global consideration like greenhouse gas emissions,
while from the standpoint of a farmer, it may be access to local resources, such as land or
water. As a result, any process to develop a conceptual framework is inherently political,
involving balance and contention among different interests and concerns.
Thus all conceptual frameworks are part of the contest of values and interests of
different people; some are powerful, others not. Because no one is immune to these
pitfalls, it is important for the assessment process to be open to critical examination
of biases and prejudices cloaked as “science” or “best practice.” Work on agroforestry in Kenya by Jerneck and Olsson (2007) provides an example of how views
held by scientists can be a barrier to understanding opportunities not just for local,
private benefits but also for enhancing the public good globally. Thus, openness to
alternative perspectives applies with particular force to the words and slogans used
by the assessment team, who must be sensitive to differing and conflicting views
when developing a framework for assessment (see Box 3.2).
This essential sensitivity to alternative views is especially important in the early
phases of the assessment process, including development of the conceptual framework. Different people can have very different conceptual starting points (i.e., their
own mental models about how the world works). The risk is that the assessment
team’s uncritical use of words and slogans—and belief in them—can inhibit the process of identifying other views that may be equally valid, relevant, and useful. If
handled effectively, then, the diversity of views is not just of academic interest but
can be an important asset in creating a framework for assessment.
3.2.5 The process is as important as the framework
The conceptual framework for an assessment depends on purpose, time, and place.
It can play a key role in communicating the results of the assessment, especially to
audiences that are more visually oriented than word or text focused. But the framework should be seen as the product of a process that has its own value rather than
simply as a starting point for assessment. If the framework is derived from a process
that truly involves multiple stakeholders and that respects and embraces differences
in their perspectives, experience, and knowledge, it can provide an important element of the legitimacy and usefulness required for a successful assessment.
Given this central importance of the conceptual framework both as a process
and as a product, and given also the fact that a high-quality process must grapple
with social and political realities, it is important to involve the right professionals,
who have the right skills and experience. In particular, skilled social and political
analysts with local knowledge and experience are essential. And professional facilitators also can help with process insights and techniques for communication and
mediation to bridge lines of tension. Indeed, if funding allows, professional facilitation of the process of creating a conceptual framework can be an excellent investment, producing rewards throughout the assessment process.
78 | A Manual for Assessment Practitioners
Box 3.2. Using conceptual frameworks for conceptual clarity
Multiple perspectives and conflicting political interests affect use and interpretation of words
within science, just as they do in the “real” world. A single word or slogan sometimes embodies
a way of thinking about relationships between people and nature. Whether used consciously
or (more often) unconsciously, these concepts can be packed with complex meanings for one
person that are neither shared nor apparent to others.
The English origins and current international use of the word “forest” represents a very
powerful example of a concept packed into two syllables, often separating nature from people
and asserting claims on behalf of the ruler or state over forest resources to the exclusion of
local people. The Magna Carta from 1215 included a promise by King Henry II to “deforest”
lands recently acquired, thereby returning the control over these lands to local communities or
elites. Michon et al. (2007) and Van Noordwijk et al. (2008) explored how use of the term forest and its implicit institutional and political intentions was intended to allow clear-felling and
replanting as legitimate forest management practices and a means to control land and other
natural resources. In contemporary times, the meaning has been inverted to ban management
involving land clearing as antithetical to sound “forestry” even if the smallholder agroforestry
systems, such as Indonesia’s kebun lindung (“protective gardens”), outperform conventional
forest management economically, socially, and environmentally.
Slogans can be even more insidious than single words, because they tend to pack more
punch, impairing communication and enshrining fuzzy thinking. “No forest, no river,” “No river,
no water,” “Forests prevent floods”—these slogans are commonly used in public discourse
and purport to be based on sound science. Across the tropics, they have been used to justify
evictions from “critical” watersheds. They also have been marshaled to justify hundreds of millions of dollars in public expenditure for reforestation under the control of forestry departments.
The condensed representation of “knowledge” used in these slogans has been very powerful in
the politics of control over the landscape and its resources in many countries.
An inclusive, participatory approach to developing conceptual frameworks may be an effective means of revealing the fallacies and logical errors in powerfully ambiguous words and
falsely concrete slogans.
3.3 How to use conceptual frameworks to focus your assessment
Section’s take-home messages
• Conceptual frameworks can provide greater focus on key issues and relationships
in assessment efforts; they also can serve a useful role in synthesis and cross-site
• The MA conceptual framework has in some ways emerged as a dominant framework
for ecosystem assessment, but there is no unified theory on the creation of conceptual
frameworks. Still, the structure and elements of the MA conceptual framework and related
examples may be useful starting points in developing a framework.
• The MA conceptual framework (or any framework) should not be adapted for use in a
new context without understanding its original purpose and deciding whether and how to
adapt or translate the diagrams used to summarize its content.
• Much of the really useful bit of the approach to conceptual frameworks in the MA was
not derived from a four-box diagram but from the higher-level idea linking human
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 79
well-being and ecosystem services—that vagueness opened opportunities for
The structure and elements of a conceptual framework cannot be comprehensive. They
need to focus on the most salient issues for users of a particular ecosystem assessment.
The structure and elements of a conceptual framework also are the foundation for
identification, prioritization, and development of appropriate indicators for conditions and
trends in ecosystems.
It is important to try to keep the framework simple, paying particular attention to possible
thresholds and turning points.
Assessment teams face important choices in seeking balance between “top down”
approaches to developing conceptual frameworks for synthesis and comparison and
“bottom up” approaches in which collaborative development of a framework can be a
means to engage stakeholders to ensure usefulness and legitimacy and thereby create a
sense of ownership of the assessment process.
A conceptual framework, in the broadest sense, is a tool for organizing ways of
thinking about the subject at hand: “A well-designed framework for either assessment or action provides a logical structure for evaluating the system, ensures that the
essential components of the system are addressed as well as the relationships among
those components, gives appropriate weight to the different components of the system, and highlights important assumptions and gaps in understanding” (MA 2003:
34). For ecosystem assessment, the “subject at hand” is the relationship between
people and nature or, more specifically, between human well-being and ecosystem
services. Of particular interest is how those relationships may be changing. There are
many examples of conceptual frameworks but no unifying theory. And, in practice,
there is a huge range in the use of the term. A conceptual framework could be a
formal model, a sketch on the back of a used envelope, or a figure produced through
collaborative effort. It could be accessible or off putting, could clarify or obscure,
and could engage others’ views or simply presume them.
A conceptual framework does not necessarily have to be represented with a
diagram, but often a visual representation increases the effectiveness in conveying
the ideas and elements of the framework to most audiences, including users and
members of the assessment team. Done well, graphical representations can help
clarify and pinpoint key issues within webs of complex relationships; they can help
bring differences into focus and also aid in developing a shared terminology and
data protocols that can facilitate interdisciplinary and cross-site collaboration. The
framework, then, is an important—a nearly indispensable—tool for design of multistakeholder, multidisciplinary, multiscale, multisite assessments.
Use of any type of figure in clarifying thinking must be balanced against the risks of
rigidity in approach and other pitfalls. On the one hand, it is well established that overly
narrow analyses of environmental issues can be highly misleading (Baumol and Oates
1988). Neglecting real complexities when framing problems can lead to overly simplistic
solutions that are not relevant or applicable to the intricacies and difficulties of real situations. Recognizing the complexity, interconnections, and trade-offs involved is critical.
However, there is an equal risk of lack of focus and, consequently, being overwhelmed
and ultimately paralyzed by detail. Conceptual frameworks can be an antidote to this
problem of “everything depends on everything and everyone.” But this also means that
compromises are inevitable and that there is no single “best” representation.
80 | A Manual for Assessment Practitioners
3.3.1 Overview of the structure and main elements of the MA conceptual
There are many approaches to thinking through and structuring graphic conceptual
frameworks (see Boxes 3.3, 3.4, and 3.5). Any graphic representation of a conceptual framework is
“simultaneously helpful and problematic, [has] an ad hoc flavor, and should be viewed
as a heuristic device. Box and arrow frameworks reflect the infancy of theoretical studies, and were designed to facilitate the quest for general principles and integrated theories.” (Lambin et al. 2006:6)
The particular structure used in the global portion of the Millennium Ecosystem
Assessment (see Figures 3.1 and 3.2) had its origins in the Drivers-Pressures-StatesImpacts-Responses (DPSIR) framework (Smeets and Weterings 1999, Pirrone et al.
2005 and, for a critical view, see Svarstad et al. 2008). A simpler version of this,
called Pressure State Response (PSR), is among the frameworks most frequently
used by international development organizations (Spreng et al. 1996). The foremost
difference between PSR and DPSIR is that the latter identifies drivers explicitly and
introduces the impact category, which helps highlight the most salient causal relationships. Both DPSIR and PSR are linear frameworks, examining how changes in
pressures affect systems but not including the full loop to consider how changes in
the system then feed back to affect pressures. The MA framework differs in fundamental ways from these in that it incorporates both this feedback loop and multiple
temporal and spatial scales.
The foundation of the application of DPSIR is a comparison of reference conditions
(or baselines) in ecosystem services with the same or similar systems under increasing degrees of human intervention and disturbance. Historically, biologists and ecologists have
sought the general rules governing natural systems by studying an idealized notion of
nature in its most “pristine” state. Understanding these rules and drawing comparisons
between “pristine” natural areas (reference conditions) and areas with varying degrees
of human intervention is the main method used to gauge the type and extent of impacts
of human activities on ecosystems (Spreng and Wils 1996). In contrast, the MA conceptual framework does not assume any “natural” state for the system and instead treats
the system condition as a dynamic response to changes in drivers.
Box 3.3. Other systems approaches to thinking and graphic representation
DPSIR is closely related to systems analysis and hierarchy theory, both of which can provide
additional insight into multiscale analysis. In these approaches, a “system” is perceived as a
whole, with elements intrinsically related because they consistently affect one another over
time and “operate towards a common purpose” (Senge et al. 1994:90). “Hierarchical systems”
often are described as a nested set of subsystems, including the key feedback loops affecting
the nature of the subsystem interactions (Simon 1962).
Figure 3.1. MA conceptual framework.
Source: MA 2005a
Figure 3.2. Links between ecosystem services and human well-being.
Source: MA 2005a
Box 3.4. Comparing different uses of DPSIR frameworks
The conceptual framework of the Intergovernmental Panel on Climate Change is perhaps the
most prominent application of the DPSIR approach (see Figure A). (Others include the UN
SCOPE project and the Organisation for Economic Co-operation and Development framework.) The IPCC has been a leader in developing the notions of differential vulnerabilities and
adaptive capacities mediating impacts and has usefully introduced some new broad classes
of response options (mitigation, adaptation). But whereas the IPCC focuses on a single direct
driver (climate change) and on impacts across a range of specific sectors (energy, food, water),
the Millennium Ecosystem Assessment considered multiple direct drivers and focused on the
effects on a range of ecosystem services. The basic contrasts between the MA and the IPCC
framework illustrate that even when both approaches follow the DPSIR structure, important
choices need to be made about what to include as major elements, which depend on the purpose of the assessment.
The Global Environment Outlook (GEO), produced by the U.N. Environment Programme
(UNEP) since 1997, uses the DPSIR framework. The fourth assessment (GEO-4) was prepared
through an intergovernmental and scientifically independent process with features similar to
those of the IPCC, MA, and IAASTD (UNEP 2007). This evolution of the GEO process was part
of a wider endeavor to strengthen the scientific base of UNEP. Different approaches of governments and experts had become apparent during consultations: some argued for the use of the
DPSIR conceptual framework used in previous GEOs and others argued for the MA framework
(UNEP 2004).
This dilemma was tackled in GEO-4 by evolving the DPSIR framework to include concepts
from the MA and IPCC frameworks. Framing an assessment of how environment contributes to
development was the aim. The interface between people and the environment was seen as the
carrying concept for such a frame (see Figure B). As in the MA, the development of the framework created tension among authors—a tension that was not fully resolved in the process due
to time constraints.
The GEO-4 framework is inspired by the MA, but because of its broader scope it places
ecosystem services and human well-being in a wider context. It prompted GEO-4 to question
some of the assumptions in the MA. A minor but illustrative conceptual difference is that the
MA considers climate change as a direct driver, while in GEO-4 it is seen as one among several
interacting forms of environmental change.
Compared with the MA, GEO-4 expands the environmental factors that determine human
well-being beyond ecosystem services. Nonecosystem natural resources and the stress that
the environment imposes on society (in the form of diseases, pests, radiation, and hazards)
were added to expand the basis for assessing trade-offs.
A core development in GEO-4 is its acknowledgement of the fact that environmental factors are interacting with demographic, social, and material factors in determining human wellbeing. A variation of the framework presented in GEO-4 illustrates the dual role of social and
economic sectors (see Figure C). The sectors are driving environmental change, and at the
same time they are instrumental in shaping the way the environment affects people. Again,
this was considered important in order to assess real life trade-offs. But even more so, it was
needed to allow in-depth analysis of people’s vulnerability to environmental change.
The GEO-4 conceptual framework condenses a number of ideas into one figure. This
approach illustrates the risk of overloading a framework, but it also demonstrates how wellknown concepts from different realms can be combined to give new insights. In this respect it
Box 3.4. continued
is worth noting the attempt to incorporate the concept of material, human, social, and natural
capital into the equation. Also, it is worth noting that the approach to “responses” echoes,
albeit broadens, the approach in the IPCC framework. In GEO-4, “responses” are seen as “formal and informal adaptation to, and mitigation of, environmental change” (UNEP 2007: figure
1, p. xxii).
Finally, the GEO-4 framework demonstrates that the concepts introduced in the MA
can further evolve and combine with other approaches such as the DPSIR framework. New
assessments, whether global or subglobal in nature, should continue to challenge the current
conceptual understanding of the complex chains of cause and effect taking place in space and
time that characterize the interactions between people and the environment.
Figure A Box 3.4. Conceptual framework of the Intergovernmental Panel on
Climate Change.
Source: IPCC 2007
Figure B Box 3.4. DPSIR framework used in GEO-4.
Source: UNEP 2007
Figure C Box 3.4 . A variation of the GEO-4 conceptual framework highlighting
the dual role of the social and economic sectors.
Source: UNEP 2007
Box 3.5. Multiple conceptual frameworks for dryland systems
It is fair to conclude that the processes of degradation of dryland systems, sometimes called
desertification and “defined as a persistent decrease in provisioning of ecosystem services”
(MA 2005a:645), remain poorly documented at the global scale (Lambin et al. 2003). To better
frame the assessment of dryland systems under the U.N. Convention to Combat Desertification (UNCCD), the European Commission (EC), through the Institute of Environment and
Sustainability of the Joint Research Centre, started a process in 2008 that illustrates the role
of multiple conceptual frameworks in the absence of a single, synthetic framework. The overall
objective is to evaluate and integrate strategic indicators and benchmarks for a comprehensive
assessment in a multistage process.
First, existing indicator systems and conceptual frameworks are critically reviewed from
the perspective of the Dahlem Desertification Paradigm (DDP). While a paradigm, in general,
relates to a set of practices that define scientific approaches during a particular period of time
(Kuhn 1962), the DDP attempts to capture the multitude of biophysical and socioeconomic
interrelationships within dryland systems from the perspective of human-environmental system
dynamics (Newell et al. 2005). There are nine assertions and implications arising from the DDP
(Reynolds and Stafford Smith 2002, Reynolds et al. 2007), so that in the end the paradigm
might turn into “a single, synthetic framework . . . (that) is testable, which ensures that it can be
revised and improved” (Lambin et al. 2007:336). These assertions are as follows (with implications not detailed here):
Desertification always involves human and environmental drivers.
“Slow” variables are critical determinants of system dynamics.
Thresholds are crucial and may change over time.
Costs of intervention rise nonlinearly with increasing degradation.
Desertification is a regionally emergent property of local degradation.
Coupled human–environment systems change over time.
Development of appropriate local environmental knowledge must be accelerated.
Systems are hierarchically nested (so, manage the hierarchy).
A limited suite of processes and variables at any scale makes the problem tractable.
Second, indicators are held against existing conceptual frameworks to test and illustrate their
potential for application. The EC initiative provides several examples of frameworks with relevance for dryland systems; two of them are illustrated here only to demonstrate their relevance
under the DDP:
• Human Ecosystem Model: Developed at the University of Idaho in the United States, this
model has the character of an organizing principle to design “a coherent system of biophysical and social factors capable of adaptation and sustainability over time.” The flow
and use of a limited set of critical resources are seen to be regulated by the social system,
creating the so-called human ecosystem (see Figure A) (Machlin and McKendry 2005). The
model is multiscale and hierarchically nested, and it is considered most useful for predicting and evaluating cascading and nonlinear effects; it is also able to synthesize a large
range of theory, method, and evidence, including dryland systems.
Box 3.5. (continued)
Figure A Box 3.5. Human Ecosystem Model.
Source: Machlin and McKendry 2005
• Adaptive Cycle Model: Developed by the Resilience Alliance, a collaboration of scientists
and practitioners to explore the dynamics of socioecological systems (Folke et al. 2002),
the model is a “tool of thought” to focus on the interaction of processes of destruction
and reorganization in complex adaptive systems (thus moving beyond traditional notions
of stable state and succession). It identifies four distinct phases with different rates of
change; the two newly added functions are rapid transitions that occur from “collapse”
(or release) to reorganization (the so-called Omega>Alpha backloop), while the so-called
foreloop (r>K) is modeled as a slow, incremental process of growth and accumulation (see
Figure B). Adaptive cycles are seen to be nested in hierarchies across time and space,
representing a panarchy (Gunderson and Holling 2002). This implies that the results
obtained from drylands would not trigger cascading instabilities of the whole system because of the stabilizing nature of nested hierarchies.
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 87
Box 3.5. (continued)
Figure B Box 3.5. Adaptive Cycle Model.
Source: Folke et al., 2002
This section briefly considers each of the main components of the MA conceptual framework:
• Indirect drivers (a subset of the “pressures” in a PSR framework); also
sometimes called underlying drivers.
• Direct drivers, which, along with indirect drivers, constitute the “pressures” in
a PSR framework. These also are called “driving forces,” “process variables,”
or “control variables” in other frameworks. Indicators in this category gauge a
process that will influence conditions and trends (state variables).
• Conditions and trends in biodiversity, ecosystem services, and human well-being; known as “state variables” in PSR frameworks.
• Impact variables, showing cause-and-effect relationships linking different elements. Impacts are symbolized by the arrows in Figure 3.1 and are elaborated in
Figure 3.2 for the specific cause-and-effect relationships linking various ecosystem services and elements of human well-being. In a PSR framework these are
unidirectional, whereas in the MA framework bidirectional impacts are allowed.
Although there is an element of bidirectionality in many relationships, doubleheaded arrows should be used sparingly and reserved for the most salient feedbacks in a conceptual framework.
• Response variables, which are efforts by people (individually or collectively) to
reduce negative impacts and enhance positive impacts in order to attain more
desirable levels in the pressure and state variables. Responses are symbolized by
the symbols marked “strategies and interventions” in Figure 3.1. Assessment
88 | A Manual for Assessment Practitioners
of these strategies and interventions is the focus of Chapter 6. Participatory
appraisal, policy analysis, cost–benefit techniques, impact assessment, and
outcome mapping are a few of the tools that can be used to assess responses.
Conditions and trends in state variables (including impacts) and assessment of responses are treated in detail in Chapters 4 and 6, respectively. This chapter devotes
additional attention to drivers.
3.3.2 Drivers of ecosystem change
Ecosystem change is always caused by multiple interacting factors originating from
different levels of organization (scales). The mix of drivers varies in time and space;
some common changes result from a combination of biophysical and socioeconomic
drivers that work gradually (e.g., declining infant mortality, the spread of salinity),
while others happen precipitously (e.g., drought, hurricane, war, economic crisis).
The specific abiotic and biotic factors that merit most attention vary among localities
and regions (Lambin et al. 2003).
As noted earlier, factors causing ecosystem change do so either directly or indirectly. A direct driver unequivocally influences ecosystem processes, while an indirect
driver operates more diffusely, by altering one or more direct drivers. The indirect
drivers are underlying (root) causes that are formed by a complex of social, political,
economic, demographic, technological, and cultural variables. Collectively, these
factors influence the level of production and consumption of ecosystem services.
The causal linkage is almost always mediated by other factors, thereby complicating statements of causality or attempts to establish the proportionality of various
contributors of change.
Meta analyses suggest that the five most important groups of indirect drivers to
consider in developing a conceptual framework for ecosystem assessment are:
Population change (demographic drivers);
Change in economic activity (economic drivers);
Sociopolitical drivers;
Cultural (and religious) drivers; and
Technological change (science and technology).
Important direct drivers to consider in a conceptual framework for ecosystem assessment include:
• Habitat changes (driven through land use/cover change, physical modification of
rivers, or water withdrawal from rivers);
• Overexploitation;
• Invasive alien species;
• Pollution; and
• Climate change.
Taking this list one step further, direct drivers of land use change and land cover
change (e.g., deforestation, desertification, human settlement) are activities such
as logging, cropland expansion, road building, and other types of infrastructure
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 89
development (Geist and Lambin 2002, Geist and Lambin 2004, MA 2005a:73–76,
Nelson et al. 2006).
The various individual causes or drivers of ecosystem change, such as those just
listed, interact directly, are linked via feedback, and thus often have synergistic effects. Conceptual efforts to frame these interactions relate to pathways, trajectories,
causal clusters, or “syndromes.” Common to such conceptual frameworks is that
not all causes of change and not all levels of organization are equally important,
that certain driver combinations appear repeatedly, and that a limited suite of processes and variables at any scale makes the problem tractable (MA 2003, Lambin et
al. 2003, MA 2005a:73–76). At least two broad strands of treatment of drivers in
conceptual frameworks can be distinguished: the concept of clusters, pathways, or
trajectories, and the syndromes concept.
The clusters, pathways, or trajectories concept was developed within the LandUse/Cover Change (LUCC) Programme (Lambin et al. 2006). This framework
helped the MA distinguish indirect drivers (“underlying driving forces”) and direct drivers (“proximate causes”) (MA 2003, Reid et al. 2006). Summarizing from
a large number of empirical–analytical cases studies, it has been found that land
change is driven by a combination of a limited number of fundamental high-level
causes (resource scarcity, changing opportunities, policy interventions, vulnerability,
sociocultural factors) that combine direct and indirect drivers in a typical situation,
making a difference between slow and fast variables (see Table 3.1) (Lambin et al.
In tropical forest and (sub)tropical dryland systems, frequent and recurrent
(“robust”) driver combinations can be distinguished (Geist et al. 2006). In tropical
forests, a contemporary pattern stems from the necessity for road construction that
is associated with wood extraction or agricultural expansion. Such expansions are
mostly driven by policy and institutional factors, but they also involve economic
and cultural drivers (e.g., frontier mentality), with variations of the pattern existing across time and space (Geist and Lambin 2002, MA 2005a:585–621). In (sub)
tropical dryland ecosystems, a recurrent pattern of causal interactions stems from
the necessity for water-related infrastructure that is associated with the expansion
of irrigated croplands and pastures, mostly driven by policy, economic, and technological factors (Geist and Lambin 2004, MA 2005a:623–62). Again, the pattern has
been found to vary in space and time (Geist 2005).
The syndromes concept was developed at the Potsdam Institute for Climate
Change in 1994–2000 as a transdisciplinary core project. The concept describes archetypical, dynamic, and coevolutionary patterns. A list of about 100 “symptoms”
of change (e.g., agricultural intensification) has been reduced to six “syndromes.”
Borrowing from medical science, a syndrome is defined as a typical cluster of symptoms and their interrelations, and it has been found that syndromes once identified
will form clusters that only weakly interact with each other. Each syndrome (Sahel,
Aral Sea, Dust Bowl, overexploitation, favela, mass tourism) links processes of degradation to both changes over time and the status of state variables (Schellnhuber
et al. 2002). For example, the so-called Aral Sea syndrome describes socioecological
deterioration as a consequence of large-scale infrastructure projects, and the Sahel
syndrome refers to the overuse of marginal agroecosystems by poor, impoverished
rural populations with little or no livelihood options, thus triggering further degradation and poverty.
Natural population growth
and division of land
Domestic life cycles that
lead to changes in
labor availability
Loss of land productivity
on sensitive areas
following excessive or
inappropriate use
Failure to restore or to
maintain protective
works of environmental
Heavy surplus extraction
away from the land
Spontaneous migration,
forced population
displacement, refugees
Decrease in land
availability due to
encroachment by other
land uses (e.g., natural
reserves or the tragedy
of enclosure)
Source: Lambin et al. 2003:224.
Resource scarcity
causing pressure of
production on resources
Capital investments
Changes in national
or global macroeconomic and trade
conditions that lead
to changes in prices
(e.g., surge in energy
prices or global
financial crisis)
New technologies for
intensification of
resource use
Increase in
and agroindustrialization
Improvement in
accessibility through
road construction
Changes in market
prices for inputs
or outputs (e.g.,
erosion of prices of
primary production,
unfavorable global or
urban–rural terms of
Off-farm wages and
Changing opportunities
created by markets
Table 3.1. Typology of the causes of land use change
Rapid policy changes (e.g.,
Government instability
Economic development
Perverse subsidies,
policy-induced price
distortions and fiscal
Frontier development (e.g.,
for geopolitical reasons
or to promote interest
Poor governance and
Insecurity in land tenure
Outside policy
Internal conflicts
Illness (e.g., HIV)
Risks associated with
natural hazards (e.g.,
leading to a crop
failure, loss of resource,
or loss of productive
Impoverishment (e.g.,
creeping household
debts, no access
to credit, lack of
alternative income
sources, and weak
buffering capacity)
Breakdown of informal
social security
Dependence on external
resources or on
Social discrimination
(ethnic minorities,
women, lower class
people, or caste
Loss of adaptive
capacity and increased
Loss of entitlements
to environmental
resources (e.g.,
expropriation for largescale agriculture, large
dams, forestry projects,
tourism and wildlife
conservation), which
leads to ecological
marginalization of the
Changes in institutions
governing access to
resources by different
land managers (e.g.,
shift from communal to
private rights, tenure,
holdings, and titles)
Growth of urban
Breakdown of extended
Growth of individualism
and materialism
Lack of public education
and poor information
flow on the environment
Changes in social
organization, in resource
access, and in attitudes
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 91
Syndromes reflect both expert opinion and local case study information. The
approach is applied at the intermediate functional scales that reflect processes taking place from the household level up to the international level. Syndromes aim at a
high level of generality in the description of mechanisms of environmental degradation rather than processes of restoration, renewal, or reorganization. As for land
degradation in (sub)tropical dryland systems, several syndromes may be applied
(e.g., “Sahel,” “Aral Sea,” and “overexploitation”). The concept and the Sahel syndrome, in particular, have been criticized because of the exclusive use of downward
spirals, inherent Malthusian thinking, tautological reasoning, and the lack of actor
orientation (Geist 2006).
As for decision making at the local level, a distinction can be made between
exogenous drivers—that is, those outside the control of local land managers and
mainly including indirect drivers (prices, markets, technology development) but also
a few direct drivers (e.g., climate change)—and endogenous drivers—that is, those
under the control of local managers and mainly including direct drivers (use of external inputs, species introductions and removals), with only a few indirect drivers (e.g., technology adaptation). The implication is that, conceptually, intervention
measures might be targeted to underlying causes (indirect drivers) rather than proximate ones (direct drivers) and to weakening the positive feedbacks that accelerate unsustainable ecosystem change while strengthening the negative feedbacks that
slow or dampen ecosystem change (see Figure 3.3). The conceptual model shows
where, during the processes of LUCC, national policy is likely to have the most impacts on land use (areas indicated with a superscript 2) and where interventions will
be more difficult (areas indicated with a superscript 1) (Reid et al. 2006:162).
3.3.3 Flows of ecosystem services, stocks of resources, or both as state
Flows of ecosystem services per unit time are closely linked to changes in the levels
(stocks) of resources needed to produce those flows: there can be little cycling of soil
nutrients (a flow of fertility) without the necessary populations of soil organisms (a
stock of biological resources).
The MA conceptual framework emphasizes flows of ecosystem services and does
not work as well regarding the stocks of resources that are essential to sustainability.
Though flows of ecosystem services may sometimes be an effective proxy for the
status of the underlying natural resource base, this is not always the case. Increasing
pumping of groundwater (an increasing service flow) can be (and often is) associated with aquifer depletion (a change in a stock).
Resource stocks are increasingly recognized as crucial variables for measurement, presumably because they can be used to determine how resources will persist
under current or future patterns of use (Victor 1991). Sustainability assessments informed by this school of thought often emphasize maintaining capital stocks (Stern
1995), which are also described as state variables (Ludwig et al. 1997, Bell and
Morse 1999). This “capital theory” grew from the extensive literature on economic
growth and finite resources that thrived in the 1970s (Victor 1991). Work by the
U.K. Department for International Development (DfID) is one prominent example
of the capital theory approach (often called the “livelihoods approach”), which focuses on “five types of capital” (natural, physical, financial, human, and social) as a
92 | A Manual for Assessment Practitioners
Figure 3.3. Conceptual model of land use and land cover change.
conceptual framework for planning, programming, and project design for sustainable development.
Arrow et al. (2004) have proposed a consistent and empirically tractable concept of sustainability that provides a useful framework for focusing on the resource
base, interpreted broadly in terms of the stocks of the five types of capital developed
by DfID. The upper right box in Figure 3.4 incorporates a list of resource stock
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 93
Figure 3.4. Modified conceptual framework for California Agroecosystem Assessment.
Source: Adapted from Millennium Ecosystem Assessment and DFID frameworks
variables within a DPSIR framework adapted for agroecosystem assessment in California. Of course, adding these variables would significantly increase the measurement challenges and costs in any attempt to monitor the full range of indicators.
3.3.4 Multiple scales in space and time
DPSIR approaches have been criticized for their linearity and promotion of onedimensional thinking about relationships among drivers and state variables. The
potential danger lies in obscuring rather than revealing critical relationships and
encouraging one-dimensional, quick-fix responses (Bell and Morse 1999). Similarly,
linear DPSIR approaches have been deemed to be overly narrow, as they are unable
to account for the background processes that determine ecosystem and environmental health (Berger and Hodge 1998). More dynamic approaches that emphasize
feedbacks and multiple scales may help reduce these criticisms.
The MA framework differs from the basic DPSIR approach in two important
ways. First, the MA framework (see Figure 3.1) incorporates multiple spatial scales
(see Figure 3.5 for additional examples), which are not commonly included in PSR.
Second, the MA incorporated feedbacks from both environmental changes and related
consequences on human well-being over time. An important aspect of this second
point is that feedback loops make the MA a more dynamic system, which can be described as circular or webbed, in comparison to the relatively linear DPSIR approach.
The MA framework’s embrace of multiple spatial scales essentially combines
DPSIR with the idea of nested spatial domains. As noted earlier, DPSIR has a great
advantage over PSR in its attention to the causality of underlying impacts, and
94 | A Manual for Assessment Practitioners
Fig. 3.5.A. Conventional nested spatial domain conceptualization as proposed by
Hägerstrand. The figure shows how the symbolic transaction domains are situated and
operate under each level above them. The actors at the lowest level are then left with a
set of freedoms to carry out their daily activities.
Fig. 3.5.B. Conceptualization of the merging of DPSIR scheme(s) into the system of
nested spatial domains. The taxonomy allows for a focus on both actor constraints and
the problem’s causal linkages at each level.
Figures 3.5 A and B. Approaches to graphic depiction of spatial scale.
Source: (Ness et al. 2008)
adding more spatial scales becomes even more powerful. This approach also can be
used to integrate actors who only operate in certain spatial domains, such as some
official agencies (see Box 3.6).
At some point, though, there is a limit to the number of spatial and temporal
scales any assessment team can handle. Since many of the differences among uses
relate also to spatial or temporal scales, it makes sense to consider dividing into a
multilevel collection of nested assessments, so that various issues can be analyzed at
appropriate levels without being overwhelmed by the complexity of dealing at the
same level of detail for the whole system (see Figure 3.6).
Box 3.6. Drivers link across scales
Summarizing a large number of case studies across various ecosytems, Geist and Lambin
(2002, 2004) found that typically three to five indirect drivers underpin two to three direct drivers. These findings also shed some light on possible cross-scale dynamics. For example, for
changes at tropical forest margins (Table A), it appears that local-to-global interplays are much
more common than in the case of dryland systems (Table B), where local-to-national interactions are much more widespread. This insight suggests that global ownership of forest ecosystem assessment appears to be achievable, but not the global ownership of dryland ecosystem
assessments. The latter might explain part of the UNCCD difficulties in achieving a comprehensive assessment and in implementing top-down measures in general (Geist et al. 2006).
Table A. Indirect drivers of tropical deforestation by scale of influence
All factors
Policy and
Cultural or
*Six cases of unspecified population pressure could not be attributed to scales.
Geist et al. 2006:64.
Table B. Indirect drivers of desertification by scale of influence
Policy and
Cultural or
*Thirty-five demography-driven and 32 climate-driven cases could not be attributed to scales.
Geist et al. 2006:65.
96 | A Manual for Assessment Practitioners
Figure 3.6. Upper tributary watersheds provide habitat for in situ conservation of
biodiversity and ecosystem goods and services valued at other levels.
(Source: Lebel et al. 2008)
Multilevel approaches also may help reduce the tensions that would otherwise
arise in a monolithic single-level approach (Lebel 2006). When ways of using and
managing different goods and services interact, then understanding both past trends
and future possibilities in quantitative, linear ways can be extremely difficult. Pluralistic devices, like developing some parallel assessments from different perspectives,
may help deal with complexity.
3.3.5 Complexity and uncertainty
Complexity and uncertainty feature in every aspect of an ecosystem assessment and
arise from a number of sources.
• Incomplete and imperfect data. Researchers never have complete information,
and all data have some associated uncertainty; “reserved words” and other
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 97
techniques for systematic assessment of uncertainty in quantitative and
qualitative data are reviewed in Chapter 4.
Uncertainty about the course of future events and about the effectiveness of responses to those events. “Pure” uncertainties about future developments can be
explored using scenarios; these and related methods are presented in Chapter 5
(also see MA 2005b:229–59). Uncertainties about impact and effectiveness of
responses are considered in Chapter 6.
Uncertainties about how complex systems work (and which parts matter most).
Lack of knowledge of key causal relationships can be dealt with by proposing
alternative conceptual frameworks and treating them as competing frameworks.
But when there are too many uncertainties of this sort, an assessment may be
very hard to conduct. Nevertheless, it still may serve the purpose of identifying
where further monitoring, observations, and research are needed (Heinz Center
Uncertainty about the shape of key empirical relationships. Jared Diamond and
others (e.g., Chapin et al. 2000) have emphasized the past and possible future
roles of rapid, unanticipated catastrophe in social and ecological history. It is
important to know when the world is getting close to the edge. But identifying which thresholds and turning points should be included in a conceptual
framework for an assessment is not easy. It depends on attaining a good enough
understanding of the ecosystem that, for example, alternative stable states can
be identified (Resilience Alliance & Sante Fe Institute 2008). Thresholds, tipping points, discontinuities, and other nonlinearities—once acknowledged—also
create challenges for representation in the textual and graphical summaries
typically developed for assessments. For example, arrows drawn to indicate relationships among actors or ecosystem components may need to be completely
reconfigured if a certain condition or circumstance occurs. One simple method
might be to show the assumptions in two extreme cases (alternative stable states
if these exist) as adjacent panels (see Figure 3.7).
Adaptive management, social learning, and muddling through. Finally, a
common source of uncertainty is the presence of multiple interests and kinds of
users, who consequently have different perceptions about what the valued or
relevant set of “ecosystems goods and services” are or what should be included
in a conceptual framework and given priority for assessment (Lebel et al.
2008, Lebel and Garden 2008). This is illustrated in the case of upper tributary
watersheds in Figure 3.6 and is considered further in the next section. Adaptive
management and evolutionary approaches focus explicitly on social learning
processes to address uncertainty. The theory of adaptive management rests on
the recognition that humans often do not have the information or capacity
necessary to manage ecosystems, which are dynamic, self-organizing systems
that undergo cycles of instability and resilience (Walters and Holling 1990).
Adaptive management has been described as a partner in scientific discovery
that surveys the environment of interest and provides updated information
and understanding that can be used to guide decision making. In environments
where uncertainty is high, there are few reliable answers, but potential solutions
can be explored through experimentation. Lee describes adaptive management
approaches “as experiments that probe the responses of ecosystems as people’s
behavior in them changes” (Lee 1999: 1). Prominent examples of this approach
include logical frameworks, strategic mapping, neighborhood sustainability
98 | A Manual for Assessment Practitioners
Figure 3.7. One way to represent thresholds as part of graphic conceptual framework
for an ecosystem assessment.
Source: Lebel 2006
indicators (Meter 1999), integrated assessments (Kasemir 1999, Bell and Morse
2003), and systematic sustainability analysis (Bell and Morse 2003).
Many of these challenges of complexity and uncertainty are best addressed with
the participation not just of the scientific and policy users of an assessment but also
the users of the ecosystem themselves. Multistakeholder processes that encourage
informed input and deliberation from a wide variety of sources may be needed to
negotiate an acceptable initial framework to guide the assessment (Dore 2007, Warner 2006).
3.3.6 Unpacking the MA conceptual framework
Experience with the MA sub-global assessments showed it is not always easy to balance “top down” and “bottom up” tensions between assessment teams and users
operating at different scales and from different perspectives. Tensions across scales
and among different groups are inevitable. From the global perspective, the process
of constructing a conceptual framework for ecosystem assessment within the framework of multilateral environmental agreements cannot help but obscure cultural
differences and political conflict. There also is an important distinction between the
treatment of complexity and uncertainty in a necessarily abstract conceptual framework, as just described, and the messy political complexity that is among the most
salient considerations for local people and political leaders.
In a world of great complexity and imbalances in political power, it is not surprising that difficulties can be encountered by teams working at local and national
scales. Moreover, regardless of the scales in question, it is important to note that
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 99
some situations cannot be “assessed” by concepts represented in a single, global
conceptual framework. The two basic MA conceptual framework diagrams (Figures
3.1 and 3.2) are commonly accompanied by a third one (see Figure 3.8), which
shows how the framework has been adapted in order to make sense to a group
of indigenous people in Peru. The juxtaposition of these three diagrams gives the
impression that the main MA figures represent, in graphic form, a single global
and scientific knowledge system, while Figure 3.8 represents one of many local and
indigenous knowledge systems. This figure does show one attempt to adapt a conceptual framework to local conditions, but it does not provide any reason to believe
that the result always should look like this.
Participants in the process of developing and debating the MA’s conceptual
framework know that it is essentially a work of compromise, a negotiated agreement to use certain words and phrases in preference to others and to adopt some
broad definitions and assumptions that constitute the lowest common denominator
of academic debate about the relationship between “people” and “nature,” between
“the economy” and “the environment,” or between “human well-being” and “ecosystems.” Without this common language, an assessment team would not be able to
produce a coherent and persuasive assessment of what is happening to these relationships. And when a team is faced with the task of rephrasing or translating this
common language for use in a social, cultural, and political environment, they need
to think through how words and phrases such as “biodiversity,” “ecosystem services,” and “human well-being” are treated within the MA’s conceptual framework
and consider whether adaptations are necessary.
Figure 3.8. Local
of the MA
for the Peru
Source: MA
100 | A Manual for Assessment Practitioners
This section includes a sample of some contentious issues that may merit particular attention by an assessment team that is working with assessment users to adapt
or create a conceptual framework. It is worth taking some time to explore these
questions (and delving into others as well) to ensure that the conceptual framework
develops in a way that not only is useful to the stakeholders but also engages them
in the assessment process from its outset. Doing so not only should help lend insight
into the goals and origins of the MA. It also may help the assessment team clarify
aspects that fit the circumstances and key elements that may need to be adjusted to
reflect local conditions or gaps that need to be filled.
For example, the definition and measurement of human well-being is an intensely political issue. What types of people and human behavior are most important in the assessment being undertaken? How do different peoples’ value systems
affect the depiction of human well-being in the conceptual framework? How can
different peoples’ knowledge systems be included? How does biodiversity figure in
these considerations of human well-being?
In many countries today, the central challenge for ecosystem management and
human well-being is the relationship between poverty (or inequality) and the environment. How will these relationships be handled in the conceptual framework? Is
it important to allow for the sort of poverty–environment relationship described in
the Brundtland Report, where “poor people are forced to overuse environmental
resources to survive from day to day, and their impoverishment of their environment
further impoverishes them, making their survival ever more difficult and uncertain”
(WCED 1987:27)? Of course, affluence may also have this kind of influence. In either case, there will be many other things that cause rich people or poor people to
behave in certain ways that may need to be reflected in the conceptual framework.
This web of relationships between ecosystem services and human well-being
deserves careful thought. In the basic MA conceptual framework (Figure 3.1), an
arrow leads upward from the box containing ecosystem services to the box containing human well-being. But unlike all the other arrows in this diagram, it looks to be
immune from “strategies and interventions.” So why might this control be limited
or entirely absent in the one relationship that is the primary focus of an ecosystem
assessment? One answer might be that human and environmental well-being are so
intimately connected that it makes no sense to conceive of them as separate things.
That answer might well appeal to the owners of an indigenous knowledge system of
the kind represented in Figure 3.8.
Consider also the appropriateness of the impacts that are indicated by arrows in
the two basic MA conceptual framework figures (Figures 3.1 and 3.2). Why might
the only arrow to escape from the box of human well-being be one that leads to
indirect drivers, and why is this arrow subject to political control? It may be fair
enough to argue that the supply of ecosystem services, or the social effects of environmental change, can only be modified by something that counts as a driver. But is
the poverty–environment relationship just another version of the basic relationship
between ecosystem services and human well-being, as depicted in the MA figures?
Or does it need to be depicted in a different way in the conceptual framework being
prepared? Which of these relationships are most important in this situation? How
would different groups engaged in the assessment depict this fundamental relationship between people and nature? Are there different perspectives that need to be
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 101
3.4 How to use conceptual frameworks to engage users
Section’s take-home messages
• Assessment processes are inherently political, involving disparities in power and authority
and conflicting interests.
• Various pragmatic approaches vis-à-vis the global conceptual framework are valid—from
adaptation to independence and including the use of multiple frameworks.
• The people who variously are (or are not) informed, consulted, and involved in the
creation of the conceptual framework for ecosystem assessment and the ways in which
their knowledge and expertise are valued (or not) will in many ways govern the entire
assessment process.
• There will be significant disparities in power and authority among groups that have
important stakes in the human–environment interactions at the heart of an ecosystem
Conceptual frameworks are used by their proponents in several different ways, and
these can be mapped against the top-down to bottom-up spectrum of ways to go
about framing an ecosystem assessment. Some uses are clearly instrumental: a conceptual framework can be a means for a group of assessors and users to agree on
basic understandings of what features of a system are important to assess and how
those features are related. Or it can be used as a one-way communication tool or
device to persuade potential contributors to, and users of, an assessment to adopt a
certain problem-framing favored by the proponents. Other uses fall somewhere in
between, being there to inform but still open for others to question, challenge, and
revise. Finally, some are really intended to be a tool for deliberation and negotiation
and can be revised, fragmented, and discarded depending on the arguments made by
the stakeholders involved.
3.4.1 Crossing boundaries between different forms of knowledge
Large international assessment efforts such as the IPCC, the MA, and the IAASTD
are examples of a broad class of “boundary crossing” organizations that straddle
the shifting divide between politics and science and are intended to effectively bridge
boundaries between science, policy arenas, and broader society while maintaining
the scientific credibility, political legitimacy, and social usefulness identified as fundamental principles in Chapter 2.
One of the practical lessons distilled from comparative studies of boundary
crossing organizations concerns the production of “boundary objects,” which can
be maps, sketches, diagrams, historical narratives, explanatory models, and conceptual frameworks, among other things. Production of boundary objects means
that scientists, policy makers, local people, civil society organizations, and other
stakeholders representing different types of knowledge work together to create and
use these objects; the exact form of the object is less important than the interactive
process that creates it (Cash et al. 2003, Clark and Holliday 2006, Guston 2001).
The conceptual framework for assessment has value as a boundary object since it
102 | A Manual for Assessment Practitioners
can become a focal point for building effective interactions between the assessment
team and stakeholders.
Studies of this boundary-crossing process have shed light on the ways “knowledge” and “action” coevolve in any multistakeholder, multidisciplinary, or multiinstitutional group and how knowledge is used as the “rationale” to present actions
that serve a purpose for the group. This is not easy work. A skeptical stance toward
existing rationalizations is needed in order to help multiple stakeholders find productive ways out of current confusion and conflicts. Some of the existing conflicts
will be based on real trade-offs and conflicting objectives; others may be at least
partially linked to stereotypes, myths, and misunderstandings. For the first group of
conflicts, a reorientation of objectives through negotiation and mediation probably
is the best way to make progress; for the second, knowledge-based assessments can
be of direct help in improving understanding.
3.4.2 A gallery of approaches
How can conceptual frameworks be made more “user-friendly” for local people?
Some of the keys lie in the principles of stakeholder engagement laid out in Chapters
1 and 2 and elaborated here in section 3.2. In light of the foregoing, and based on
the experiences of the sub-global assessments undertaken between 2002 and 2005
as part of the MA, it is not altogether surprising that the MA conceptual framework
tended to be more readily applied at coarser than finer scales due to the difficulty
of capturing local-scale interactions within the framework and the need to include
stakeholder perspectives based on alternative knowledge systems and world views.
Capturing these multidimensional perspectives required considerable time and the
use of innovative participatory methods (MA 2005b:72–73).
The South Africa Regional Assessment (discussed in section 3.2) and the Tropical Forest Margins Assessment (discussed here) used one or more frameworks in addition to the MA framework. In a similar vein, as described in this section, the MA
framework provided an important entry point for launching new assessments, such
as the Northern Range and Caribbean Seas assessments. Other community-based or
national assessments, such as those in Japan, Peru, and Costa Rica, have adapted the
MA conceptual framework through reframing and reinterpretation in terms consistent with their communities’ worldviews—often producing a radical departure from
the original framework. Still others, such as Papua New Guinea (PNG), have chosen
to go their own way to develop frameworks that are essentially independent from
the MA version.
Tropical forest margins: Using multiple frameworks
The Partnership for the Tropical Forest Margins (known as ASB, from its former
name, the Alternatives to Slash-and-Burn Program) adapted the MA conceptual
framework but also continued to use its own framework for integrated natural
resource management (iNRM). Figure 3.9 represents the main components of the
iNRM approach and the need to ensure that work on the different components is
not undertaken in isolation. The “solutions” arrived at in such an approach do not
“maximize” any one element or factor in the agroecosystem; they are options for
optimizing trade-offs among the different elements.
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 103
Policy Implementation
Figure 3.9. The main components of the iNRM approach.
Source: developed by Anne-Marie Izac
The concept of integrated natural resources management was developed in the
context of the research undertaken by the international agricultural research centers.
The kind of research conducted there evolved over time. Initially, crop improvement
research (breeding “improved” germplasm to increase crop productivity) was the
raison d’etre of the centers. As productivity gains occurred, scientists observed that
various constraints—from inappropriate institutions and policies to limited natural
resources and human resources—had to be factored in if further gains in overall
production were to take place. Research on natural resources management, aimed
at maintaining the properties if not the integrity of the natural resource base of agricultural systems, thus became accepted as a necessary complement to crop improvement research.
Integrated natural resources management emphasizes that the two complementary types of research need to be implemented in an integrated manner rather than
simply in parallel. At the core of the iNRM approach lies the realization that there
are trade-offs among different ecosystem services and among different objectives.
The reductionist approach to science that is embedded in “maximizing” productivity (or profits, or water availability, or nutrient release) leads scientists to provide
“solutions” that do not resolve the complex problems of agroecosystems and natural resource management because they do not address them.
104 | A Manual for Assessment Practitioners
ASB was particularly well positioned to adapt many of the elements of the MA
conceptual framework that were directly useful while also maintaining the iNRM
framework that had played an important role in building shared understanding of
the program. The iNRM framework and the MA framework have many similarities
among their components: driving forces, conditions and trends in resources, poverty
(human well-being), and interventions (responses). Moreover, both approaches recognize that ecosystem services are important locally and globally and that policies
and actions at remote (national, regional, and global) scales can affect local actions
and conditions (and vice versa). But use of the MA framework enabled the ASB assessment to move to new levels by helping make more specific links between ecosystem services and human well-being.
Uniquely, ASB is a sub-global assessment on tropical forest margins that cuts
across the humid tropics, so it includes a (nearly) global scale (see Tomich and Palm
2005 for additional information). This made it possible to see more clearly that
there can be a disconnection between the provisioning and regulating functions of
local ecosystems, particularly when the main drivers are phenomena (migration,
globalization, poverty, food insecurity) that cut across scales from local to global. In
such cases, community stewardship of services may not be sufficient to sustain them.
And changes in some ecosystem services (e.g., carbon storage) may be felt globally,
but the global populations with a stake in these services may be unaware or uninterested in the local conditions and decisions that drive these changes.
The lesson learned in this case is that sometimes a pre-existing conceptual framework can provide the glue for collaboration between the assessment team and stakeholders, while adaptation of the MA framework enhances capabilities for synthesis
across broader spatial scales.
The Northern Range and Caribbean Sea: Finding the right entry points
The Northern Range and Caribbean Sea Assessments—undertaken concurrently, and
jointly led by The Cropper Foundation and the University of the West Indies—generated different experiences in the application of the MA conceptual framework.
In the case of the Caribbean Sea, the authors involved in the assessment agreed
that it was most meaningful to use the two priority ecosystem services provided by
the sea—fisheries and tourism—as a starting point for the assessment. This decision
arose from the realization that the capacity, time, and financial resources available
would not allow for an assessment of all ecosystem services, driving forces, and
response options required at the regional, subregional, and national levels. Further,
it was agreed that because of the economic importance of fisheries and tourism to
the region, the focus for formulating response options for the Caribbean Sea should
be placed on these services. Having established this basis, the Caribbean Sea Assessment then proceeded to apply all components of the MA conceptual framework,
using the ecosystem services box as a starting point.
In the case of the Northern Range Assessment, however, the mountain ecosystem was first subdivided into three subsystems—forest ecosystems, freshwater ecosystems, and coastal ecosystems—because it was felt that this would allow for the
most effective packaging and thus assessment of the information. Because of its
cross-cutting nature, biodiversity was treated separately but in the same manner
as a subsystem. For each of the subsystems, the authors then identified all relevant
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 105
components of the MA conceptual framework—the ecosystem services, links to human well-being, driving forces, and response options—that were applicable, and
these were all assessed. Having applied this approach, the authors were more effectively able to make the linkages across the subsystems, especially related to driving
forces and responses. Because of this they were able to focus a very complex and
integrated set of issues in a clear, discretely laid out set of response options. When
juxtaposed with the Caribbean Sea Assessment, the Northern Range Assessment
used a more system-based entry point into the MA Framework.
Comparing these two experiences, it is evident that—depending on needs and
issues—it is possible to apply the MA framework in different ways to answer the
ultimate question: how should we respond? It is therefore extremely important that
assessment practitioners gain a clear understanding of needs very early in the assessment process in order to steer the assessment methodology, including application of
the conceptual framework, in the right direction.
Japan: Framing cultural concepts of ecosystems
The Japan sub-global assessment is the first of its kind in the country. It uses two
terms at the core of its conceptual framework that are difficult to translate from the
Japanese: satoyama (a traditional rural landscape) and satoumi (marine and coastal
ecosystems with human interaction). The Japan sub-global assessment is intended to
follow and apply the framework of the MA sub-global assessments by identifying the
condition and trends of biodiversity, ecosystem services, and human well-being; examining the drivers of changes in these components; assessing the responses; developing
conceptual models linking condition and trends, drivers, and responses; and building
scenarios with a local and/or national focus on the issues of satoyama and satoumi.
While the assessment aims to provide inputs into the tenth Conference of the
Parties (COP-10) of the Convention on Biological Diversity in Nagoya, Japan, in
2010, it seeks to strengthen the scientific basis for action needed to conserve and
sustain satoyama and satoumi and to enhance human well-being. Although the assessment has been planned and developed since late 2006, the team in Japan spent
more than a year explaining what the MA was, consulting with stakeholders regarding the planned assessment, and establishing the governance structure in a situation in which few knew about the MA or the conceptual framework it used. The
development of a conceptual framework adapted to the Japan assessment still is in
process, and it has taken several steps to create the platform where various types
of stakeholders—users as well as experts and scientists from different disciplines—
interact and share ideas.
The term satoyama encompasses several types of ecosystems, including secondary forests, agricultural land, ponds, grasslands, and settlements, and accounts for
approximately 40% of the land area of Japan. Satoyama also is a way of life and a
perspective on interactions between humans and the environment. More recently,
this concept of interaction in satoyama has been expanded to include marine and
coastal ecosystems, giving rise to the idea of satoumi. Although there still is some
debate about this, the mosaic patterns of ecosystems appear to maintain a high level
of biodiversity. While the government of Japan has become committed to promoting the term satoyama to the international community, especially at COP-10, there
is no common definition of satoyama and there also are various similar phrases and
106 | A Manual for Assessment Practitioners
terminologies in Japanese. This is because numerous groups and individuals have
attempted to define satoyama from their own background and interests, and some
refer to it as secondary forests.
The assessment is trying to look into various definitions that include different
types of ecosystems and different area categories as well as historical narrative contexts of satoyama and satoumi to identify the unit and the scale of mosaic-type
systems with human interaction and also to determine the linkages to nature and to
urban areas, given that satoyama usually exist between primary forests and cities.
In addition, one of the major issues concerning satoyama and satoumi in Japan is
their underuse (versus an overuse of many ecosystems or their services) through aging and the loss of population in many localities. This is expected to provide some
views on pressure on ecosystems that would be different from most assessments
in developing countries—in particular in terms of pressures on mountain and forest ecosystems—and to suggest the extent to which humans should intervene to
maintain the ecosystems that have been abandoned and, as a result, are undergoing
natural succession.
Although these are distinctively Japanese terms and they are difficult to translate in ways that do justice to their full meaning, these types of multi-ecosystem
landscapes are not unique to Japan. Such complex landscape mosaics are found
throughout the world. Top-down global approaches consistently miss these important integrative elements at the landscape scale. The Japanese sub-global assessment
may produce insights that could be useful to other assessment teams.
Peru: Developing locally relevant conceptual frameworks
Indigenous communities conducting an assessment in the Vilcanota region of Peru
created a conceptual framework using the Quechua understanding of ecological and
social relationships (see Figure 3.8 in section 3.3.6). Within the Quechua vision of
the cosmos, concepts such as reciprocity (Ayni), the inseparability of space and time,
and the cyclical nature of all processes (Pachakuti) are important components of the
Inca definition of ecosystems. Love (Munay) and working (Llankay) bring humans
to a higher state of knowledge (Yachay) about their surroundings and are therefore
key concepts linking Quechua communities to the natural world. Ayllu represents
the governing institutions that regulate interactions between all living beings. These
concepts have existed since the time of the Incas and are deeply ingrained within
Quechuan communities.
The communities, and particularly those leading the assessment, felt that the
assessment work would have less relevance if the MA conceptual framework were
used as a base for the assessment process. The framework that was developed and
used has similarities with the MA’s, but the divergent features were considered important to the Quechua people conducting the assessment. The various similarities
and differences between the conceptual frameworks were discussed to build understanding of the broader spectrum of perceptions about the relationship between
ecosystem services, human well-being, and drivers of change.
The Vilcanota conceptual framework also includes multiple scales (Kaypacha,
Hananpacha, Ukupacha); however, these represent both spatial scales and the cyclical relationship between the past, present, and future. Inherent in this concept of
space and time is the adaptive capacity of the Quechua people, who welcome change
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 107
and have developed resilience through an adaptive learning process. (It is recognized
that current rates of change may prove challenging to the adaptive capacities of the
communities.) The cross shape of the Vilcanota framework diagram represents the
Chakana, the most recognized and sacred shape to Quechua people, and orders
the world through deliberative and collective decision making that emphasizes reciprocity. Pachamama is similar to a combination of the “ecosystem goods and services” and “human well-being” components of the MA framework. Pachakuti is
similar to the MA drivers (both direct and indirect). Ayllu (and Munay, Yachay,
and Llankay) may be seen as responses and are more organically integrated into the
cyclic process of change and adaptation.
In the Vilcanota assessment, the Quechuan communities directed their work
process to assess the conditions and trends of certain aspects of the Pachamama
(focusing on water, soil, and agrobiodiversity), how availabilities of these goods and
services are changing, the reasons behind the changes, the effects on the other elements of the Pachamama, how the communities have adapted and are adapting to
the changes, and the state of resilience of the Quechua principles and institutions for
dealing with these changes in the future.
Developing the local conceptual framework from a base of local concepts and
principles, as opposed to simply translating the MA framework into local terms,
allowed local communities to take ownership of the assessment process and gave
them the power both to assess the local environment using their own knowledge
and principles of well-being and to seek responses to problems within their own
cultural and spiritual institutions. Strong involvement from the Quechua community in the process produced a framework that reflects local views on the place of
humans within the natural world but that also has many similarities with the MA
framework. The similarities were important to discuss in order to build a process
that matched the goals of the MA.
Costa Rica: Adaptation of local ecological knowledge for resource management
The Asociacion Ixacavaa de Desarrollo y Informacion Indıgena began the Bajo
Chirripo assessment with the idea of developing a management plan for the community’s resources. Through discussions and meetings with community members,
it soon became apparent that in the past a strict “management plan” had existed
and was based on norms and beliefs regarding interactions between humans and
their environment. The concept of reciprocity between humans and the rest of the
environment was key. The erosion of traditional knowledge and of the norms and
beliefs that have guided how the Cabecar have traditionally managed their resources
became the major impetus for developing a local, participatory assessment process.
Ixacavaa therefore focused their ecosystem assessment process on the recovery
of lost knowledge that in the past safeguarded the integrity of the environment and
ensured the sustainability of human activities. Because of this, it was important to
make sure that the conceptual framework used was one that fit this goal and emphasized the importance of local knowledge. The discussion and development of a
conceptual framework was also an opportunity to initiate discussions within the
community about lost knowledge and how to recuperate it.
In Bajo Chirripo, the basic concepts of the MA conceptual framework were
understood to some extent, but it was nevertheless quite foreign to community
108 | A Manual for Assessment Practitioners
members’ way of thinking. Recognizing a need to understand these concepts in the
language and from the perspective of the community, the assessment team and the
Cabecar community invested considerable time in revising the conceptual framework. A full local framework that could be used in assessment was not completed,
but components were identified and discussed in the community.
One of the components identified as central to ecosystem management is the
Cabecar conical shaped house, which represents the natural world. The flip side
(also a conical house) is the spirit world and is equally important. Communities had
noticed a big decline in the number of animals and the quantity of important natural
resources available to them in their territory. They explain this by saying that the
animals have left the natural world and are hiding in the spirit world. When humans
begin to act more responsibly and with greater reciprocity, it is believed that the
animals will return to the natural world. Because of the discussion of these ideas in
the process of the assessment, community members in one village decided to build
a conical house (which can also be physically constructed on earth and becomes a
spiritual icon to the communities), which has now been completed.
Papua New Guinea: An independent approach
Papua New Guinea is a nation of indigenous peoples, with several hundred languages and cultures, who generally like to think of themselves as “customary landowners.” Members of this society are not very keen on flow charts as a way to think
about environmental issues, and although there are many different forms of traditional environmental knowledge, they do not seem to constitute “systems” of the
sort portrayed in Figure 3.8. In this context, it was felt that the best way to adapt the
MA conceptual framework was to use the simplest of all diagrammatic forms, the
four-cell matrix, as a way to explore the difference between scientific and indigenous
perspectives on the definition of crucial terms in the environmental policy process.
The first home-grown diagram in the conceptual framework chapter of the PNG
assessment therefore offers four ways of defining an ecosystem (see Figure 3.10).
It recognizes the ambiguity already present in the scientific perspective of the MA
conceptual framework and then suggests a parallel ambiguity in the “political” perspective that would make sense to a nation of customary landowners who have
good reason to consider the territorial domain of a traditional political community
as an ecosystem in its own right, one that is internally divided by landscape elements
that are normally distributed in equal measure between smaller groups within the
Figure 3.10. Defining an ecosystem.
Source: PNG sub-global assessment (unpublished)
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 109
The same logic is later applied to the identification of significant species within
an ecosystem in order to get a handle on the evident gap between the way in which
scientists and local people think about biodiversity or “life on Earth” (see Figure
3.11). Biologists distinguish between an endemic species, which is unique to a certain type of ecosystem and whose survival therefore depends on the survival of that
ecosystem, and a keystone species, which makes a unique contribution to the survival of a certain type of ecosystem, even if it is not endemic (Mills et al. 1993). A
keynote species, by contrast, is one whose services are essential to the survival of a
specific form of traditional or indigenous culture, while a totemic species (at least in
the Melanesian context) is one whose services are recognized in magic and mythology and hence in the value that local people attribute to its reproduction.
Figure 3.11. Identifying significant species within an ecosystem.
Source: PNG sub-global assessment (unpublished)
Another four-cell matrix is applied to the analysis of values, drivers, and responses (see Figure 3.12), establishing a contrast between the production systems
and management regimes that operate at a community scale (where local people
define ecosystems as territorial domains) and those that operate at a wider social
Figure 3.12. Analyzing values, drivers and responses.
Source: PNG sub-global assessment (unpublished)
This way of thinking about the world has obvious drawbacks if it leads people
to put things into one of four pigeonholes to which they do not really belong. But
that is not the point. In the last case the link made between indigenous production systems and traditional territorial domains allows a number of useful insights.
First, it does away with the less helpful distinction often made between the cash and
subsistence sectors of a national economy in the valuation of ecosystem services,
for many of the products of indigenous production systems are actually sold in domestic and international markets. Second, indigenous knowledge can be thought of
as a set of loosely connected practical understandings embedded in different forms
of production rather than a picture of the world as a whole, and this does seem to
make more sense of the way that Melanesians (and many other people) deal with
their natural environment. Third, the innovative and adaptive nature of what might
otherwise be called “traditional” production systems is clearly recognizable, avoiding the romantic and misguided idea that customary landowners simply follow in
110 | A Manual for Assessment Practitioners
the footsteps of their ancestors. Although it is then possible to ask how indigenous
and industrial production systems may compete or collaborate in specific economic
sectors (forestry or agriculture, for example), this still allows recognition of the existence of production systems (like dive tourism or illegal fuelwood harvesting) that
occupy an informal or intermediate space between these two categories.
The PNG assessment is a reminder that conceptual frameworks, knowledge systems, and computer programs are all built out of basic binary distinctions. Yet there
is no reason to assume that distinctions that make sense in one context or to one
group of people can be absorbed into a conceptual framework that makes sense in
all contexts and to all possible users of an ecosystem assessment. The best way to
test the validity and usefulness of any conceptual framework is to check the significance of such distinctions to the people who “relate to nature” or “use ecosystem
services” in the place where an assessment is conducted.
Arrow, K. J., P. Dasgupta, and K.-G. Maler. 2004. Evaluating projects and assessing sustainable
development in imperfect economies. In The economics of non-convex ecosystems., ed. P. Dasgupta and K.-G. Maler. Dordrecht, the Netherlands: Kluwer Academic Publishers.
Baumol, W. J., and W. E. Oates. 1988. The theory of environmental policy. Cambridge, U.K.:
Cambridge University Press.
Bell, S., and S. Morse. 1999. Sustainability indicators: Measuring the immeasurable. London: Earthscan Publications.
Bell, S., and S. Morse. 2003. Measuring sustainability: learning by doing. London: Earthscan
Berger, A. R., and R. A. Hodge. 1998. Natural change in the environment: A challenge to the
pressure-state-response concept. Social Indicators Research 44 (2): 255–65.
Carney D. 1998. Sustainable rural livelihoods: what contribution can we make? DfID, London
Cash, D.W. W.C. Clark, F. Alcock, N.M. Dickson, N. Eckley, D. Guston, J. Jäger, and R. Mitchell.
2003. Knowledge systems for sustainable development. Proceedings of the National Academy
of Sciences of the United States of America 100: 8086–8091.
Chapin, F. S. III, E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L. Reynolds, D. U.
Hoope et al. 2000. Consequences of changing biodiversity. Nature 405:234–42.
Clark, W.C., and L. Holliday. 2003. Linking Knowledge with Action for Sustainable Development.
National Research Council of the National Academies: Washington, D.C.
Dore, J. 2007. Multi-stakeholder platforms (MSPS): unfulfilled potential. In Democratizing water
governance in the Mekong region, ed. L. Lebel, J. Dore, R. Daniel, and Y. Koma, 197–226.
Chiang Mai, Thailand: Mekong Press.
Folke, C., S. Carpenter, T. Elmquist, L. Gunderson, C. S. Holling, and B. Walker. 2002. Resilience
and sustainable development: Building adaptive capacity in a world of transformations. Ambio
31 (5): 437–40.
Geist, H. J. 2005. The causes and progression of desertification. Ashgate Studies in Environmental
Policy and Practice. Aldershot, U.K.: Ashgate.
Geist, H. J. 2006. Change or collapse? A theoretical approach to global environmental change
and land use in rainforest and arid zone hotspots. (In German.) Geographische Zeitschrift 94
Geist, H. J. and E. F. Lambin. 2002. Proximate causes and underlying driving forces of tropical
deforestation. BioScience 52 (2): 143–49.
Geist, H. J. and E. F. Lambin. 2004. Dynamic causal patterns of desertification. BioScience 54 (9):
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 111
Geist, H. J., E. Lambin, C. Palm, and T. Tomich. 2006. Agricultural transitions at dryland and tropical forest margins: Actors, scales and trade-offs. In Agriculture and climate beyond 2015: A new
perspective on future land use patterns, ed. F. Brouwer and B. A. McCarl, 53–73. Dordrecht,
Netherlands: Springer-Verlag.
Gunderson, L. H., and C. S. Holling. 2002. Panarchy: Understanding transformations in human
and natural systems. Washington, DC: Island Press.
Guston, D. 2001. Boundary organizations in environmental policy and science: an introduction.
Science, Technology and Human Values 26: 399–408.
Heinz Center. 2006. Filling the gaps: priority data needs and key management challenges for national reporting on ecosystem condition. Washington, DC: Heinz Center for Science, Economics
and the Environment.
IPCC (Intergovernmental Panel on Climate Change). 2007. Climate change 2007—IPCC fourth
assessment report. Cambridge, U.K.: Cambridge University Press.
Jerneck, A., and L. Olsson. 2007. More than Trees: Contextualising agro-forestry and identifying
opportunities in subsistence farming. Internal report to World Agroforestry Centre, December.
Kasemir, B., M. B. A. Asselt, and G. Durrenberger. 1999. Integrated assessment of sustainable
development: Multiple perspectives in interaction. International Journal of Environment and
Pollution 11 (4): 407–25.
Kuhn, T. 1962. The structure of scientific revolutions. Chicago: University of Chicago Press.
Lambin, E. F., H. J. Geist, and E. Lepers. 2003. Dynamics of land use and cover change in tropical
and subtropical regions. Annual Review of Environment and Resources 28:205–41.
Lambin, E. F., H. J. Geist, and R. R. Rindfuss. 2006. Introduction: Local processes with global impacts. In Land-use and land-cover change: Local processes and global impacts. Global Change—
The IGBP Series, ed. E. F. Lambin and H. J. Geist, 1–8. Berlin: Springer-Verlag.
Lambin, E. F., H. J. Geist, J. F. Reynolds, and D. M. Stafford Smith. 2007. Integrated human–
environment approaches of land degradation in drylands. In Sustainability or collapse? An integrated history and future of people on earth. Dahlem Workshop Series 96, ed. R, Costanza, L.
J. Graumlich, and W. Steffen, 331–39. Cambridge, MA: The MIT Press.
Lebel, L. 2006. The politics of scale in environmental assessment. In Bridging Scales and Knowledge
Systems: Concepts and Applications in Ecosystem Assessment, ed. W. V. Reid, F. Berkes, T. J.
Wilbanks, and D. Capistrano, 37–57. Washington, DC: Island Press.
Lebel, L., and P. Garden. 2008. Deliberation, negotiation and scale in the governance of water resources in the Mekong region. In Adaptive and integrated water management: coping with complexity and uncertainty, ed. C. Pahl-Wostl, P. Kabat, and J. Möltgen, 205–25. Berlin: Springer.
Lebel, L., R. Daniel, N. Badenoch, and P. Garden. 2008. A multi-level perspective on conserving
with communities: Experiences from upper tributary watersheds in montane mainland southeast
Asia. International Journal of the Commons 1:127–54.
Lee, K. N. 1999. Appraising adaptive management. Conservation Ecology 3 (2): 3. [online] URL:
Ludwig, D. B., B. Walker, and C. S. Holling. 1997. Sustainability, stability, and resilience. Conservation Ecology 1 (1): 7.
Machlin, G. E., and J. E. McKendry. 2005. The human ecosystem as an organizing concept in ecosystem restoration. HESG Publications 05-01. Moscow, ID: University of Idaho.
Meter, K. 1999. Neighborhood sustainability indicators guidebook. Minneapolis, MN: Crossroads
Resource Center.
Michon, G., H. de Foresta, P. Levang, and F. Verdeaux. 2007. Domestic forests: a new paradigm for
integrating local communities’ forestry into tropical forest science. Ecology and Society 12(2): 1.
[online] URL:
MA (Millennium Ecosystem Assessment). 2003. Ecosystems and human well-being: A framework
for assessment. Washington, DC: Island Press.
MA. 2005a. Ecosystems and human well-being, Vol. 1: Current state and trends. Washington, DC:
Island Press.
112 | A Manual for Assessment Practitioners
MA. 2005b. Ecosystems and human well-being, Vol. 4: Multiscale assessments. Washington, DC:
Island Press.
Mills, L. S., M. E. Soule, and D. F. Doak. 1993. The keystone-species concept in ecology and conservation. BioScience 43 (4): 219–24.
Nelson, G. C., E. Bennett, A. A. Berhe, K. Cassman, R. DeFries, T. Dietz, A. Dobermann et al.
2006. Anthropogenic drivers of ecosystem change: An overview. Ecology and Society 11 (2):
online journal.
Ness, B., S. Anderberg, and L. Olsson. 2008. Structuring problems of unsustainability in the Swedish sugar sector with the multilevel DPSIR framework. Sustainability Science. Accepted for
Newell, B. C., C. L. Crumley, N. Hassan, E. F. Lambin, C. Pahl-Wostl, A. Underdal, and R. Wasson.
2005. A conceptual template for human–environment research. Global Environmental Change
15 (4): 299–307.
Pirrone, N., G. Trombino, S.Cinnirella, A. Algieri, G. Bendoricchio, and L. Palmeri. 2005. The
Driver-Pressure-State-Impact-Response (DPSIR) approach for integrated catchment–coastal
zone management: preliminary application to the Po catchment, Adriatic Sea coastal zone system. Regional Environmental Change 5: 111–137.
Reid, R. S., T. P. Tomich, J. C. Xu, and A. S. Mather. 2006. Linking land-change science and policy:
Current lessons and future integration. In Land-use and land-cover change: Local processes and
global impacts, ed. E. F. Lambin and H. J. Geist, 157–71. Berlin: Springer.
Reid, W., et al. 2006. Bridging Scales and Knowledge Systems: Linking Global Science and Local
Knowledge in Assessments. Washington, D.C.: Island Press.
Resilience Alliance and Sante Fe Institute. 2008. Thresholds and regime shifts in ecological and
social-ecological systems. Resilience Alliance.
Reynolds, J. F., and D. M. Stafford Smith, eds. 2002. Global desertification: Do humans cause
deserts. Dahlem Workshop Series 88. Berlin: Dahlem Unversity Press.
Reynolds, J. F., D. M. Stafford Smith, E. F. Lambin, B. L. Turner II, M. Mortimore, S. P. J. Batterbury, T. E. Downing et al. 2007. Global desertification: Building a science for dryland development. Science 316 (5826): 847–51.
Schellnhuber, H.-J., M. K. B. Lüdeke, and G. Petschel-Held. 2002. The syndromes approach to
scaling/describing global change on an intermediate functional scale. Integrated Assessment 3
(2/3): 201–19.
Senge, P., R. Ross, C. Roberts, B. Smith, and A. Kleiner. 1994. The fifth discipline fieldbook: Strategies and tools for building a learning organization. London: Nicholas Brealy.
Simon, H. A. 1962. The architecture of complexity. Proceedings of the American Philosophical
Society 106 (6): 467–82.
Smeets E., R. Weterings. 1999. Environmental indicators: typology and overview. Technical Report 25, European. Environmental Agency, Copenhagen, Denmark.
Svarstad, H., L. Kjerulf Petersen, D. Rothman, H. Siepel, and F. Watzold. 2008. Discursive biases of
the environmental research framework DPSIR. Land Use Policy 25: 116–125.
Spreng, D., and A. Wils. 1996. Indicators of sustainability: Indicators in various scientific disciplines. Multidimensional approaches to sustainability: The framing project. Alliance for Global
Sustainability: 1–27.
Stern, D. 1995. The capital theory approach to sustainability: A critical appraisal. Boston University
Center for Energy and Environmental Studies Working Papers Series No. 9501: 1–44.
Tomich, T. P., and C. A. Palm et al. 2005. Forest and agroecosystem tradeoffs in the humid tropics: A crosscutting assessment by the Alternatives to Slash-and-Burn consortium conducted as
a sub-global component of the Millennium Ecosystem Assessment. Nairobi: Alternatives to
Slash-and-Burn Programme.
Tomich, T. P., D. Timmer, J. C. Alegre, V. Areskoug, D. Cash, A. Cattaneo, P. Ericksen et al. 2007.
Integrative science in practice: Process perspectives from ASB, the Partnership for the Tropical
Forest Margins. Agriculture Ecosystems and Environment 121 (3): 269–86.
Conceptual Frameworks for Ecosystem Assessment: Their Development, Ownership, and Use | 113
UNEP (U.N. Environment Programme). 2004. Synthesis of responses on strengthening the scientific base of the United Nations Environment Programme, Report by the Executive Director.
UNEP. 2007. Global Environment Outlook–4. Nairobi: UNEP.
van Noordwijk, M., D.A. Suyamto, B. Lusiana, A. Ekadinata, and K. Hairiah. 2008. Facilitating
agroforestation of landscapes for sustainable benefits: tradeoffs between carbon stocks and local
development benefits in Indonesia according to the FALLOW model. Agriculture Ecosystems
and Environment 126: 98–112.
Victor, P. A. 1991. Indicators of sustainable development: Some lessons from capital theory. Ecological Economics 4:19l–213.
Walters, C., and C. S. Holling. 1990. Large-scale management experiments and learning by doing.
Ecology 71 (6): 2060–68.
Waltner-Toews, D., and J. Kay. 2005. The evolution of an ecosystem approach: The diamond schematic and an adaptive methodology for ecosystem sustainability and health. Ecology and Society 10 (1): 38.
Warner, J. F. 2006. More sustainable participation? Multi-stakeholder platforms for integrated
catchment management. Water Resources Development 22:15–35.
WCED (World Commission on Environment and Development). 1987. Our Common Future. Oxford: Oxford University Press.
Assessing State and Trends in Ecosystem
Services and Human Well-being
Robert Scholes, Reinette Biggs, Cheryl Palm, and
Anantha Duraiappah
What is this chapter about?
This chapter provides practical guidance on gathering, evaluating, and presenting information
related to the supply and consumption of ecosystem services and the status of human wellbeing, especially to the extent that it depends on ecosystem services. The chapter suggests
the types of measurements that can be used as indicators of both ecosystem services and
human well-being. It shows how information from a variety of sources can be integrated to
provide coherent and robust insights into trends in the adequacy and security of ecosystem
service provision. It suggests ways of communicating the information in an effective and responsible manner, including where information is incomplete or uncertain.
4.1 How to set the scope of a condition and trend assessment
Section’s take-home messages
• It is assumed that the geographical scope and content of an assessment have already
been broadly determined through the stakeholder engagement processes.
• One part of the scope-setting exercise that is particular to the condition and trend
assessment is agreeing on the time period for assessing conditions and for analyzing
trends. It is helpful to think of the period for assessing trends as “the relevant past to the
predictable future.”
The “state” of human well-being or of ecosystem services is a snapshot of its condition in a given area and at a particular time, usually the present or recent past. State is
not synonymous with “health,” “integrity,” or “degradation” of ecosystems, though
all of those related concepts may be part of an ecosystem condition assessment. State
can be measured in many ways, including, for instance, the yield of a service, the stock
of natural capital that permits that yield, the economic value of either, and various indicators of human well-being. “Trend” is an analysis of the change in state over time.
116 | A Manual for Assessment Practitioners
It is assumed that the geographical scope and content of an assessment have
already been broadly determined through the stakeholder engagement processes described in detail in Chapter 2. An awareness of the technical issues and data implications for condition and trend analysis is a necessary input to the discussion that
leads to this decision, which implies that setting the scope is best done iteratively.
Either the technical experts or the stakeholders can make the first proposal, which
then needs to be debated and modified successively by both users and producers of
the assessment until a workable consensus is reached.
A part of the scope-setting exercise that is particular to the condition and trend
assessment is agreeing on the time period for assessing conditions and for analyzing
trends. A condition assessment applies to a particular time, which is seldom an exact
instant but is a relatively short period. Many statistics are collected on an annual
basis, but either there is considerable interannual variation or the trends are quite
slow, making reporting on longer periods (such as a decade) more useful. Unlike the
objective of trend analysis, the goal is to detect change between periods, not within
them. The window must be long enough to collect reliable data, but not so long that
too much change occurs within it.
It is helpful to think of the period for assessing trends as “the relevant past to the
predictable future.” Ecosystem services typically exhibit some “path dependency”;
in other words, their current levels depend to some degree on their history. But it
is not necessary to push the record as far back as is technically possible unless the
path dependency of the particular services and processes being examined requires
it. A single human generation (three decades) is often enough. It is also untrue that
extrapolation into the future is impossible. Ecosystem service and human well-being
trends are generally quite predictable for a limited period, and all rational managers make such extrapolations. The length of that period varies, depending on the
processes involved—“slow processes” such as the growth of long-lived trees can
be extrapolated for decades, whereas “fast processes” such as deforestation may
only be predictable a few years ahead. For times beyond these limits, fundamental
uncertainties make such predictions indeterminate, and scenario analysis is the appropriate tool (see Chapter 5).
4.2 How to select what to measure
Section’s take-home messages
• An indicator, which is usually quantitative, is a single variable with some logical connection
to the process or object of concern. Compound indices are measures made up of several
different indicators, combined in a particular way to increase their sensitivity, reliability, or
ease of communication.
• Assessments of ecosystem condition rely on sets of indicators. Ideally, individual
indicators should be policy relevant, scientifically sound, simple to calculate and easy
to understand, practical and affordable, sensitive to relevant changes, suitable for
aggregation and disaggregation, and usable for projections of future scenarios.
• The different sources of information for the assessment include peer-reviewed literature,
statistical databases, maps and remotely sensed images, computer models, indigenous
technical and traditional knowledge, and Internet sources.
• Some things do not fit easily into an ecosystem service framework. Biodiversity, for
Assessing State and Trends in Ecosystem Services and Human Well-being | 117
example, is generally not an ecosystem service itself but a necessary condition for such
services to be delivered. There are numerous ways to measure and express biodiversity.
• Since ecosystem services are place and time specific and do not necessarily aggregate
upward in a simple additive way or disaggregate downward by simple proportionality,
care must be taken in relation to scale questions in both time and space. It is important to
work at an appropriate scale for both ecological and human processes.
4.2.1 Metrics, indicators, and indices
A metric is a quantitative measurement. An indicator can be anything that contains
useful information (a bad smell is an indicator of rotten food, for instance), but in
the assessment context it is usually quantitative. Indicators are single variables with
some logical connection to the process or object of concern—they reflect in some unambiguous way its status, causes (drivers), or outcome. It is quite common (but not
essential) to “index” such variables to a particular reference state, for example by
• Setting them to a value of 100 at a particular time, and then expressing all
previous and subsequent values relative to that number;
• Rescaling them between 0 and 1 by subtracting the minimum value from
the observed value and dividing by the difference between the minimum
and maximum value (a variant of this is expressing the observed value as a
percentage of the maximum possible value); or
• Normalizing them by subtracting the value from the mean, and dividing by the
standard deviation.
When indexed in this way, the indicators are nondimensional and can be added
together, for instance in compound indexes. Compound indexes are measures made
up of several different indicators, combined in a particular way to increase their
sensitivity, reliability, or ease of communication. For instance, a geometric mean is
often used rather than a simple arithmetic mean, to prevent very high values in one
component from dominating the calculation.
Compound indexes should be used with great care: they run the risk of canceling out underlying trends when the component indicators change in different directions, and they always have value judgments encoded in their weighting rules, either
explicitly or implicitly. Even “unweighted” indexes suffer this problem, since they
assume that all factors are equally important.
The attributes of good indicators are given in Box 4.1. There are advantages
to adopting indicators that have already been agreed to by a political and technical
process. They can be adopted by themselves or in conjunction with other indicators
as appropriate. For example, a lengthy process of discussion and experimentation
has resulted in agreed lists of key indicators for monitoring biodiversity under the
Convention on Biological Diversity (a summary is given in Table 4.1). Similarly,
for human well-being a framework of eight goals, 18 targets, and 48 indicators
to measure progress toward the Millennium Development Goals was adopted by
a consensus of experts from the United Nations Secretariat and the International
Monetary Fund (IMF), the Organisation for Economic Co-operation and Development (OECD), and the World Bank (see section 4.4).
Box 4.1. Principles for choosing indicators
Individual indicators should ideally be:
1. Policy relevant
Indicators should provide policy-relevant information at a level appropriate for decision making. Where possible, indicators should allow for assessment of changes in ecosystem status
related to baselines and agreed policy targets.
2. Scientifically sound
Indicators should be based on clearly defined, verifiable, and scientifically acceptable data,
collected using standard methods with known accuracy and precision or based on traditional
knowledge that has been validated in an appropriate way.
3. Simple to calculate and easy to understand
Indicators should provide clear, unambiguous information that is easily understood. It is important to jointly involve policy makers, major stakeholders, and experts in selecting or developing
indicators to ensure that the indicators are appropriate and widely accepted.
4. Practical and affordable
Obtaining data on the indicator should be practical and affordable.
5. Sensitive to relevant changes
Indicators should be sensitive and able to detect changes at time frames and spatial scales
that are relevant to the decision making. At the same time, they should be robust to measurement errors or random environmental variability in order to prevent “false alarms.” The most
useful indicators are those that can detect change before it is too late to correct the problems.
6. Suitable for aggregation and disaggregation
Indicators should be designed in a manner that facilitates aggregation to higher scales or disaggregation to lower scales in space or time, for different users. Indicators that can be expressed in relation to ecosystem boundaries as well as political boundaries are very useful.
7. Usable for projections of future scenarios
Indicators that allow cause-effect relationships to be quantified and projected forward allow
for scenario analyses. This can enable evaluation of alternative policy options or management
Sets of indicators: No single indicator can provide information on all policy-relevant changes.
Assessments of ecosystem condition therefore rely on sets of indicators. Ideally, the chosen
set should include a relatively small number of individual indicators and be representative of
the relevant issue. The smaller the total number of indicators, the lower the cost and the easier
it is to communicate the findings to policy makers and the public. However, the set of indicators should not be so small or simple that they ignore important aspects of the issue being
Source: Based on CBD 2003.
Assessing State and Trends in Ecosystem Services and Human Well-being | 119
Table 4.1. Indicators developed for the Convention on Biological Diversity, in relation to their 2010
Focal Area
Status and trends of the
components of
biological diversity
• Trends in extent of selected biomes, ecosystems, and
• Trends in abundance and distribution of selected
• Coverage of protected areas
• Change in status of threatened species
• Trends in genetic diversity of domesticated
animals, cultivated plants, and fish species of major
socioeconomic importance
• Area of forest, agricultural, and aquaculture
ecosystems under sustainable management
• Proportion of products derived from sustainable
• Ecological footprint and related concepts
• Nitrogen deposition
• Trends in invasive alien species
• Marine Trophic Index
• Water quality of freshwater ecosystems
• Trophic integrity of other ecosystems
• Connectivity / fragmentation of ecosystems
• Incidence of human-induced ecosystem failure
• Health and well-being of communities who depend
directly on local ecosystem goods and services
• Biodiversity for food and medicine
• Status and trends of linguistic diversity and numbers of
speakers of indigenous languages
• Other indicator of the status of indigenous and
traditional knowledge
• Indicator of access and benefit-sharing
Sustainable use
Threats to biodiversity
Ecosystem integrity and
ecosystem goods and
Status of traditional
knowledge, innovations
and practices
Status of access and
Status of resource transfers
• Official development assistance provided in support of
the Convention
• Indicator of technology transfer
Source: UNEP/CBD/COP/7/21/Part 2, Decision VII/30; summarized in Balmford et al. (2005). Available at
4.2.2 Different sources of information
The purpose of assessments is to collate, organize, analyze, evaluate, and present
information. The sources of this information can be extremely diverse, but they
should all pass the same basic rules of admissibility:
• Can they be traced? If nobody can get access to the fundamental information
other than those who prepared it, it has little credibility. This means that the
120 | A Manual for Assessment Practitioners
sources have to be attributed, in the public domain (i.e., not secret, although
some restrictions on making the information freely available can be applied in
rare cases), and accessible (in a library or archive somewhere).
• Can they be tested? This is a hallmark of the “scientific method.” Ecosystem
assessments are part of the scientific knowledge system, even when they draw
on nonformal technical knowledge. If the information is of such a nature that
it could never be validated or it has no generality or predictive capacity (for instance, if every case is completely unique), it does not lend itself to assessment.
Peer-reviewed literature
Peer-reviewed literature is the main source of assessment information, precisely because it represents information that is already in the public domain and has already
been tested. Assessments are not meant to generate new primary knowledge (i.e., assessments are not research projects, although they often spawn such projects). They
are meant to draw mostly on existing information. Passing the test of peer review is
not a guarantee of correctness, but it does increase the confidence with which material
can be used. “Peer-reviewed” usually means published in an academic journal or book,
but other pathways are possible. Any credible and documented process by which the
information has undergone reasonably independent checking is acceptable.
Correct attribution (“referencing”) of source material is essential for traceability. Using a consistent format will ensure that all the necessary information is present (see Box 4.2). It is essential to reflect the information used in an honest, fair, and
balanced way. In other words, it is not appropriate to quote sources that have not
been read or to give selective quotations that misrepresent the overall message of the
source. It is the job of the review editors to be vigilant about such abuses. Correct
attribution of information and ideas is an issue of professional ethics—it is not acceptable to simply cite a recent source when the original idea preceded it. The reference lists should contain all author names (not just et al.). And cultural, language,
discipline, and geographic bias should be minimized in any literature consulted.
Often assessments are greatly assisted by being able to use “gray literature.”
This is material in some form of document, such as an internal report or a low-circulation periodical, that is in principle already in the public domain but is effectively
unavailable to the broader community. Such material may be used, provided a copy
is placed in a public domain repository, such as library or an archive.
Assessments wish to use the most up-to-date information, but some of this may
be in the form of “submitted,” “in review,” or “in press” papers. In press articles
(which have been accepted by the peer review process and are simply awaiting publication) may be cited. Submitted or in review papers can be included in drafts of
the assessment, but if they are not published or in press by the time the assessment
is in final proofs, they must be removed and any conclusions critically dependent on
them must be revised.
Statistical databases
Often the quantitative information in assessments is based on databases (usually
electronic, but sometimes on paper) that are not peer-reviewed or published in the
same sense as the literature just described. The correct acknowledging citation of the
Assessing State and Trends in Ecosystem Services and Human Well-being | 121
Box 4.2. Referencing information sources
The ability to trace information to its source is a key indicator of credibility. What follows is not
a prescribed reference style, but a list of essential information for different types of sources.
Use whatever reference style is most appropriate for the assessment’s intended audience. For
instance, some users prefer extensive footnotes or numbered endnotes to the “scientific” style
of embedding author and date in the text.
1. An article in a periodical: author(s), year, title of paper, name of journal, volume, and page
2. A book or report: authors(s), year, title of volume, ISBN, publisher, publisher’s address, and
number of pages.
3. A chapter in a book: author(s), year, title of chapter, editors of book, title of book, publisher,
publisher’s address, and page range.
4. A conference proceeding: author(s), year, title of presentation, name of conference or proceeding, date and place, organizing institution, address, and page range in proceeding or
published abstract.
5. A personal communication: (i.e., something told by an informant). Name and initials of
the person, date, and an address or email contact. Note: this is generally a low-credibility
source and should be used sparingly unless there is no alternative. If the person has credentials or experience that give them credibility, note that.
6. A database: author(s), date of access, version, name of database, and responsible
7. A Web site: html address, date of access, and author (where apparent).
8. A photograph: photographer, date, and place (for repeat photography, also the lens used
and direction of the photo).
databases (i.e., traceable and fair to the people who have put the effort into them)
is important (see Box 4.2 for an example). Note that although the copyright of the
data resides with the assessment, if the data have been altered substantially (expressed in a different form, for instance), it is still necessary for reasons of courtesy
and transparency to cite the originator. Since many such databases are “live,” it is
important to record the date on which the information was retrieved.
Determining the quality of data in databases is important. One approach is to
use information from different sources and see if it is in reasonable agreement. A
second check is for internal consistency. Do the row and column totals add up correctly? Are the location data in fact in the place where the study is meant to cover?
Be aware of (and acknowledge) potential biases, uncertainties, and gaps in the data,
especially where independent corroboration is difficult. Often databases will be accompanied by a peer-reviewed paper that describes the data sources and how they
have been manipulated. Databases with poor metadata and no such statements
about their strengths and weaknesses have lower credibility.
Assessments are themselves prolific sources of data. Ensure that the assessment,
in leaving behind legacy data, passes the tests just described. All the information
used should be copied onto a secure medium after the assessment and placed in
an archive that will outlive the project. Putting data into platform and softwareindependent formats is helpful where it may only be retrieved decades from now.
122 | A Manual for Assessment Practitioners
Make sure the metadata are adequate and, where possible, are embedded in the data
files themselves, as self-describing column headings and units (see the simple guide
to preparing datasets for sharing and archive by Cook et al. 2001).
Spatial data
Maps and remotely sensed images are valuable sources of information and potentially
powerful communication tools. There are well-developed standards for documenting
the origin and processing of such data, which should be adhered to both in the report
and (in greater detail) in any electronically archived datasets (e.g., Moellering 2005).
There are trade-offs to be made with remotely sensed data between scale, resolution, content detail, frequency of acquisition, and cost. Scale is the extent of the coverage, and resolution is the smallest object that can be discerned, usually described
by the pixel size. In general, large scales are associated with lower (coarser) resolution but also with higher frequency of acquisition and lower cost. The scale, resolution, and frequency need to be appropriate for the pattern expected in the image,
which ultimately relates to the purpose of the study and the scale of the processes
underlying the generation and use of ecosystem services. Higher spatial resolution
and more spectral channels do not necessarily equate to “better” data. Remember
that the cost of using remotely sensed information is not just the cost of acquiring
the data but also the cost of analyzing it.
The usefulness of the information depends on both the content it represents and
the level of processing it has undergone. Is it just one band, or is it multiple spectral
bands? Are the bands in a part of the wavelength that contains the information
being looked for? How well have the images have been corrected for geographical and atmospheric distortions? Uninterpreted images have limited use in assessments except as backdrops to maps or to show certain very obvious features such
as large-scale land transformation. The most useful products are not the raw images
themselves but products derived from them that relate directly to ecosystem services
supply and use. For example, the São Paulo sub-global assessment used a Landsat 5
thermal infra red image (band 6) to illustrate the temperature-ameliorating service
that green belts provide in urban heat islands. Land cover maps, tree cover maps,
productivity maps, digital elevation models, and fine-resolution maps of the distribution of people are all derived from remote sensing. Such products are usually accompanied by a publication that describes how they were derived and validated and
what the reliability of the product is.
The distinction between “models” and “observations” is not nearly as sharp as many
people think. Almost all measurements have some underlying model, and modern
information sources such as satellite images have very elaborate models as part of the
processing chain. Models are very useful in assessment because they allow gaps in
space and time to be filled in a consistent way and they permit extrapolation, within
reasonable limits. The citation for models must include the version used and should
include a citation of the data used as input to the model. Many models are not truly
in the public domain, and the same rule applies as for “gray literature”—a copy of the
code needs to be placed in a public archive if it is to be used in an assessment.
Assessing State and Trends in Ecosystem Services and Human Well-being | 123
Models that have passed peer review may be used in an assessment to conduct
new analyses that do not need peer review (other than that provided by the assessment itself). This is the one form of “new research” that often is part of an assessment. For example, the Intergovernmental Panel on Climate Change commissions
dedicated runs of many climate models for its assessments.
Just as with databases, it is important to assess and communicate the level of
confidence in the models. This can be done by checking the quality of their documentation, by comparing them to other models or independent observations, or by
any of the standard techniques of model validation. Models that do not permit any
form of uncertainty analysis (i.e., where no confidence limit can be given) should be
treated with caution.
Experiential, traditional, and indigenous knowledge
“Indigenous technical knowledge” and “traditional knowledge” are two of the
terms used for information that may be well established and highly relevant to the
assessment but that is not typically considered “scientific information.” Assessments
are encouraged to use such information, provided the rules of traceability and testability are satisfied. In addition, ethical rules relating to acknowledgement and use
of intellectual property must be observed, and cultural sensitivities regarding privileged information must be respected. But if information from any source is so secret
that it cannot be shared, it should not be offered to an assessment process nor used
by such a process. The issues and approaches relating to use of traditional knowledge and similar sources are dealt with in more detail in Chapter 1.
Internet sources
Increasingly information is sourced from the Internet. Be aware that such information may not have passed a peer-review test and can be transient (i.e., untraceable in
the future). It is also important not to assume that if information is not revealed by
an Internet search, it must not exist. This discriminates against pre-Internet knowledge and knowledge in parts of the world with poor connectivity.
These caveats do not apply equally to all Web-accessible sources. There is no
difference in credibility between an electronically published peer-reviewed journal
and a traditionally published one, and the same applies to electronic databases
that are supported by metadata and adequate descriptions of their origin. There is
some evidence that open-source knowledge communities (such as Wikipedia) are
of similar reliability as encyclopedia articles sourced from commissioned experts.
Citizen–science-derived, Internet-based information systems (for instance, of bird
observations) have proved very useful, especially if they have a built-in mechanism
for quality checks.
4.2.3 Fitting biodiversity into an ecosystem service framework
The ecosystem services approach is based on a “utilitarian” concept of nature. In
other words, nature is valuable because of its usefulness to humans. This is a defensible argument from the perspective that the concept of “value” is, as far as is known,
an entirely human one, and it can only be conceived of from a human perspective.
124 | A Manual for Assessment Practitioners
Even humans’ aesthetic, spiritual, cultural, or ethical appreciation of nature is, in this
sense, “utilitarian.” However, almost all researchers working in this field would concede that if humans were not present, nature would nevertheless have a “value.” This
is its “intrinsic value,” in the strictest sense. (Some people use the term intrinsic value
for the hard-to-monetize, nonconsumptive aesthetic, cultural, ethical, and spiritual
values.) So an approach like that in the Millennium Ecosystem Assessment (MA) does
not dismiss the existence of intrinsic value; it simply says that it is (by definition) unable to quantify intrinsic value and therefore cannot assess it. In this sense, utilitarian
value is complementary to nonutilitarian value; not a replacement for it.
Biodiversity is in general not an ecosystem service itself but a necessary condition for ecosystem services to be delivered. Yet there are some important exceptions
where the “diversity” part of biodiversity is itself the service—for instance, in some
types of nature-based tourism and in the prospecting for new pharmaceuticals, genetic resources, and nature-inspired products. But in most cases of natural resource
use, it is the “bio” part of biodiversity that is the service—as in the thousands of situations where a single particular species provides a food, medicine, or useful product. Here the product can be valued, but that does not itself constitute a value for
biodiversity. Another species could provide the same or similar product, and often
biodiversity is eliminated by the monocultural propagation of the preferred species.
When biodiversity is one or two steps removed from the service—for instance, in
most regulating and supporting services—attempts to value biodiversity directly are
pointless and amount to double accounting. The economic value of biodiversity is
embedded in the value of the services. This view (see Figure 4.1) represents an elaboration on the somewhat vague and simplistic representation of biodiversity (e.g., in
the first versions of the MA conceptual framework, MA 2003) as a cloud of “life on
earth” surrounding ecosystem services.
In general, the relationship between biodiversity and the delivery of ecosystem
services is indirect. There is a large and growing body of evidence that the yield of
many provisioning services is on average higher from systems with more biodiversity
than from monocultures, although the maximum service yield from a monoculture
may be equal to the maximum yield from a mixed system. The relationship between
ecosystem service quantity and biodiversity approaches a maximum at a fairly low
level of biodiversity (in the order of 10 species rather than hundreds). The relationship is clearer when biodiversity is expressed in terms of the number of functional
types rather than the number of species.
A key value of biodiversity may be in reducing the variability of ecosystem services, (equivalently, reducing the uncertainty or risk), especially in the face of disturbances or changes in the environment. In other words, the value of biodiversity is
expressed less via the average yield of the service (µ in figure 4.1) than via its higher
moments, such as the variance (σ) or spatial diversity (γ).
Biodiversity can be measured and expressed in any number of ways (see a comprehensive treatment by Magurran 2004). Some of the key approaches are listed in
Table 4.2.
4.2.4 Dealing with multiple scales
Assessments are typically carried out at particular scales: local, national, regional, or
global, to name a few common ones. But they often rely on information collected at
Assessing State and Trends in Ecosystem Services and Human Well-being | 125
Figure 4.1. A conceptual model of the relationship between biodiversity and human
well-being, via ecosystem services. The symbol µ indicates that the influence operates
mostly through mean amount of the service that is generated, and that matters most in
this influence, whereas σ indicates that the influence on temporal variability is the key
factor, and γ indicates that the influence is mostly on spatial heterogeneity.
Source: Unpublished work by R. J. Scholes, C. Perrings and A. Kinzig.
greater or smaller scales, or they must interface with assessments or decision-making
processes at somewhat different scales, either larger or smaller. The characteristic
scale at which ecosystem services, their drivers, or the decision-making processes
that influence them operate is seldom exactly the scale of the assessment. Since ecosystem services are place and time specific and do not necessarily aggregate upward
in a simple additive way or disaggregate downward by simple proportionality, care
must be taken in relation to scale questions in both time and space. It is important
to work at an appropriate scale for both ecological and human processes, right from
the start. (For a theoretical review of scale-related issues pertinent to ecosystem assessments, see MA 2003, chapter 7.)
As noted earlier, it is important to not confuse scale (the maximum extent or, in
the time dimension, duration) with resolution (the smallest event detectable at that
scale, sometimes called the “grain”). Usually the problems are caused by resolution
mismatches rather than scale itself. Even fine-scale processes can be adequately analyzed at large scales, if the resolution is appropriate and there is enough computing
126 | A Manual for Assessment Practitioners
Table 4.2. Selection of biodiversity indicators
Area of ecosystem types (e.g., forest, agriculture, built-up)
Current extent of natural vegetation types relative to their
pre-industrial era extent
IUCN Red List species (IUCN 2002)
Trends in representative or key species
Number and status of endemic species
Number and status of migratory species
Number and share of livestock breeds and agricultural plant
Number of endangered varieties of livestock breeds and
agricultural crops
Road density
Change in mean annual temperature and precipitation
Damming and canalization of rivers
Pollutants exceeding soil, water, or air standards
Number of invasive species
Amount harvested per species
Carbon stored in forests
Total revenue from ecotourism
Percentage protected area by IUCN category (IUCN and UNEP
Number of threatened and invasive species with a
management plan
Conservation policy capacity, in number of people
Number of local site support groups and volunteer monitors
Number of physical and chemical standards
Hotspots (high endemism with high human impact) (Myers et
al. 2000)
Living Planet Index (Loh 2002)
Natural Capital Index (Ten Brink 2000)
Biodiversity Intactness Index (Scholes and Biggs 2005)
Human footprint (Sanderson et al. 2002)
Total Pressure Index (UNEP 2002)
Pressures and threats
Composite indicators
Source: Based on Biggs et al. 2007, CBD 2003.
Provisioning services can generally be aggregated to higher scales if they are expressed in consistent and absolute units. Conservation of mass means that the supply of a mass or volume-based quantity of a service at a larger scale is the sum of the
services at subsidiary scales, provided there is an unbiased sampling scheme. But in
general even provisioning services expressed at a coarse resolution cannot be disaggregated into finer resolved form without using some form of probabilistic model. If
the underlying drivers of the spatial or temporal variation in the service are known,
and there is a representation of the fine-resolution variability (in time or space) of
Assessing State and Trends in Ecosystem Services and Human Well-being | 127
the driver, the driver field can be used as a covariate to predict a likely distribution.
This process is used, for instance, to derive highly disaggregated population data
from national census numbers, using information such as the distribution of nighttime lights or roads as predictor variables.
The major problems arise for indicators where these “simple” rules do not apply. For instance, many biodiversity measures are scale dependent. An example is
species richness: there is a strongly nonlinear relation between the number of species
and the area under consideration. If the scale relationship is known, scaling laws
can be applied. For example, in the species richness case, the scaling relationship
is called the species-area curve, and it usually follows an exponentially saturating
form with ecosystem-specific coefficients. When nonlinear scaling relationships are
involved, the basic rule is that averaging (which is a form of aggregation) must take
place at the end of the calculation rather than the beginning. In other words, the
result, not the drivers, get aggregated.
A related problem is preserving the detail in findings when determining averages. Information often needs to be averaged in order to provide data summaries
that are comprehensible. But the real world is invariably patchy, and there is a risk
of homogenizing this spatial, temporal, and social variability in the average. It is
possible to conclude, for instance, that there is no poverty at all because the average
income is above some specified level, when in fact a large proportion of the population is very poor, but that is balanced by a small number of very rich individuals.
One solution to this loss of information is to present the distribution of the values
as well as the mean. This can take several forms, here listed in increasing levels of
complexity: give the range, the standard deviation, higher order moments such as
skewness and kurtosis, or the quantiles of the distribution. A graphic (for instance,
a box-and-whisker plot) often captures this information more clearly than a table
with lots of numbers. Spatial patchiness can be represented by color-shaded maps,
while temporal variability can be shown by time-course graphics.
4.3 How to gather and assess information on ecosystem services
Section’s take-home messages
• It is essential to be clear about whether an ecological stock or an ecosystem service flow
is being measured. In general, stocks are expressed in units of quantity, while flows are
expressed as quantities per unit time. Usually, ecosystem services are flows, both on the
supply side and the demand side.
• Ecosystem integrity can be expressed through a combination of various indicators: the
extent of the ecosystem relative to some reference state, the degree of fragmentation
of remaining patches, the change in community composition of the ecosystem, and the
capacity of the ecosystem to deliver a given service or basket of services per unit area.
• The “ecological footprint” indicator converts demand for a range of ecosystem services
into the equivalent land or sea areas needed to supply that service and then adds them
up. It can be calculated for an individual, an enterprise, or a nation.
This section describes the techniques used to collect and evaluate information about
ecosystem services and the metrics used to express them. Table 4.3 provides a list of
Table 4.3. Some indicators and possible proxies for the main ecosystem services assessed in
MA-type assessments
Ecosystem service
Potential Indicator
Possible proxies
Food crops
Livestock production
Yield of crop product
Offtake of animals or their
Livestock as assets, draught
animals, or cultural icons
Livestock biomass or
metabolic equivalent
mass (e.g., tropical
livestock unit)
Offtake of given species
Stocks of species
Area planted to crop
Turnover or gross profit in
meat, dairy and hide
Livestock numbers by
Wild-harvested food
(fisheries, hunting)
Energy crops, including
fuelwood and charcoal
Fiber (cotton, hemp, wool,
silk, paper pulp, etc.)
Wood as timber
Fresh water
Harvest of products, usually
as m3, but also in local
units such as board-feet
or number of poles
m3 of fit-to-use water (for
large flows, km3 are used)
Harvest of known medicinal
species (tons, or number
of organisms)
CO2 sequestration
Net CO2 flux out of
Denitrification, P fixation, S
Difference in concentration
of toxin in input and
output stream
Height and duration of flood
Attenuation of coastal
flooding, erosion and
damage to infrastructure
or resources
Intensity, duration, and
extent of outbreaks of
undesirable species
N,P, and S removal
Detoxification of waste
Flood attenuation
Coastal protection
Yield (MJ) of given primary
or secondary energy
Yield of given product (tons)
Pest, pathogen and weed
Turnover or gross profit
of fisheries or hunting
% of biofuels in energy
Turnover or gross profit of
textile sector or papermaking industries
Turnover or gross profit of
forestry sector
Per capita water use
Price of water
Cost of water purification
Depth to groundwater
Number of people using
natural medicines
Change in C stock
Downstream NO3, PO4
and SO4
Illnesses attributable to
toxins, incidence of
fish kills
Losses of life and property
due to flooding
Km of coast with intact
Cost of coastal damage
Expenditure on biocides
Area occupied by alien
Number of alien species
Assessing State and Trends in Ecosystem Services and Human Well-being | 129
Table 4.3. continued
Ecosystem service
Potential Indicator
Possible proxies
Recreation & amenity
Recreational opportunities
Area of landscape in
attractive condition
Presence of sites,
landscapes, or species
of spiritual or cultural
Presence or area of sites or
species of scientific or
educational value
Tourism sector turnover or
gross profit, number of
Visitor opinion polls
Visits to beauty spots
Number or area of
important sites,
protection status
Spiritual and cultural
Scientific and educational
Energy capture
Nutrient cycling
Net Primary productivity
N mineralization
P mineralization
Cation availability
% of flowers pollinated
within a species
Area of suitable habitat for a
given species
Number of school visits
Number of papers
Cover by N-fixing plants
% mycorrhizae
CEC in the profile, % base
Populations of pollinating
Vegetation type area
Fragmentation indices
ecosystem services and some of the indicators that have been found to be appropriate for quantifying them. The list is indicative rather than exhaustive; it covers the
main, widely reported services. There are many more potential ecosystem services—
sub-global assessments, for example, may reveal new ones—and each has many possible indicators and proxies.
4.3.1 Stocks and flows
It is essential to be clear about whether an ecological stock or an ecosystem service flow is being measured. For example, the amount of tree biomass in a forest
is a stock; it supports several potential ecosystem service flows, such as the annual
harvest of wood or an annual uptake of carbon dioxide. In general, stocks are expressed in units of quantity (e.g., metric tons, m2, or ha), while flows are expressed
as quantities per unit time (e.g., kg/year or m3/second). Usually, ecosystem services
are flows, both on the supply side and the demand side. Stocks and flows need to
balance: if the demand exceeds the supply over a given period, the stock will be
depleted by an equivalent amount. For a renewable resource, if the extraction rate
is less than the natural replenishment rate, the stock will rise, usually to some maximum level.
It is usually necessary to express ecosystem services in both flow and underlying
stock terms. The significance of a particular flow is hard to judge unless the size of
the stock is known (and for renewable resources, the maximum flow that could be
extracted from it without depleting the stock). Similarly, a stock by itself seldom
says anything useful about the ecosystem service flows that are actually, or potentially could be, derived from it.
130 | A Manual for Assessment Practitioners
Both stocks and flows can be expressed in economic (monetary) terms or in
physical quantity terms. The economic value of a flow is calculated as the quantity per unit time multiplied by an average price. The equivalent economic value
of a stock that underlies a flow can be calculated from the present and assumed
time course of future ecosystem service flows, using an appropriate discount rate
(typically quite small, 1–3% per annum). This number is called the natural capital
(Dasgupta and Maler 2001). Note that a stock or flow that is essential for survival
and that cannot be substituted with another resource or with money would have an
infinite value.
“Ecosystem services” are any benefit that people derive from nature, and they
need not be restricted to living or renewable resources. Nonrenewable natural resources, such as ore bodies, fossil aquifers, and deposits of coal, oil, or gas, can also
be regarded as natural capital stocks delivering a flow of services that end up supporting human well-being. For nonrenewable resources, the accounting must take
into account their declining, nonself-replenishing nature—in other words, that the
natural capital is depleted as the product is converted to financial or social capital.
This must be factored into the net change in “inclusive wealth.”
Not all ecosystem services are “consumed” when they are used. For instance, admiring a cultural landscape or a biodiversity icon does not necessarily make it unavailable to be admired by someone else. Even water is not destroyed when it is used: it is
typically converted to another form (e.g., somewhat polluted) that may be unsuitable
for immediate reuse for the same purpose but may be useful for another purpose.
The flows of provisioning services can often be directly measured, as a harvest
yield over a period of time. Alternatively, they can sometimes be measured as a
change in the stock over a given period. For instance, many ecosystems provide a climate regulating service by sequestering carbon from the atmosphere. It is possible to
measure this flux of carbon dioxide (CO2) from the atmosphere into the ecosystem
directly, but the equipment needed is expensive and difficult to use. Over time, the
net flux will show up as a change in the stock of carbon in the biomass, soil, sediment, or water body, and this is easier to measure.
4.3.2 Measures of ecosystem integrity
Ecosystem integrity, which is largely synonymous with ideas such as ecosystem
health, quality, or intactness and is the converse of ideas such as degradation, can be
expressed with combinations of the following indicators.
• The extent of the ecosystem relative to some reference state, such as its former
extent at some time during the period of historical record, or the extent inferred
from paleo-ecological data, or a “potential” distribution inferred from climate
or substrate requirements. This is the simplest indicator of ecosystem loss, but it
is a measure of stock rather than ecosystem service flow.
• The degree of fragmentation of the remaining patches, which can be expressed
using a wide variety of metrics, including average patch size, the perimeter:area
ratio, and the degree of connectivity between patches. These measures are particularly important with respect to the habitat-supporting service, since the number and type of species that can persist in a patch depends, in the long term, on
the size of the patch and its exposure to disturbance along its perimeter.
Assessing State and Trends in Ecosystem Services and Human Well-being | 131
• The change in community composition of the ecosystem, usually expressed
relative to some reference state (such as the “natural state” inferred from some
examples believed to be in this state). The difference metric must be able to cope
with simultaneous change across many species. Some are quite simple (such as
the Euclidean distance to the reference state in the “hypervolume” defined by
the abundances of each species), while others use advanced statistical concepts,
such as collapsing the many variables onto a few “principal components.” Composition measures are very sensitive—perhaps oversensitive—to natural fluctuations and variations in species composition. They can be hard to interpret: is an
observed change in composition a good or bad thing? An alternative approach
is to reduce the community diversity to an index, such as the species richness or
the Simpson’s Diversity index, and then track changes in that. Note that the link
between compositional diversity and ecosystem service delivery remains generally unclear.
• The capacity of the ecosystem to deliver a given service, or basket of services,
per unit area. This is a sensitive and useful indicator and forms a robust and defensible basis for the definition of concepts such as degradation. The MA Desertification Synthesis defines degradation as a persistent reduction in the capacity
of the land to deliver one or more ecosystem services. Service-based integrity indicators pick up the subtle changes in ecosystem functioning that are missed by
simple loss-of-area measures—for instance, the high grading of valuable timber
out of forests without clear-cutting or the consumption of all the bushmeat from
under the canopy.
4.3.3 The balance of ecosystem service supply and use
Some intuitive ecosystem service indicators combine measures of ecosystem service
demand (consumption) with estimates of the supply capacity. The simplest of these
calculates the difference between supply and demand (see Box 4.3 for an example).
This is a better approach than calculating a supply:demand ratio, which becomes
undefined if either tends to zero.
A composite form of this kind of indicator is the ecological footprint. It converts demand for a range of ecosystem services into the equivalent land or sea areas needed to supply that service and then adds them up (Rees and Wackernagel
1994, see also The ecological footprint can be
calculated for an individual, an enterprise, or a nation. A footprint is unsustainable
when it is greater than the area of ecosystems that are equitably available to that
entity. The ecological footprint is a very graphic and easily communicated way of
expressing ecosystem service supply versus demand, but it has several problems.
First, although all the components are expressed in the same units (hectares) and
can therefore be mathematically added, it is not clear that they can be ecologically
added, since often the same hectare can deliver several services simultaneously, or
with only weak trade-offs between them. Second, the area needed to supply the
service may not be based on the yield where the service is actually derived (which is
almost impossible to trace or calculate) but on some “mean yield,” which is easily
calculated but may not be relevant.
132 | A Manual for Assessment Practitioners
Box 4.3. Calculation of the fuelwood deficit in Southern Africa
In the Southern Africa Millennium Ecosystem Assessment, the availability of fuelwood as an
ecosystem service was assessed by calculating the difference between supply and demand.
Fuelwood supply was calculated at a 5 × 5 km resolution based on climate and satellite-derived tree cover data. Fuelwood demand (b) was calculated by scaling average consumption
in rural and urban areas by a function of ambient temperature and woodfuel availability. The
results show clearly that, contrary to popular conception, woodfuel scarcity in southern Africa
is confined to very specific parts of the region.
Box 4.3 Figure. Supply (a), demand (b) and shortfall or surplus (c) of fuelwood in
southern Africa in 2000.
Source: Scholes and Biggs (2004)
4.4 How to gather and assess information on human well-being
Section’s take-home messages
• In the MA, human well-being is considered to have many dimensions, including but not
restricted to monetary income. In this view, poverty is defined as the absence of wellbeing.
• Values expressed in monetary units that are accrued over a multi-year period are
corrected for the effects of inflation, and are often converted to an internationally
comparable currency.
The MA takes a particular position on human well-being: that it is multidimensional
and, in particular, includes considerations in addition to monetary income (see MA
2003, Chapter 3). In this view, poverty is defined as the absence of well-being rather
than simply as not having a certain minimum amount of money. Some important axes
of human well-being, and their indicators, are summarized in Table 4.4. (There are several comprehensive discussions of this topic; see Alkire 2002, Prescott-Allen 2001.)
Assessing State and Trends in Ecosystem Services and Human Well-being | 133
Table 4.4. The constituents of human well-being recognized by the Millennium Assessment, and
some potential indicators for them
HWB constituent
Potential Indicators
Basic material for a
good life
Access to resources
to sustain
Offtake of food, fiber, fuel, medicines,
construction materials and
freshwater to meet basic needs
Security of resource or land tenure
Income (including from sale of above)
Measures of income distribution
Protein intake per day
Digestible energy in food
Deviation from target weight per height
Expected longevity at birth
Childhood mortality
Disease-adjusted life years
Exceedance of guideline limits
Prevalence of indicator health
Ability to keep warm or cool
People in adequate housing
Risk of death, injury, or property loss
through natural hazards
Deaths, injuries, property, or
infrastructure loss; number of
displaced people due to armed
Happiness measures
Exposure to toxins
Exposure to risk
Good social relations
Absence of conflict
Sense of belonging
Freedoms and choice
Participation in
decision making
Level of education
Gender bias in education
Level of corruption
Fairness of elections
The assessment team needs to make a number of decisions on technical issues.
For measures that are expressed in monetary terms, income over a multiyear period
is usually expressed as the equivalent at a given time, to allow for the effects of
inflation. For instance, an assessment over the period 1991 to 2000 may choose to
express its income in equivalent year 2000 terms. To do this, econometric time series
of inflation indexes (for example, the consumer price index) are needed. Second, the
income is often expressed not only in local currency terms but also in an internationally comparable currency, such as the U.S. dollar. To do this, a rate of exchange and
a date must be specified. Often this rate of exchange is not based on the market rate
but on the purchasing power parity (PPP) conversion rate, which is more reflective
of the impact on human well-being.
Many indicators have the tendency to mask pockets of low well-being by averaging
across the entire population. For instance, gross income indicators (such as the GDP/
capita) by themselves are less informative than income plus some measure of its distribution in society: the quintiles of income, for example, or some composite measure of
income equality, such as the Gini coefficient, or some measure such as the fraction of the
134 | A Manual for Assessment Practitioners
population below a specified income limit. Similarly, health measures such as longevity
or child mortality are more informative if they also give an indication of the distribution of the indicator in the population. They can sometimes be disaggregated spatially
(since the primary data source is usually the local clinic). If possible, they should be
expressed per vulnerable class (e.g., gender, ethnic group, income class, or age).
A number of composite indices of human well-being have been attempted, and
some are routinely calculated and available at the national scale. The U.N. Development Programme (UNDP) annually reports the Human Development Index for every country. This combines three dimensions: health (measured by life expectancy at
birth), education (measured by the adult literacy rate and the combined gross enrolment ratio across all educational levels), and income (measured by the logarithm of
gross domestic product per capita, at purchasing power parity in dollars). UNDP’s
Human Poverty Index is a geometric average of three indicators: the probability of
surviving to age 40, the fraction of adults who are literate, and the average of the
fraction of the population that has access to safe water supplies and the fraction of
children who are underweight for their age.
The Wellbeing of Nations prepared by Robert Prescott-Allen combines, for each
country, 36 indicators of health, population, wealth, education, communication,
freedom, peace, crime, and equity into a Human Wellbeing Index. It also combines
51 indicators of land health, protected areas, water quality, water supply, global
atmosphere, air quality, species diversity, energy use, and resource pressures into an
Ecosystem Wellbeing Index. The ratio of the two indexes is an index that measures
how much human well-being each country obtains for the amount of stress it places
on the environment (Prescott-Allen 2001).
The Millennium Development Goals adopted by the United Nations aim to raise
the well-being of people all over the world. They are associated with specific targets,
each of which has one or more indicators. Data on these indicators are often available
at national resolution, and sometimes at sub-national resolution (see Table 4.5).
4.5 How to assess the link between ecosystem services and human
Section’s take-home messages
• The first step in establishing the link between an ecological resource or service and human
benefit is to sketch out a causal pathway linking the service in question to the elements of
human well-being it is thought to influence.
• Economic valuation of ecosystem services is a complex technical field that can be difficult
for untrained and inexperienced people to master. Yet several relatively simple steps can
lead in the direction of economic valuation, such as expressing the service flows and their
underlying stocks in quantitative terms.
• It is important to calculate the economic value of ecosystem services—the all-inclusive
value to society as a whole, over the entire life cycle, taking into account both taxes
and subsidies and adding in the externalities—rather than just the financial value to the
immediate beneficiary.
• A commonly used tool to assess trade-offs is the social preference function. Tradeoffs occur when the extraction and use of one service has an impact—positive or
negative—on the benefit that can be realized from another service.
Table 4.5. Indicators used to quantify progress toward the U.N. Millennium
Development Goals
Goal and targets
Eradicate extreme poverty
and hunger. Target 1: Halve,
between 1990 and 2015, the
proportion of people whose
income is less than one dollar a
day. Target 2: Halve, between
1990 and 2015, the proportion
of people who suffer from
Achieve universal primary
education. Target 3: Ensure
that, by 2015, children
everywhere, boys and girls alike,
will be able to complete a full
course of primary schooling.
1. Proportion of population below $1 (1993 PPP)
per day (World Bank)
2. Poverty gap ratio [incidence × depth of
poverty] (World Bank)
3. Share of poorest quintile in national
consumption (World Bank)
4. Prevalence of underweight children under five
years of age (UNICEF–WHO)
5. Proportion of population below minimum level
of dietary energy consumption (FAO)
6. Net enrollment ratio in primary education
7. Proportion of pupils starting grade 1 who
reach grade 5 (UNESCO)
8. Literacy rate of 15–24 year olds (UNESCO)
Promote gender equality and
empower women. Target
4: Eliminate gender disparity
in primary and secondary
education, preferably by 2005,
and in all levels of education no
later than 2015.
9. Ratio of girls to boys in primary, secondary,
and tertiary education (UNESCO)
10. Ratio of literate women to men, 15–24 years
old (UNESCO)
11. Share of women in wage employment in the
nonagricultural sector (ILO)
12. Proportion of seats held by women in national
parliament (IPU)
Reduce child mortality. Target 5:
Reduce by two-thirds, between
1990 and 2015, the under-five
mortality rate.
Improve maternal health. Target
6: Reduce by three-quarters,
between 1990 and 2015, the
maternal mortality ratio.
13. Under-five mortality rate (UNICEF–WHO)
14. Infant mortality rate (UNICEF–WHO)
15. Proportion of 1-year-old children immunized
against measles (UNICEF–WHO)
16. Maternal mortality ratio (UNICEF–WHO)
17. Proportion of births attended by skilled health
personnel (UNICEF–WHO)
Combat HIV/AIDS, malaria, and
other diseases. Target 7:
Have halted by 2015 and begun
to reverse the spread of HIV/
AIDS. Target 8: Have halted by
2015 and begun to reverse the
incidence of malaria and other
major diseases.
18. HIV prevalence among pregnant women aged
15–24 years (UNAIDS–WHO–UNICEF)
19. Condom use rate of the contraceptive
prevalence rate (UN Population Division)
19a. Condom use at last high-risk sex (UNICEF–
19b. Percentage of population aged 15–24 years
with comprehensive correct knowledge of
19c. Contraceptive prevalence rate (UN
Population Division)
Table 4.5. continued
Goal and targets
20. Ratio of school attendance of orphans to
school attendance of nonorphans aged 10–14
21. Prevalence and death rates associated with
malaria (WHO)
22. Proportion of population in malaria-risk
areas using effective malaria prevention and
treatment measures (UNICEF–WHO)
23. Prevalence and death rates associated with
tuberculosis (WHO)
24. Proportion of tuberculosis cases detected
and cured under DOTS (internationally
recommended TB control strategy) (WHO)
Ensure environmental
sustainability. Target 9:
Integrate the principles of
sustainable development
into country policies and
programs and reverse the loss
of environmental resources.
Target 10: Halve, by 2015, the
proportion of people without
sustainable access to safe
drinking water and sanitation.
Target 11:
By 2020, achieve a significant
improvement in the lives of at
least 100 million slum dwellers.
25. Proportion of land area covered by forest
26. Ratio of area protected to maintain biological
diversity to surface area (UNEP–WCMC)
27. Energy use (kg oil equivalent) per $1 GDP
(PPP) (IEA, World Bank)
28. Carbon dioxide emissions per capita
(UNFCCC, UNSD) and consumption of
ozone-depleting CFCs (ODP tons) (UNEP–
Ozone Secretariat)
29. Proportion of population using solid fuels
30. Proportion of population with sustainable
access to an improved water source, urban
and rural (UNICEF–WHO)
31. Proportion of population with access
to improved sanitation, urban and rural
32. Proportion of households with access to
secure tenure (UN–HABITAT)
Develop a global partnership
for development. Target
12: Develop further an open,
rule-based, predictable,
nondiscriminatory trading and
financial system. Target 13:
Address the special needs of
the least developed countries
(includes tariff and quota-free
33. Net ODA, total and to LDCs, as percentage of
OECD/Development Assistance Committee
(DAC) donors’ gross national income (GNI)
34. Proportion of total bilateral, sector-allocable
ODA of OECD/DAC donors to basic social
services (basic education, primary health
care, nutrition, safe water, and sanitation)
Table 4.5. continued
Goal and targets
access for least developed
countries’ exports, enhanced
program of debt relief for
heavily indebted poor countries
(HIPC) and cancellation of
official bilateral debt, and more
generous ODA for countries
committed to poverty reduction).
Target 14: Address the special
needs of landlocked developing
countries and small island
developing states (through the
Programme of Action for the
Sustainable Development of
Small Island Developing States
and the outcome of the twentysecond special session of the
General Assembly). Target 15:
Deal comprehensively with the
debt problems of developing
countries through national and
international measures in order
to make debt sustainable in
the long term. Target 16: In
cooperation with developing
countries, develop and
implement strategies for decent
and productive work for youth.
Target 17: In cooperation with
pharmaceutical companies,
provide access to affordable
essential drugs in developing
countries. Target 18: In
cooperation with the private
sector, make available the
benefits of new technologies,
especially information and
Source: based on UNGA 2001.
35. Proportion of bilateral ODA of OECD/DAC
donors that is united (OECD)
36. ODA received in landlocked developing
countries as a proportion of their GNIs
37. ODA received in small island developing
states as proportion of their GNIs (OECD)
38. Proportion of total developed-country
imports (by value and excluding arms)
from developing countries and from LDCs,
admitted free of duty (UNCTAD, WTO, World
39. Average tariffs imposed by developed
countries on agricultural products and textiles
and clothing from developing countries
(UNCTAD, WTO, World Bank)
40. Agricultural support estimate for OECD
countries as percentage of their GDP (OECD)
41. Proportion of ODA provided to help build
trade capacity (OECD, WTO)
42. Total number of countries that have reached
their HIPC decision points and number that
have reached their HIPC completion points
(cumulative) (IMF, World Bank)
43. Debt relief committed under HIPC initiative
(IMF, World Bank)
44. Debt service as a percentage of exports of
goods and services (IMF, World Bank)
45. Unemployment rate of young people aged
15–24 years, each sex and total (ILO)
46. Proportion of population with access to
affordable essential drugs on a sustainable
basis (WHO)
47. Telephone lines and cellular subscribers per
100 population (ITU)
48. Personal computers in use per 100
population and Internet users per 100
population (ITU)
138 | A Manual for Assessment Practitioners
Ecosystem services support human well-being via many (but not infinitely many)
paths, as indicated earlier in Figure 4.1. The crucial first step in establishing the link
between an ecological resource or service and human benefit is to sketch out a causal
pathway, which can be thought of as a hypothesis, linking the service in question to
the elements of human well-being it is thought to influence. This pathway may have
several steps, especially if the service in question is a supporting or regulating service.
Quantifying the key elements in this pathway is the next step, and this will rely on
appropriate indicators at each step of the pathway. These indicators may be derived
from observations, from models, or (most likely) from some combination of both.
It is important to bear in mind that human well-being is multidimensional and
is influenced by factors that may be unrelated to ecosystem services. For instance,
growth of the manufacturing economy or the availability of foreign aid may increase
well-being, but they are very indirectly and distantly related to ecosystem services.
Furthermore, impacts on human well-being may be slow to materialize. Therefore
it will often be hard to unequivocally demonstrate that a particular service has been
solely responsible for an observed change in human well-being. This is not a problem unique to ecosystem services. It is an intrinsic feature of impact analysis in
highly connected environments.
4.5.1 Ecosystem service valuation
The economic valuation of ecosystem services is not essential to an ecosystem service assessment, but when it is possible to do one in a reasonably rigorous way, it is
very useful for three reasons:
• The process imposes a high level of rigor on the assessment.
• Being able to express different services in a common denominator (economic
value) allows trade-offs to be explicitly evaluated.
• Communication of the importance of ecosystem services within a policy
environment is greatly facilitated if it is accompanied by credible economic
A word of caution is needed. Economic valuation of ecosystem services is a complex
technical field that can be difficult for untrained and inexperienced people to master.
Often, such evaluations do not already exist when an assessment is launched, so the
assessment would need to undertake the studies itself: a violation of the guideline
that assessments should not engage in new research. Furthermore, such studies may
be costly and time consuming. Only venture into economic valuation as far as your
team’s skills, funds, and time allow.
On a more positive note, several relatively simple steps can lead in the direction
of economic valuation and are very useful. For instance, expressing the service flows
and their underlying stocks in quantitative terms is an important step. Describing
the way in which they lead to human benefits is another key step, even if it cannot
be quantified. Often it is relatively straightforward to calculate some of the components of total economic value, even if they cannot all be estimated—and this may
be enough to make the point that ecosystem services are valuable. For instance, use
values are often easier to estimate than nonuse values, and within use values, actual
direct use values are easier to estimate than indirect use values or option values.
Assessing State and Trends in Ecosystem Services and Human Well-being | 139
Economists are quick to point out that economic value is not synonymous with
monetary value. Money is simply a convenient way of expressing the value that
society applies to things. Some values can be denominated in other terms, which
may be more appropriate than money in certain cases—the number of lives lost, for
instance. For the purposes of ecosystem services assessment, it is the economic value
that matters (i.e., the all-inclusive value to society as a whole, over the entire life
cycle, taking into account both taxes and subsidies and adding in the externalities)
rather than the financial value (which is calculated at the level of the entity being
accounted only, excludes costs and benefits relating to the broader context, and is
expressed only in monetary terms). Many things that are financially profitable are
economically loss making, and vice versa.
Some ecosystem service assessments calculate the total value of the service (e.g.,
Costanza et al. 1997). Note that “total value” can have several meanings. In contrast to marginal value, it means the accumulated value to all beneficiaries of the
total consumption of a single service, over the entire area and duration of the assessment. Total economic value can also mean the value accrued by all the mechanisms
illustrated in Figure 4.2. Total value could also mean the value accumulated across
all the different ecosystem services that exist in an area or just the subset of services
for which value can be calculated. The phrase “global value” is also sometimes used
as a synonym to total value, which gets confusing in the context of global versus
sub-global ecosystem assessments, so we do not recommend using it.
Total Economic Value
Figure 4.2. The Total Economic Value classification of the components of value of
ecosystem services Note that “non-use value” should not be confused with “nonconsumptive use”—i.e., use that does not reduce the stock. It means that the person who
is assigning value is not the person who uses the service. Further, “existence value” should
not be confused with “intrinsic value” (the ethically based value of other species to exist,
independent of the benefits they may yield to humans). If humans can value it, however
esoterically, it is by definition an extrinsic value. Note that in the indirect use values should
be completely reflected in the direct values and should not be counted twice.
Source: Adapted from DEFRA 2007
140 | A Manual for Assessment Practitioners
Usually the more useful value to calculate is the marginal value—in other words,
the extra value that is added by one more unit of the service over and above the current supply. This is more applicable for policy purposes where small adjustments in
the balance between services are envisioned rather than the total replacement of a
service. The marginal value is the relevant information for calculating the trade-off
with other options within a small range on either side of the current state. Marginal
values are calculated by well-established methods that typically depend on solving
economic equilibrium models.
Often the value of an ecosystem service is equated to the financial turnover of
that commodity—in other words, the volume multiplied by the price. This is sometimes all that is possible to calculate, and it is permissible if it is used in a comparative sense, in relation to values for other services or economic activities that
are calculated in the same way, as a sort of relative index of importance. Strictly
speaking, it is the value addition of a service that should be calculated—that is, the
realized income from that service less the cost of production. For instance, the value
of the provisioning service provided by a particular crop is not the market price
multiplied by the yield; it is this price less the cost of the inputs needed to achieve the
yield and the cost of externalities (the costs borne by society at large, such as pollution of a river). It should also be corrected for subsidies.
A technical question that always arises is the appropriate discount rate to use for
bringing future costs and benefit flows to present terms (see European Communities
2008, Chapter 3). There is no consensus on this issue, but general agreement that a
fully commercial interest rate (in other words, the time cost of money that a commercial bank would charge to a private entity) is inappropriately high for ecosystem
service assessments. Instead, a social discount rate should be used, especially if the
uncertainty about the future is high (Carpenter et al. 2007). This rate represents the
degree to which society as a whole values present benefits relative to future benefits, taking into account all the risks involved and not allowing for profits in the
There are arguments that the “ethical” discount rate would be zero, since people
today have no right to discount the options that future generations can have without their participation. This leads to unhelpful outcomes in most cases—it basically
means that the future value of all renewable resources is infinite. Furthermore, even
if the ethics argues for zero discounting, there remains a time cost of money, which
may be quite low in real terms. Therefore, assessments often perform key valuation
calculations using a range of discount rates between the social (1–3 %) and the near
market (10–15%). Some work suggests using negative discount rates (i.e., valuing
the future above the present) in addressing crucial environmental issues like climate
change and biodiversity loss (Dasgupta 2007).
There are many detailed treatments of the techniques of ecosystem service valuation. (A useful guide in the context of ecosystem services is CBD 2007; in the same
Technical Series, also see TS 3 on wetlands, TS 4 on forests, and other relevant titles.
See also DEFRA 2007.) The techniques rely either on “revealed preference” (actual
choices made in real markets, usually only applicable to use values) or on “stated
preferences,” based on hypothetical choices elicited through questionnaires. Revealed preference methods are usually based on “willingness to pay,” whereas stated
preferences may be based either on willingness to pay for a service or willingness to
accept payment for a service. Table 4.6 gives an overview of the techniques that may
Table 4.6. Methods of ecosystem service valuation
Brief explanation
What is the impact on
the production of a
marketed good caused
by a change in the
supply of an ecosystem
Price times volume gives
the value
Linked to actual
Needs a clear
and quantified
understanding of the
causal relationship
between the service
and the product
Applicable to directly
marketed ecosystem
services only (i.e.,
provisioning services
and some cultural
Can overestimate the
actual value
Market value
Usually calculated as what
(also called
someone would have to
pay to get the service in
another way or to restore
a service if it were lost or
Cost of illness
Calculates the equivalent
or morbidity
value of the loss of
(also called
earnings or life that
human capital
would result if the
service were not
Hedonic pricing
and choice
Travel cost
Random utility
Value that shows up in the
prices of other goods,
for instance the increase
in property value where
the supply of a service
is high
Based on questions such
as “what would you be
willing to pay for this
What people are shown to
be willing to pay to travel
to destinations where
cultural services (beauty,
recreation) are on offer
Extension of travel cost
method; examines the
impact of changing
quality or quantity of an
ecosystem service at a
given site
Data readily
available and
Applicable to
such as the
of toxins and
pollutants or
the control of
Linked to markets,
so data are
available and
Controversial when applied
to mortality using
inequitable “value of
Applicable to all
services, and to
use and nonuse
Based on
Subject to bias and high
Data intensive, and mostly
related to property
Limited to some cultural
services; can
overestimate if a trip
is made for many
Limited to use values,
mostly for recreational
142 | A Manual for Assessment Practitioners
be useful in particular situations. (For more details, references, and examples, see
CBD 2007 and Defra 2007.)
In some circumstances nonmonetary valuation is more appropriate than monetary valuation—for instance, in barter economies. It may be ethically inappropriate
to place a value on human life because such values often implicitly devalue the lives
of the poor relative to the rich. In these cases a variety of techniques can be used—
for instance, preference ranking for certain services, undertaken using participatory
methods (Howarth and Wilsdon 2006), or measurement in terms of other human
well-being indicators, such as life span, disease-adjusted life years, infant mortality,
or weight-for-age curves.
4.5.2 Assessing trade-offs
A trade-off occurs when the extraction and use of one service has an impact on the benefit that can be realized from another service (Scholes 2009). This impact can be positive
or negative. A positive interaction is called a synergy or, in popular terms, a win-win
situation. For instance, many of the supporting, regulating, and cultural ecosystem services are synergistic—actions taken to strengthen one also strengthen others.
Unfortunately, the more common and worrying case is a negative trade-off (a
win-lose or lose-lose situation). For instance, increasing the production of one provisioning service usually reduces the availability of other provisioning services and of
regulating, supporting, and cultural services. The issue of assessing trade-offs is discussed further in Chapter 6. Some of the methods that can be used are very sophisticated. If these are beyond the assessment’s capacity, at least show which services are
likely to be involved in trade-offs, and try to give an indication of the nature of the
trade-off (positive or negative) and its approximate magnitude, in a relatively simple
matrix form (see, for example, Table 4.7).
A commonly used tool to assess trade-offs is the social preference function. For
individuals, this reflects the preference ordering of a range of constituents of wellbeing that the individual has reason to value. The list of constituents as well as their
weights will differ among individuals. The preference ordering provides information
on the perceived relative importance of various ecosystem services as determinants
of the various constituents of human well-being. A commonly used approach to solicit the social preference ordering is through deliberative participatory techniques,
whereby groups of individuals are asked to rank their preferences over a range of
constituents of well-being with and without the knowledge of the links between the
various constituents of well-being and ecosystem services.
4.6 How to communicate state and trends
Section’s take-home messages
• Assessments are a communication tool between researchers and decision makers. If they
are technically proficient but fail to communicate, they fail overall.
• For many issues there are vast amounts of data and information available. Having specific
questions to be answered will guide selection of the appropriate data and the techniques
for collating and synthesizing the data.
Source: Tomich, T. P. et al. 2005, (Table 1).
$ + consumption
$ + consumption
$ + consumption
$ + consumption
Prices are based on 1996 averages, and expressed in December, 1996 R$ (US$ = R$1.04), discounted at 9% per annum.
Food Security, consumption, and $ reflect, respectively, whether the technology generates food for own consumption or income that can be used to buy food, or both.
0.5 to 1
0.5 to 1
0.5 to 1
0.5 to 1
Entitlement path
(operational phase)
Household food
0 to 0.5
0 to 0.5
(private prices)
Returns to
Indicates time averaged above-ground carbon.
0 to 0.5
0 to 0.5
0 to 1
0 to 1
ha–1yr –1
Smallholders’ concerns/Adoptability by
Crop protection
Returns to land
(private prices)
R$ ha–1
National policy makers’ concerns
NM = not measured; NA = not applicable.
Improved pasture
Improved fallow
Species per
Aboveground t C
ha–1 (time
Plot-level production sustainability
Managed forestry
Land-use systems
Agronomic sustainability
(0 = no issue, 1 = major issue)
Global environmental
Table 4.7. An example of a trade-off matrix. This one evaluates divergent potential uses of a tropical rainforest ecosystem.
144 | A Manual for Assessment Practitioners
• Assessments should report variation in ecosystem services over space and time. It is
often the variability in the production of the service that most affects human well-being
rather than the mean availability.
• It is fundamental to good scientific practice, and especially important in the context of
assessments, to accompany key assertions with some measure of the confidence in
those findings. For quantitative data, a statistical approach is possible; for more qualitative
findings, a set of agreed phrases regarding the evidence and amount of agreement can
be used.
• Care should be taken in presenting complex information in easily understood language
and in different formats that are suitable for a variety of audiences.
Assessments are a communication tool between researchers and decision makers. If
they are technically proficient but fail to communicate, they fail overall. Therefore
a great deal of attention must be given to choosing the best ways of presenting the
information to the intended audience and making the overall product readable, understandable, and unambiguous.
4.6.1 Collate, summarize, analyzes, and synthesize
If assessments do not collect new data, what value are they? They add value by
bringing the data together, analyzing it, evaluating it, and showing connections and
meaning that were not available in the raw data.
Any data collation or synthesis exercise should rest on careful consideration
of what information is relevant and useful in terms of the particular issue being
examined. Although collecting and analyzing data can be tedious, it should never
be approached as a mindless task. Data should not be reported simply because they
are available. Before collating and synthesizing the data, it is usually a good idea to
brainstorm and test the questions the intended audience would like answered. For
many issues there are vast amounts of data and information available. Having specific questions to be answered will guide selection of the appropriate data and the
techniques for collating and synthesizing the data.
The methods used for summarizing data will depend on the information to be
provided. Some of the most common types of information provided in assessments
of ecosystem services are measures of “central tendency” (i.e., the mean, median, or
mode, as appropriate) and measures of variability over space or time (i.e., the range
and/or standard deviation). Analyzing how different variables relate to one another
can be very informative, but it may require fairly sophisticated statistical techniques.
For any information provided in an assessment, it is important to note the degree of
confidence that can be associated with the information. This can take the form of
formal statistics (e.g., a confidence interval) for quantitative data or language that
conveys uncertainty for qualitative data (see next section).
Assessments often focus largely on reporting means, but they should probably
pay as much or more attention to reporting variability in ecosystem services over
space and time. It is often the variability in the production of the service that most
affects human well-being rather than the mean availability. For instance, freshwater availability in the driest month of the year—not the average annual freshwater
availability—is the critical constraint on agricultural production. The most common
Assessing State and Trends in Ecosystem Services and Human Well-being | 145
measures of variability or spread are the range (the difference between the largest
and smallest value) and the standard deviation (a measure of the average difference
between each individual data point and the mean of all data points). “Box and
whisker plots” are a very nice graphic tool for summarizing the mean, standard
deviation, and range of a particular dataset. Maps are powerful tools for displaying
spatial variation in ecosystem services. When comparing variation among different
variables, it is most appropriate to use the coefficient of variation, which measures
the ratio of the standard deviation to the mean (usually as a percentage) and is therefore dimensionless.
Understanding relationships between different ecosystem services or between
drivers of ecosystem change and ecosystem services can be very useful for assessing
trade-offs between services or proposed management interventions. Relationships
are often best conveyed graphically as biplots (i.e., the one value plotted against the
other value, as points or as lines). The simplest statistical measure of a relationship
between two or more variables is the correlation coefficient. But beware: a correlation between two variables does not provide proof of a causal relationship. Causal
relationships, and predicting changes in one variable from a change in another, are
typically examined using regression analysis, but regression does not “prove” a link
either. It just creates a more plausible case. Regression analyses range from simple
linear regression to highly sophisticated models with nonlinear relationships, many
predictor variables, time delays, and causal hierarchies. Gelman and Hill (2007)
provide good coverage of these techniques.
4.6.2 Conveying certainty and uncertainty
The purpose of assessments is not to fill knowledge gaps but to reveal them. It is
important, therefore, not try to paper over the cracks but to be honest about what
is not known and what is uncertain. Sometimes small missing pieces of information
can be obtained during the assessment process, and small gaps can be estimated
using extrapolation. But there will invariably still be information that is either completely missing or to some degree inadequate. An important function of assessments
is to establish the research priorities for the future.
It is fundamental to good scientific practice, and especially important in the
context of assessments, to accompany key assertions with some measure of the confidence in those findings. This obligation will not always be demanded by the audience, and it may even be resisted (“Can’t these scientists make up their mind?”).
Not every statement in the entire text needs to be accompanied by a qualification or
an error range, but the key ones do. Be vigilant in eliminating overqualification of
findings. For example “The following result may occur with a low probability...” is
a redundancy.
Certainty (or uncertainty) can be presented several ways in assessments (Schneider et al. 1998, Moss and Schneider 2000, MA 2003:175–76). The method used
depends on the type of information involved. For richly quantitative data, a statistical approach (in other words, the familiar confidence limits in tables, graph, or text)
is possible and can be interpreted in text with special “reserved words” (see Box
4.4). For more qualitative findings, a set of agreed phrases can be used.
There are two broad philosophical approaches to estimating uncertainty using
statistics: the frequentist framework (the basis for most standard statistics) and the
146 | A Manual for Assessment Practitioners
Box 4.4. Communicating uncertainty
In the key statements of high-level summaries, it is helpful to use agreed and calibrated language to express your level of certainty and uncertainty.
For quantitative analyses that lend themselves to formal statistical treatment, or for judgments where the experts are comfortable assigning broad probability ranges, the following
reserved language can be used:
Virtually certain
Very likely
Medium likelihood
Very unlikely
Exceptionally unlikely
Greater than 99% chance of being true or occurring
90–99% chance of being true or occurring
66–90% chance of being true or occurring
33–66% chance
1–33% change of occurring or being true
Less than 1% chance of occurring or being true
For more qualitative statements, this language could be used:
Level of agreement
Amount of evidence
Agreed but unproven
Agreed but incompletely
Tentatively agreed
by most
Provisionally agreed by
Generally accepted
Suggested but
Source: Moss and Schneider 2000.
Bayesian framework. In the frequentist framework, uncertainties are derived by hypothetical repetitions of the data collection process (i.e., taking multiple independent samples). In the Bayesian framework, uncertainties are derived from the laws
of probability. In many cases the two approaches will produce the same estimate,
and the choice of technique is mainly an issue of practicality or preference.
One advantage of the Bayesian framework is that it provides a means of combining different datasets or sources of information. This can be done in two ways:
through use of a “prior,” whereby the Bayesian model updates existing information
in the light of new information, or through hierarchical models, whereby different
datasets or information sources form their own “submodels,” which are then combined at higher levels of the hierarchy to obtain estimates based on all the datasets.
Meta-analysis is one such hierarchical approach and has been very useful in combining information from a large number of studies of tropical deforestation in order to
determine the importance of different drivers of deforestation (Geist and Lambin,
Whether using a frequentist or Bayesian approach, it is important to keep in
Assessing State and Trends in Ecosystem Services and Human Well-being | 147
mind that all measures of uncertainty tend to underestimate the true uncertainty,
since by definition only the “known unknowns” can be quantified—not the “unknown unknowns.”
Assessments are often undertaken because contradictory or conflicting information exists. The assessment may not be able to resolve all of these. Some may be a result of legitimately different interpretations of the same data. Others may be because
the data are inadequate for a definitive test of one theory over another. In these cases
the task of the assessment is to be fair to all valid viewpoints, and it is the job of the
review editors to ensure that this is achieved. Being fair does not necessarily mean
equal space, and there is no need to provide a platform for discredited ideas. Assessments should not only point out disagreements. They should note where there is
consensus or near consensus as well.
4.6.3 Presenting complex information
The body of the assessment is invariably words. The use of jargon and technical
language should be minimized, and where it is unavoidable it should be supported
by a glossary. Sentences that are short and simple work best, as do paragraphs that
are short and on a single topic. In summaries for policymakers, paragraphs can
start with the conclusion and then proceed to build the case for the lead statement.
Varying the presentation of information can keep it interesting—for instance, by
putting a table or picture on each double-page spread. Bullet lists can create punchy
summaries of items that need not fall in a particular order, and numbered lists can
be used for items in order of priority. In both cases, lists of between three and seven
items are the most reader-friendly. Boxes or sidebars can provide information that
is relevant but include a level of detail inappropriate for the main flow of the text
or a slight digression. Appendixes are used for the same purpose, but more so—that
is, for material that is relevant but bulky and not essential for a top-level reading of
the text.
Tables give the reader the actual data. They are often there for reference purposes (so that someone can look up an exact value for a place or time). It is very
hard to comprehend tables with more than six elements in their totality. Readers can
be helped by providing row or column totals, highlighting certain cells, and providing guidance in the table caption about what they should especially notice. Vertical
and horizontal lines in tables should be used sparingly and judiciously, in order to
make the table uncluttered and easy to read. All columns need to be headed with a
description and the units of measure. Superscripts in the table can be linked to footnotes to supply essential detail on data sources, exceptions, or comments.
Graphics are used to show relationships and trends in the data. Scientists tend
to follow certain conventions in graph layout, in terms of how the axes are labeled
and in keeping the graphic uncluttered. Graphs with fewer than six lines have been
found to work best. Continuous variables should be presented as line graphs rather
than bar graphs. The standard deviation (or confidence interval) should appear on
the graph. Fancy effects (such as 3-D graphs, picture backgrounds, etc.) are best
avoided unless they are essential to the message. Captions for graphs and tables need
a generous amount of information. Many people read assessments by just looking
at the pictures and reading the captions, so it important to make sure they can stand
alone. It is not necessary to repeat in the text the information in a caption, and in
148 | A Manual for Assessment Practitioners
general it is not advisable to show exactly the same information both as a table and
as a graphic.
Maps and satellite- or model-derived images form an important part of many
assessments. Although this makes them expensive to print, color is often indispensable for these forms of communication. Even if it is not essential, color printing is
often worth the investment to make an attractive report. (The same applies to using
good-quality paper and the skills of a professional layout person.) The use of maps
and images in communication materials is an art form in itself, but some broad
guidelines are available. The amount of detail in an image needs careful thought.
This relates to how many categories there are in the legend, the colors or patterns
chosen to represent them, and how large the map or image will be in the final report.
It helps to reduce clutter—but without deleting essential orienting information such
as latitude and longitude grid marks, scale bars, north arrows, and place names.
Map or image legends should be sufficiently detailed that the user does not have
to search in the text for basic explanatory detail. And the images should not be so
small that they cannot be read.
Photographs and line drawings (e.g., cartoons or sketches) can often reinforce
a message very effectively and make the overall text much more readable and appealing. Photographs that include people need the informed consent of those in the
Many decision makers find the forest of technical graphs and tables unintelligible; they often consider word-based narratives (anecdotes or stories) and case
studies much easier to understand. It is important, therefore, to cater for both styles
of information acquisition. Some kinds of information, such as describing the pathway between ecosystem services and human well-being impacts, lend themselves far
better to a narrative style or diagrams than to quantification.
Alkire, S. 2002. Dimensions of human development. World Development 30 (2): 181–205.
Balmford, A., L. Bennun, B. ten Brink, D. Cooper, I. M. Côté, P. Crane, A. Dobson, N. Dudley,
I. Dutton, R. E. Green, R. D. Gregory, J. Harrison, E. T. Kennedy, C. Kremen, N. LeaderWilliams, T. E. Lovejoy, G. Mace, R. May, P. Mayaux, P. Morling, J. Phillips, K. Redford,
T. H. Ricketts, J. P. Rodríguez, M. Sanjayan, P. J. Schei, A. S. van Jaarsveld, and B. A. Walther.
2005. The Convention on Biological Diversity’s 2010 Target. Science 307:212–213.
Biggs, R., R. J. Scholes, B. J. E. Ten Brink, and D. Vackár. 2007. Biodiversity indicators. In Sustainability indicators: A scientific assessment, ed. T. Hák, B. Moldan, and A. L. Dahl. Washington,
DC: Island Press.
Carpenter, S. R., W. A. Brock, and D. Ludwig. 2007 Appropriate discounting leads to forwardlooking ecosystem management. Ecological Research 22 (1): 10–11.
CBD (Convention on Biological Diversity). 2003. Monitoring and indicators: designing nationallevel monitoring programmes and indicators. Montreal: CBD Secretariat.
CBD (Convention on Biological Diversity). 2007. An exploration of tools and methods for valuation of biodiversity and biodiversity resources and functions. Montreal: CBD Secretariat.
Cook, R. B., R. J. Olson, P. Kanciruk, and L. A. Hook. 2001. Best practices for preparing ecological
datasets to share and archive. Bulletin of the Ecological Society of America 82 (2): 138–41.
Costanza, R., R. d’Arge, R. de Groot, S. Farber, M. Grasso, B. Hannon, S. Naeem, K. Limburg, J.
Paruelo, R.V. O’Neill, R. Raskin, P. Sutton, and M. van den Belt. 1997. The value of the world’s
ecosystem services and natural capital. Nature 387:253–260.
Assessing State and Trends in Ecosystem Services and Human Well-being | 149
Dasgupta, P., and K-G. Maler. 2001. Wealth as a criterion for sustainable development. Henley-onThames, U.K.: NTC Economic & Financial Publishing.
Dasgupta, P. 2007. The socio-economics of science and sustainable development. Plenary Lecture
prepared for the International Conference on Science and Technology, Tokyo, September 7–8.
DEFRA (Department of Environment, Food and Rural Affairs). 2007. An introductory guide to
valuing ecosystem services. London: DEFRA Publications.
Geist, H. J., and E. F. Lambin. 2001. What drives tropical deforestation? A meta-analysis of proximate and underlying causes of deforestation based on subnational case study evidence. University of Louvain, Belgium: LUCC International Project Office.
Gelman, A., and J. Hill. 2007. Data analysis using regression and multilevel/hierarchical models.
New York: Cambridge University Press.
Howarth, H. B., and Matthew A. Wilsdon. 2006. Theoretical approach to deliberative valuation:
Aggregation by mutual consent. Land Economics 82 (1): 1–16.
IUCN (International Union for Conservation of Nature). 2002. 2002 IUCN red list of threatened
species. Gland, Switzerland: IUCN.
IUCN (International Union for Conservation of Nature) and UNEP (U.N. Environment Programme).
2003. World database on protected areas. Gland, Switzerland, and Nairobi: IUCN and UNEP.
Kinzig A., C. Perrings and R.J. Scholes. 2009. Ecosystem Services and the Economics of Biodiversity
Conservation, ecoSERVICES Group Working Paper, Phoenix.
Loh, J., ed. 2002. Living planet report 2002. Gland, Switzerland: World Wide Fund for Nature.
MA (Millennium Ecosystem Assessment). 2003. Ecosystems and human well-being: A framework
for assessment. Washington DC: Island Press.
Magurran, A. E. 2004. Measuring biodiversity. Oxford: Blackwell.
Moellering, H., ed. 2005. World spatial metadata standards: Scientific and technical characteristics, and full descriptions with crosstable. Amsterdam: Elsevier, for International Cartographic
Moss, R.H. and S.H. Schneider, 2000. Uncertainties in the IPCC TAR: Recommendations to lead
authors for more consistent assessment and reporting. In: Guidance Papers on the Cross-Cutting
Issues of the Third Assessment Report of the IPCC, R. Pachauri, T. Taniguchi, and K. Tanaka
(eds.), World Meteorological Organization, Geneva, 33–51.
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403:853–58.
Prescott-Allen, R. 2001. The wellbeing of nations: A country-by-country index of quality of life and
the environment. Washington, DC: Island Press.
Rees, W. E., and M. Wackernagel. 1994 Ecological footprints and appropriated carrying capacity:
Measuring the natural capital requirements of the human economy. In Investing in natural capital: The ecological economics approach to sustainability, ed. A. Jansson, M. Hammer, C. Folke,
and R. Costanza. Washington, DC: Island Press.
Sanderson, E.W., M. Jaiteh, M. A. Levy, K. H. Redford, A. V. Wannebo, and G. Woolmer. 2002.
The human footprint and the last of the wild. BioScience 52 (10): 891–904.
Schneider, S.H., B.L. Turner, and H. Morehouse Garriga, 1998. Imaginable surprise in global change
science. Journal of Risk Research, 1(2), 165–185.
Scholes, R. J. 2009. “Ecosystem services: Issues of scale and tradeoffs.” In The Princeton guide to
ecology. S. R. Carpenter, H. Charles, J. Godfray, A. P. Kinzig, M. L., J. B. Losos, B. Walker &
D.S. Wilcove. Eds. Princeton, NJ: Princeton University Press.
Scholes, R. J., and R. Biggs (eds.). 2004. Ecosystem services in southern Africa: A regional assessment. Pretoria, South Africa: CSOR.
Scholes, R. J., and R. Biggs. 2005. A biodiversity intactness index. Nature 434:45–49.
European Communities. 2008. TEEB: The economics of ecosystems and biodiversity. Brussels:
European Communities.
Ten Brink, B. J. E. 2000. Biodiversity indicators for the OECD environmental outlook and strategy.
Bilthoven, Netherlands: National Institute for Public Health and the Environment.
150 | A Manual for Assessment Practitioners
Tomich, T.P., C.A. Palm, S.J. Verlarde, H. Geist, A.N. Gillison, L. Lebel, M. Locatelli, W. Mala,
M. van Noordwijk, K. Sebastian, D. Timmer, and D. White. 2005. Forest and Agroecosystem
Tradeoffs in the Humid Tropics. A Crosscutting Assessment by the Alternatives to Slash-andBurn Consortium conducted as a sub-global component of the Millennium Ecosystem Assessment. Alternatives to Slash-and-Burn Programme, Nairobi, Kenya
UNEP (U.N. Environment Programme). 2002. Global environmental outlook 3. London:
UNGA (United Nations General Assembly). 2001. Road map towards the implementation of the
United Nations Millennium Declaration, Doc A/56/326, New York, 6 September.
Scenario Development and Analysis for
Forward-looking Ecosystem Assessments
Thomas Henrichs, Monika Zurek, Bas Eickhout, Kasper Kok,
Ciara Raudsepp-Hearne, Teresa Ribeiro, Detlef van Vuuren,
and Axel Volkery
What is this chapter about?
This chapter offers guidance on how to set up a scenario exercise and how to develop, analyze, and use scenarios within ecosystem assessments. It sets out to provide a detailed overview of all the important steps that need to be kept in mind when embarking on a scenario
exercise, providing in-depth analysis and indispensable background material on all key decisions that need to be taken.
Section 5.1 introduces scenario development and analysis as an approach to exploring
uncertain and complex future developments in a structured manner. Also, it reflects on how
a scenario exercise may support an ecosystem assessment. Section 5.2 summarizes practical considerations in setting up scenario exercises. Although such exercises are ideally tailor
made to fit their context, an exemplary approach to developing scenarios based on the socalled deductive method is outlined in section 5.3. The implications of assumptions made in
scenarios can be analyzed either by qualitative or quantitative means; section 5.4 introduces
different approaches to analyzing the implications within individual scenarios as well as comparing outcomes of assumptions about the future across sets of ecosystem scenarios. Finally,
section 5.5 reflects on the use of scenarios for scientific exploration and research, for education and public information, and for decision support and strategic planning.
The information provided in this chapter offers a starting point to planning a scenario
exercise in the context of an ecosystem assessment. However, we stress that each scenario
process is different and ideally requires a tailor-made approach. Thus the chapter does not
provide a universal step-by-step approach nor does it attempt to provide a set of authoritative
guidelines to scenario development.
5.1 How to explore the future with a scenario exercise
Section’s take-home messages
• In addition to analyzing current ecosystem state and trends, forward-looking assessments
need to explore the prospects of future developments: scenario exercises provide a
structured approach to addressing related uncertainties and complexity.
152 | A Manual for Assessment Practitioners
• Scenarios can be defined as plausible and often simplified descriptions of how the future
may unfold based on a coherent and internally consistent set of assumptions about key
driving forces, their relationships, and their implications for ecosystems.
• Scenario exercises can serve different purposes: to support scientific exploration and
research, to inform education and collaborative learning processes, or to underpin
decision processes and strategic planning.
The state and functioning of ecosystems, and thus their ability to provide services,
are subject to constant change resulting from the complex interplay of various driving forces. Ecosystem assessments need to not only provide a picture of the current
state but also look into “the future” to assess the effectiveness of options for addressing environmental change (WRI 2008). However, the aspiration to understand
“future” changes requires assessing developments that—while they may have their
origins in past or current trends—have not happened yet (and may or may not happen). As a result, assessments of the future of ecosystems need to deal with considerable degrees of both complexity and uncertainty.
The complex nature of ecosystems results from the various interactions that
govern ecosystem dynamics as well as from the multiple anthropogenic stressors
that have an impact on the environment or lead to changes in the provision of ecological services. Such stressors (or direct drivers) include pollution, climate change,
hydrological change, resource extraction, and land degradation and conversion. In
turn, these direct drivers result from long causal chains of indirect socioeconomic
drivers, such as demographic, economic, and technological developments. Finally,
changing patterns of human values, culture, interest, and power set the conditioning
framework (or ultimate drivers) for unfolding socioecological systems (MA 2005a;
see also Chapters 3 and 4). This complexity makes it important for any forwardlooking ecosystem assessment to capture as many of these interactions as possible by
using a systemic framework that includes key economic, social, and environmental
subsystems and the links between these (see Chapter 3). Another important feature
of complex systems is that changes do not necessarily occur gradually or linearly
but can be abrupt or accelerate once certain critical thresholds are crossed. Indeed,
crossing a threshold can have substantial—often irreversible—impacts on ecological
and social systems (WRI 2008).
Two principal types of uncertainty further complicate assessments of future environmental change. The first uncertainty arises from an incomplete understanding
of the interactions and dynamics within ecosystems. Through recent assessments
scientists have been able to considerably enhance their understanding of processes
within ecosystems, but it is still important to recognize that there are elements that
cannot fully be explained, and it is likely that there are dynamics within ecosystems
that the scientific community is unaware of to date. (See Chapters 3 and 4 for a
more detailed discussion of this first type of uncertainty.) As a result of this ignorance, any assessment of future changes starts off from an incomplete understanding
of the current situation.
A second type of uncertainty is the indeterminacy of all future developments.
Three distinct sources of such indeterminacy are ignorance, surprise, and volition (Raskin et al. 2002, MA 2005a). Ignorance here refers to limits in scientific
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 153
knowledge in the understanding of possible future dynamics. This can be—but need
not be—a result of complexity, and it is similar to the first type of uncertainty. Surprise is uncertainty due to the inherent unpredictability of complex systems that
can exhibit emergent phenomena and structural shifts (for example, due to their
underlying, determining socioeconomic and ecosystems dynamics). Volition is the
unique uncertainty that is introduced when human actors are internal to a system
under study and the future is subject to human choices that have not yet been made
(MA 2005a).
Scenario exercises are seen as being particularly useful for assessing the prospects
of future developments within complex and uncertain systems, such as ecosystems.
In the Millennium Ecosystem Assessment (MA), scenarios are defined as “plausible
and often simplified descriptions of how the future may develop based on a coherent
and internally consistent set of assumptions about key driving forces and relationships” (MA 2005a). This definition captures the key features of most scenarios,
although alternative definitions of what exactly constitutes a scenario have also
been put forward (see, for example, IPCC 2000, UNEP 2002, EEA 2005). Nearly
all definitions have in common that scenarios explore a range of plausible future
changes—and they usually stress that scenarios are neither predictions nor forecasts
or attempts to show the most likely estimates of future trends (see Figure 5.1).
While the remainder of this chapter focuses on describing scenario exercises
as a tool, we do note that a host of other approaches can be used—and often have
been—to address the prospect of “future” changes in ecosystems assessment. Some
of these methods can be used instead of scenario-based approaches, while others are
valuable alongside or as part of a scenario exercise. Such methods to assess future
trends include trend analysis and trend extrapolation, forecasting (e.g., Armstrong
2001), cross-impact analysis (e.g., Gordon and Hayward 1968), future workshops
(e.g., Jungk and Müllert 1987), Delphi-type expert-based estimates (e.g., Helmer
1983), role playing, gaming and simulation, and future state visioning (e.g., Stewart
1993), as well as developing future histories, science-fiction writing, or even wild
Within the field of scenarios, different types can be distinguished. Several typologies have been offered to characterize scenarios according to their key characteristics. Commonly, scenarios can be grouped according to their principal format,
the main issue or type of question addressed, the process applied to develop the
scenario, or the epistemology underlying a scenario exercise (see, for example, Box
5.1; see also Ducot and Lubben 1980, Ogilvy and Schwartz 1998, EEA 2001, van
Notten et al. 2003, Börjeson et al. 2006, Westhoek et al. 2006, Alcamo and Henrichs 2008, Wilkinson and Eidinow 2008).
Maybe the most straight-forward distinction between scenario types relates
to the format of a set of scenarios—that is, the differentiation between qualitative
and quantitative scenarios. Qualitative scenarios are predominantly presented as
narrative descriptions of future developments, commonly in the form of phrases,
storylines, or images. Quantitative scenarios, in contrast, expand on numerical estimates of future developments—presented as tables, graphs, and maps—and are
often based on the output of simulation modeling tools (Alcamo and Henrichs
2008). While this distinction appears clear cut in theory, many scenarios published
to date are hybrids of these two types: selected aspects of qualitative scenarios may
154 | A Manual for Assessment Practitioners
Figure 5.1. Scenarios can
help address uncertainty
in complex systems;
note that scenarios
differ from facts,
forecasts, predictions,
and speculations.
Source: Zurek and
Henrichs 2007
be underpinned by numerical estimates, or the numerical estimates of a quantitative
scenario may be bound together and explained by a consistent storyline. Indeed,
many of the recent international environmental assessments have developed and
analyzed scenarios that explicitly combine qualitative and quantitative information
(see EEA 2001).
An important function of scenario analysis—particularly in the context of ecosystem assessments—is that it provides an approach to reflect on and think through
the possible implications of alternative decisions in a structured manner. Simply
put, a scenario exercise offers a platform that allows decision units (individuals,
a company, an organization, or even a country) to reflect on how changes in their
respective context (that is, developments not within their immediate spheres of influence) may affect their decisions. This approach of testing whether different policy
and management approaches are robust is sometimes referred to as “wind tunneling.” However, for scenarios to be useful to “test” or “wind-tunnel” decisions in
such a manner, the individual unit’s actions and the decision context need to be
distinguished clearly—which is all the more difficult to do in complex systems, as
numerous feedbacks between the decision unit’s behavior and contextual developments exist. Methodologically, this provides a significant challenge, and many past
scenario exercises in an environmental or ecosystem context do not separate these
spheres clearly, making it difficult to use the scenarios developed by them in concrete
decision-making situations (see section 5.5).
Scenario-based approaches have evolved to be a useful and much applied approach to support environmental assessments over the past few decades (see Box
5.2). Also, ecosystem assessments have increasingly made use of scenarios to explore
the potential future implications of different approaches for sustaining ecosystem
services in the face of growing pressures. The MA, for example, offers four global
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 155
Box 5.1. Examples of scenario typologies
A first example distinguishes scenarios according to the type of question about future developments that a scenario exercise sets out to address. Three principal types can be differentiated
by these criteria.
• Reference scenarios. Sometimes also referred to as “predictive scenarios.” Generally set
out to address the question “what is expected to happen?” and include forecasts as well
as what-if analyses.
• Explorative scenarios. Attempt to map “what can or might happen?” and explore what
future developments may be triggered either by exogenous driving forces (developments
that are external and cannot be influenced by the decision makers in question), by endogenous driving forces (developments that are internal and can be influenced by decision
makers), or by both.
• Normative scenarios. Sometimes referred to as “anticipatory scenarios.” Aim to illustrate
“how can a specific target be reached?” or “how might a specific threat be avoided?” and
thus include both backcasting studies and planning exercises. Börjeson et al. (2006) offers
a more detailed discussion of these types of scenarios.
A second example groups scenarios based on the epistemologies that underpin their exercises. Again, three principal types can be differentiated.
• Problem-focused scenario exercises center on the factors shaping future developments
and usually emphasize the product rather than the process.
• Actor-centric exercises focus on the relationship of specific actors to their environment
and primarily see scenarios as a basis for strategic conversations (particularly in an organizational learning context).
• Reflexive interventionist scenario processes are developed around the interactions between various actors and their environment (and vice versa) with the aim to inform action
learning (especially in a public policy context). Wilkinson and Eidinow (2008) offer a more
detailed discussion of these types of scenarios.
scenarios based on the implications of different assumptions regarding approaches
toward governance and economic development (regionalized versus globalized) and
toward ecosystem service management (reactive versus proactive), as illustrated in
Figure 5.2 (see MA 2005a, Carpenter et al. 2006).
Besides global-level scenario based assessment, many regional or sub-global
ecosystems assessments—including several of those associated with the MA—have
either used the MA scenarios to help frame regional assessments or have developed
new regionally specific scenarios to inform their analyses (see Table 5.1; see also MA
2005b, Pereira et al. 2005, Kok et al. 2007).
5.1.3 Why, when, and for whom to develop scenarios within ecosystem
There can be several reasons to look in the future in a structured manner. Ultimately,
it can be argued, the aim is to attempt to anticipate possible consequences of current
156 | A Manual for Assessment Practitioners
Figure 5.2. Millennium Ecosystem Assessment scenarios: plausible future developments
through 2050.
Sources: MA 2005a
developments and options to either prevent, counter, prepare for, enhance, or benefit
from future changes—and to better understand the implications of the uncertainties
that surround assumptions about how the future may unfold. Supporting this overall aim, future studies in general—and scenario exercises in particular—can be used
for multiple purposes (see Jaeger et al. 2007), including:
Aid in recognition of “weak signals” of change;
Avoid being caught off guard—“live the future in advance”;
Challenge “mental maps”;
Raise awareness (e.g., about future risks or critical thresholds);
Test strategies for robustness using “what if” questions;
Provide a common language (e.g., by unveiling different perceptions and
• Stimulate discussion and creative thinking;
• Provide better policy or decision support; and
• Stimulate engagement in the process of change.
These and the other purposes underlying scenario exercises may be roughly grouped
into three overarching clusters (based on, e.g., Alcamo and Henrichs 2008).
Scenarios developed and analyzed in support of research and scientific exploration can help to better understand the dynamics of (complex) systems by exploring
the interactions and linkages between key driving forces. This cluster includes the
group of “scientific scenarios” that sets out to examine the possible long-term behavior of biophysical systems as perturbed by human influence (MA 2005a). Also,
it can offer a platform to bring together information from different research strands
Box 5.2. History of scenario development and analysis in the context of ecosystem
The formalized use of scenarios as an approach to deal with uncertainty about future developments goes back more than 50 years. While humans have always made implicit and explicit
“what-if” type assessments to guide their decisions, the earliest cited scenario studies are
game analyses and military planning exercises published during the cold war (see, for example,
Kahn 1960, Kahn and Wiener 1967). In the 1970s and 1980s the concept of scenario analysis
was further developed and used for strategic planning within businesses (see, for example,
Wack 1985). At the same time, the first scenario studies that explored environmental concerns
emerged, often based on mathematical simulation models (see, for example, Meadows et al.
1972, Mesarovic and Pestel 1974).
Scenario analysis gained increasing prominence in the context of environmental assessments during the late 1980s and the 1990s—largely prompted by concerns with climate change
and sustainable development (Raskin 2005). Various quantitative scenarios were used extensively to explore the consequences of technological and economic assumptions for energy
use, greenhouse gas emissions, and climate change (see, for example, IPCC 1992, Alcamo
et al. 1998). At the same time, more-qualitative scenarios were used to sketch out alternative
environmental pathways (see WBCSD 1997). Also, the first examples of combined qualitative
and quantitative scenarios were elaborated (see, for example, Toth et al. 1989, Gallopin et al.
1997, Raskin et al. 1998).
More recently, many—if not most—major global environmental assessment exercises
have included a scenario-based component addressing future challenges (including, for example, Gallopin and Rijsberman 2000, IPCC 2000, IPCC 2001, UNEP 2002, MA 2005a, IPCC
2007, UNEP 2007; for an overview see EEA 2001 and Rothman 2008). It is worth noting that
in recent assessments there has been some convergence toward a set of “stereotypical” scenario logics—which share the perspective on key uncertainties and assumptions about different driving forces (see, for example, van Vuuren et al. 2008):
• Economic optimism scenarios, which have a strong focus on market dynamics and
economic optimism usually associated with rapid technology development (e.g., the A1
(IPCC) or Markets First (UNEP) scenarios).
• Reformed market scenarios, which also focus on market dynamics but include some additional policy assumptions aimed at correcting market failures with respect to social development, poverty alleviation, or the environment (e.g., the Global Orchestration (MA) or
Policy First (UNEP) scenarios).
• Regional competition scenarios, which assume that regions will focus more on their more
immediate interests and regional identity, often assumed to result in rising tensions among
regions and/or cultures (e.g., the A2 (IPCC), Security First (UNEP), or Order from Strength
(MA) scenarios).
• Global sustainable development scenarios, which see a strong orientation toward environmental protection and reducing inequality, based on global cooperation, lifestyle change,
and efficient technologies (e.g., the B1 (IPCC), Sustainability First (UNEP), or Technogarden (MA) scenarios).
• Regional sustainable development scenarios, which focus on finding regional solutions for current environmental and social problems, usually combining drastic lifestyle
changes with decentralization of governance (e.g., the B2 (IPCC) or Adapting Mosaic (MA)
Table 5.1. Summary of the scenario exercises in selected subglobal assessments under the
Millennium Ecosystem Assessment framework
San Pedro de
Caribbean Sea
Coastal BC
India Local
Sweden KW and
Northern Range
Tropical Forest
Western China
Bajo Chirripo
Stated goals of exercise
(primary purpose of scenarios, see
section 5.2.2)
Main methods used
communication with stakeholders (i.e.,
education & information)
stimulate thinking about the future
(i.e., education & information)
workshops and expert work
assess influences of external forces on
local community (i.e., education &
information // decision support)
change ways of thinking about the future
(i.e., education & information)
based on “what . . .
if” questions for
management options
assessment and implications
of the past; expert
workshops and expert work
for users and decision makers (i.e.,
scientific exploration // education &
tool for planning/actions particularly
at local scales (i.e., education &
information // decision support)
prepare for surprises, information for
planning; obtain stakeholder input (i.e.,
scientific exploration // education &
information )
stimulate thinking about the future (i.e.,
education & information)
analyze natural resource management
options, future planning; enhance
participation; inform policy makers
(i.e., education & information)
visualize the future, information for policy
makers, input for models
(i.e., education & information // scientific
information for the government, input for
models (i.e., scientific exploration)
for local communication (i.e., education &
get in touch with user needs (i.e.,
education & information)
improve response options; inform policy
makers (i.e., education & information //
scientific exploration)
workshops and expert work
workshops and modeling
participatory workshops
including community
theatre (at local level);
modeling and expert
work (at basin and
regional level)
expert work
focus groups for developing
expert work (in Mae Chaem);
participatory scenarios
quantitative modeling
workshops—qualitative and
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 159
Table 5.1. continued
Stated goals of exercise
(primary purpose of scenarios, see
section 5.2.2)
São Paulo
envision the future; change bad situation
(i.e., education & information)
India Urban
share information with partners (i.e.,
education & information)
improve ecological management (i.e.,
scientific exploration // decision
Main methods used
assessment and implication
of the pilot expert
individual consultations;
literature review
initial expert assessment and
scenario development;
participatory scenarios
workshop; scenario
redrafted by experts
Information in the original table was based on information from specially designed questionnaires.
Source: Based on MA 2005b.
and scientific disciplines to better examine the complexity of ecosystems or to explore whether systems are likely to cross critical thresholds beyond which changes
are irreversible—and the implications of doing so. When the principal purpose of
a set of scenarios is to support scientific research, it is important to ensure that
the procedure for building them unequivocally conforms to good scientific practice
and that the assumptions behind scenarios are scientifically plausible (Alcamo and
Henrichs 2008). A large number of examples of this type of scenario exercises can
be found in the scientific literature. Prominent international examples include the
scenarios elaborated by the IPCC (IPCC 2000, IPCC 2007), the Millennium Ecosystem Assessment (MA 2005a), and the Global Environmental Outlook (UNEP 2002,
UNEP 2007). (See also section 5.5.2.)
In the wider context of education and public information, scenarios can provide an approach for structuring, conveying, and illustrating differing perceptions
about unfolding and future developments. At the same time, scenarios can help to
highlight and explain the implications and long-term consequences of current trends
and choices that may lie ahead. This cluster combines two groups of scenarios suggested in MA (2005a)—“new conversation scenarios” (those aimed at exploring
new and unknown topics that be used as an educational tool for wide audiences,
de facto offering a tool for collaborative learning exercises) and “groups-in-conflict
scenarios” (those used to help understand differences in worldview and perceptions
of groups and to jointly explore consequences of actions). If the main goal of building scenarios is to inform the general public or a particular target group, then it is
particularly important that the scenarios are perceived to be credible, stimulating,
thought provoking, and—most important—relevant to the audience (Alcamo and
Henrichs 2008). (See also section 5.5.3.)
Scenarios also have a long history in decision support and strategic planning
(see Leemhuis 1985, Ringland 2002)—to solicit views and opinions about expected
160 | A Manual for Assessment Practitioners
future developments, to “test” different options for decision units to respond effectively to changing decision contexts, to evaluate the implications of specific decisions, to help prepare for risks and trends, or to analyze the trade-offs related
to specific future pathways. Especially in the context of business and private enterprises, “business strategy scenarios” have been used to explore uncertainty in a
decision context that an individual decision unit does not control, in order to test
the robustness of options and to identify opportunities and challenges (MA 2005a).
Many of these scenarios remain internal to the organization that develops them (as,
especially in the realm of business, scenarios may give a competitive advantage; see
Wack 1985, Schwartz 1991). By contrast, “public interest scenarios” or “strategic
conversation scenarios” are commonly necessarily open to wider debate. These aim
at shaping the future by articulating a common agenda and language between actors and highlighting potential actions and their consequences (MA 2005a) and by
“unearthing” the assumptions about future developments that often guide decision
making implicitly. In the field of public policy, Ringland (2002) shows a number of
examples of scenario exercises that were directed to support policies at local scales.
Furthermore, it is worth noting that in the context of environmental change, scenarios originating from a research context have played a major role in supporting
decision-making processes (see also section 5.5.4).
Ecosystem assessments generally set out to be “a social process to bring the
findings of science to bear on the needs of decision makers” (MA 2005a)—in other
words, they aim to contribute to varying degrees to all three clusters just described
(see also Box 5.3). As noted in Chapter 1, assessments have been seen to have the
most impact when they are perceived to combine three characteristics: saliency to
potential users (“Is it relevant?”), credibility with regard to use of scientific methods
(“Is it sound and convincing?”), and legitimacy in the way the exercise is designed
and conducted (“Is it inclusive and unbiased?”). (See also Mitchell et al. 2006.)
Alcamo and Henrichs (2008) argue that this is true also for scenario exercises, particularly those developed in the context of an assessment process.
Ideally, any scenario exercise would thus aim to be relevant, credible, and legitimate at the same time. Furthermore, scenarios can benefit from aiming to be creative
and to challenge prevailing expectations and worldviews (see Alcamo and Henrichs
2008). Practically, however, time and resource limitations often require setting priorities in meeting one or two of these characteristics rather than all of them. But also
conceptually these notions may be somewhat mutually exclusive within scenario
exercises: in many circumstances it may be difficult to be fully inclusive and still be
convincing (for example, when different scenarios cater to contrasting worldviews
and beliefs). Or when a scenario exercise is designed to explore the implications of
possible future surprising events, this may entail assessing nonlinearities that cannot
be constructed using scientific methods alone. Thus while each of the above characteristics is worthy of pursuit in its own right, there may be particular need to emphasize one of them depending on the main purpose of the scenario exercise—and to be
aware of what trade-offs between these concepts are acceptable. (This is discussed in
more detail in section 5.5.)
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 161
5.2 How to set up a scenario exercise
Section’s take-home messages
• A scenario exercise can have both process- and product-related outcomes: typically such
exercises lead to an intense learning experience for those involved as well as a set of
scenarios for further analysis and use.
• It is crucial to clearly define the purpose and goals of a scenario exercise at its outset,
as this will affect the set up, scope, and planning of the exercise—it will also guide the
scenario development process and the usefulness of the outcomes.
• Key aspects to consider when setting up a scenario exercise are—among others—its
context, timing, budget, degree of stakeholder participation and expert involvement,
desired complexity, and geographical scale of analysis.
• Scenario exercises are complex processes and require both an enabling authorizing
environment and a sound organizational set up: The value of involving experienced
scenario practitioners when designing the process cannot be emphasized enough.
There is no “one-size-fits-all” approach to conducting a scenario exercise. How a
scenario exercise is set up and carried out very much depends on its context, goals,
participants, and so on. Nevertheless there are a number of stages and steps common to many, if not most, exercises. The following sections describe some of the key
steps and offer a basic approach based on ecosystem assessment experiences with
scenarios. However, the steps of the modular process outlined here can and should
be adapted to fit specific needs. The way to carry out a scenario exercise depends
a great deal on its context; and to avoid being bogged down by the confusion and
preconceptions that often surround discussions about the future, it is thus crucial to
be clear about the goals and the set up of the exercise.
Generally speaking, careful planning in the early stages of a scenario exercise
significantly improves the quality of any scenario development process and its outcomes. Putting together a plan at the outset will help guide the development process
through the identification of the specific goals, steps, and resources that are necessary for a meaningful analysis. Nevertheless, any plan should offer enough flexibility
for revisiting and adjusting any of the steps outlined here, as much will be learned
throughout the process, and the approach decided on in the initial phase should not
be seen as static (Jaeger et al. 2007).
Before embarking on a scenario exercise to assess the future of ecosystem services and human well-being, a few practical issues should be considered, which are
explored in the next five sections:
• What to expect from a scenario exercise—possible outcomes;
• How to frame the purpose of the scenario exercise and decide on the scenario
• How to define the scope of a scenario exercise;
• How to establish an authorizing environment and project team for a scenario
exercise; and
• How to facilitate participation throughout a scenario exercise.
Box 5.3. Reflections on the role of scenario exercises in ecosystem assessments
The eventual focus of a scenario exercise very much depends on when in an assessment process it is performed. It is useful to distinguish four principal settings. The four roles described
here are not mutually exclusive—but clarity about the primary aim of a scenario exercise within
an ecosystem assessment is essential for ensuring that scenarios are developed, analyzed,
and used in the best possible fashion.
First, a scenario exercise at the outset of an ecosystem assessment may be a useful
process to help participants develop a common understanding and framework about those
dynamics that shape future ecosystem developments. In particular, it may aid in arriving at a
common language about current trends, threats, opportunities, and options. Such a scenario
process should ideally involve key experts and preferably also important stakeholders in the
outcome of an assessment. However, it may not be necessary to embark on a full scenario
exercise, as the key aim of this exercise is to arrive at a shared way of thinking about future
Second, a scenario exercise may be run in parallel with other activities within an assessment—with the aim of having the different components of an assessment inform each other. In
the Millennium Ecosystem Assessment, the Scenarios Working Group worked at the same time
as the Current Status and Trends and the Policy Responses Working Groups, so that ideas
developed in the Current Status and Trends Group helped shape the discussion of drivers in
the Scenarios Working Group, and vice versa. However, running a scenario exercise in parallel to other activities may make it difficult to incorporate key findings across different Working
Groups unless sufficient time is set aside for iterations between them to allow fully exploring
and harvesting respective findings.
Third, if an assessment is structured in a more linear fashion, a scenario exercise can
provide a link between a discussion on current trends and the potential of policy responses to
counter undesirable developments. In such cases, a scenario exercise would build fully on the
outcomes of an analysis of the current state and trends within an ecosystem and adapt—to the
extent possible—the framework, structure, and language developed. Similarly, the viability and
robustness of policy responses would be tested against the backdrop of the different scenarios
developed. This would greatly help achieve overall consistency across different parts of an assessment; it also entails the risk of limiting discussions.
Fourth, a scenario exercise may be limited to supporting outreach activities only, using the outcomes and findings of an assessment to develop scenarios at various scales with
stakeholders and decision makers, molded around their immediate concerns. As an example
of such a setup, an assessment would describe current trends and outline possible policy
responses—and based on the wealth of information developed there, different future scenarios
might be explored. In this case, the discussion of scenarios describing future developments is
not a part of the actual assessment; rather, the assessment is the input in one or more followup scenario exercises.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 163
5.2.1 Setting up a scenario exercise—what outcomes to expect
Generally speaking, scenario exercises have two main types of outcomes: a process
(i.e., a learning experience for actors involved in the scenario development) and a
product (i.e., the actual set of scenarios themselves, which can be used in different
Outcome 1: The process—benefits of the scenario development process
Scenario exercises can have three primary process-related benefits. First, those who
participate in a scenario development process gain a better understanding of interactions, assumptions, and trade-offs related to ecosystem services and human
well-being. Some of the direct outcomes of scenario exercises are an increased understanding of the linkages between the different parts of the socioecological system,
the identification of beliefs and assumptions about how a policy or a chosen development pathway may alter some or all of the system, the identification of potential
long-term consequences for ecosystem services of choices made in the near future,
and the identification of factors important for a successful outcome of a decision.
Second, scenario exercises create a platform to talk across interest groups, disciplines, and philosophies. Uncertainty about the future has an equalizing effect:
no one discipline or sector can predict the future. Scenario work requires scientists, governments, and citizens to collaborate and piece together plausible stories
about what might occur in the future. The result is a process that can accommodate
thoughtful, creative, and nonthreatening discussion about topics that might otherwise be politically charged. Less powerful groups can be empowered through such
a process, and more powerful groups can gain invaluable insight into how their
practices and policies affect others. Providing space for multiple forms of knowledge
(e.g., traditional and practitioner) can lead to deeper and more nuanced reflections
on social and ecological change within a system.
Finally, the discussion of and reflection on different scenarios can create the
grounds to reveal conflicts, exchange information, and help to build consensus or
at least an understanding over controversial issues related to ecosystem services and
the choice of polices for sustaining services. Indeed, scenario exercises can be used to
air conflicts or build consensus among diverse stakeholders over what a desirable future might look like. Managing natural resources often involves trade-offs between
different values and economic activities. Getting stakeholders around the same table
to discuss their visions of future land management or economic development helps
build understanding of these trade-offs and agreement on appropriate policy. When
developing and discussing a set of scenarios jointly, hidden values and assumptions
are uncovered, highlighting potential shared values and the root of conflicts. Taking stakeholders away from the present day to focus on possible futures facilitates
discussion, allowing participants to develop a greater understanding of each other’s
point of view. While there is no guarantee that increased mutual respect will carry
over to resolving current conflicts, it increases that possibility.
164 | A Manual for Assessment Practitioners
Outcome 2: The product—key elements of scenarios
Most scenarios, including those developed in an environmental and/or ecosystem
context, tend to have a number of key elements in common: a representation of the
initial situation, key driving forces, a description of step-wise changes, and image(s)
of the future (see, for example, EEA 2001 or Alcamo and Henrichs 2008).
The first key element of any scenario is an understanding of and representation
of the initial situation of the system, including an understanding of how past trends
have shaped the current state. This information may be gathered or developed as
input into a scenario exercise or as part of it. The description of the current state of
the system can stem from other parts of an ecosystem assessment. In the Millennium
Ecosystem Assessment, for example, information from the Condition and Trends
Working Group, which described the prevailing conditions and trends of ecosystem services and human well-being and their relation to past and current drivers of
change, was used to inform the scenarios developed.
Driving forces are the main factors that influence future developments and dynamics of the system a scenario focuses on. Main categories of driving forces in
scenario exercises focusing on ecosystem services include the so-called STEEP drivers (social–cultural, technological, economic, environmental, and political driving
forces). It has proved useful to distinguish between direct driving forces, also known
as structural drivers (those that unequivocally influence the system) and indirect
driving forces (those that alter the level or rate of change of one or more direct drivers). Another practical distinction is between endogenous drivers (which are in the
control of the decision makers in a given system) and exogenous ones (which are outside of the decision makers’ control). (Also see Chapter 3 for more information.)
A further important element of any scenario is a plausible and consistent description of step-wise changes that are assumed to unfold in the future. These changes
are based on assumptions about how key driving forces develop and interplay, and
how this affects the state of a system at different points in time. These changes can
be depicted as numbers and figures (see section 5.3) or as sets of phrases, illustrative vignettes, and/or detailed storylines. The number of time steps described within
a given period may vary according to the focus of the scenarios developed and the
need for analytical underpinning (see Alcamo and Henrichs 2008).
Finally, one of the end products of a scenario exercise is a description of an image or several images of the future, describing in great detail what the future may
look like as a consequence of the assumptions made about drivers and their stepwise changes. Often, the system’s “end state” (its state at the time horizon of the
scenario) is presented in the form of a narrative description (which may be more or
less extensive, depending on the objectives of the exercise). Indeed, illustrated and
narrative descriptions of the future have often proved useful for communicating the
outcome of scenario assumptions to a wider audience (see section 5.5).
5.2.2 How to frame the purpose of a scenario exercise
Scenario exercises are much more likely to be useful if their purpose is clearly identified and spelled out right from the beginning. The purpose will guide how the process
should be organized, who should be involved in the exercise, and what the scenario
should focus on. Because it is possible that many of the audiences and participants
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 165
in a scenario exercise will be unfamiliar with such efforts, it is important to have
clarity about the main goals and outcomes in order to communicate effectively with
potential participants and end users, achieve good buy in, and encourage effective
participation. More important, clarity about the goals of a scenario exercise will
help to determine what type of scenario is needed, to what degree the scenario will
need to be analyzed using simulation tools, to what level an exercise needs to be
geared toward strategic conversation rather than detailed analysis, to what extent a
set of scenarios should aim at covering expected trends versus being mind-stretching
exercises, and so on. The purpose(s) of the exercise may evolve or be expanded, but
a clear starting point is necessary.
Thus, at the outset those in charge of putting together the scenario exercise will
need to discuss and decide on specific goals of the exercise. As part of this process,
stakeholders and decision makers might be interviewed to help focus the goals further. Through examination of the goals of the assessment and consideration of the
resources available (including time, funds, and expertise) as well as the needs and interests of the different stakeholders, the assessment team and principal stakeholder
groups can decide which general type of scenario exercise they will develop.
Questions that may be useful to determine the goals of a scenario exercise
• Why is the scenario exercise being initiated?
• What stakeholders are most interested in the scenarios component of the assessment and what kind of information are they interested in?
• What should be gained from developing scenarios in terms of concrete actions?
• What policies/plans/projects will be informed by the scenario exercise?
• What type of scenario (for example, qualitative or quantitative scenarios) is required to produce information that is useful to the principal stakeholders?
• What limitations are relevant to this exercise in terms of time, expertise, and
The answer to these and similar questions very much depends on the general context
of the exercise—that is, whether it is geared toward scientific exploration, education and information, or decision support. The general context can guide what type
of scenario exercise might usefully be developed, and along what parameters. The
following paragraphs present some ideas on how the scenario exercise might be
developed depending on which context it fits into. In practice, scenario exercises
will often have several goals and therefore may fit more than one context, and the
process should be adjusted accordingly.
In a scenario exercise that aims to support research and scientific exploration
of future links between ecosystem service trends and human well-being, the focus
is more strongly on the scenarios themselves rather than the process. The participation of people from a range of scientific disciplines—and to some extent, from
stakeholder groups—is key to gaining new insights into the links between ecosystem
services and human well-being. In order to complement the assessment of current
and past conditions and trends, it may be important for the scenario process to
be coordinated time-wise with the rest of the assessment so that accurate information about ecosystem service trends can be used as a starting point from which
to develop scenarios. (This is also true when the goal of the exercise is to inform
166 | A Manual for Assessment Practitioners
decision making.) Several brainstorming sessions with a large scenarios team can
help promote the creative development of storylines. The scenario team can try to
incorporate more complexity into the scenarios and examine cross-scale interactions. Particular attention should be paid to thresholds, risks, probabilities, and the
possibility of surprise in the system of interest. End results may include qualitative
storylines and quantitative models; the identification of key scientific uncertainties,
risks for human well-being, and the sustained flow of ecosystem services; and future
research needs. (See also section 5.5.2.)
In a scenario exercise that aims primarily to provide education or (public) information, broad participation is crucial. A range of stakeholders and decision makers needs to be involved, and the role of scientist and researchers may be limited
to providing input rather than being involved in the actual scenario development
process. Discussions might be aimed at challenging mental models, creating buy in
for the assessment, or communicating with key decision makers in a nonthreatening
setting. Stakeholder workshops can be used to discuss the system of study and possible future trajectories in a way that is engaging and emphasizes thinking “outside
the box.” Such an exercise can also be empowering for marginalized stakeholder
groups through knowledge sharing and building adaptive capacity for dealing with
ecosystem change. Scenario sessions can be run strategically by a neutral facilitator
to ensure a tight focus, clear understanding on the part of the participants, and as
much active participation as possible. Inspiring texts, theatre, art, and public forums
can be used as alternative formats to spread the results to a larger audience. (See also
section 5.5.3.)
Finally, in scenario exercises geared toward decision support and strategic planning, a greater degree of caution is required in order to develop scenarios that are
aligned with the problems of interest to decision makers and at the same time based
on a process that is perceived to be legitimate. (Key questions here are “who has developed the scenarios?” and “with what agenda did they do so?”) Policy or normative scenarios may support decision making by providing an analytical framework
for finding the most suitable policy options with regard to a specific policy target.
(See also Box 5.4.) Exploratory scenarios may aim at providing a backdrop for
“strategic conversations” that challenge and sharpen the mental models of decision
makers (in this, the purpose does not differ vastly from the goals of education and
information). Either way, in the context of decision making it is crucial to engage
in a constant dialogue with those who take the decisions. It is worth noting that
the effectiveness of scenarios in this context is particularly dependant on their acceptability to the dominant decision culture. Ideally it is the decision makers and
stakeholders themselves who drive the scenario development process, with scientists acting as “resource people.” However, in some situations—in particular at the
global scale—this is not possible. Separate scenario sessions with different groups of
experts and decision makers may be a strategic, alternative way to develop plausible
and useful scenarios and ensure cross-fertilization without overly taxing the time
of decision makers. It is important to iterate several times between groups in order
to provide the specific information desired by decision makers. (See also section
After discussing the general goals of the scenario exercise, an assessment team
can put together a list of process parameters and objectives that seem to be important for achieving the goals of the exercise, without getting into too much detail
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 167
Box 5.4. Scenarios in the policy cycle
In the case of scenario exercises geared toward supporting policy-making processes, the goals
and design will depend on when in the policy process the scenarios are developed. The role of
scenarios varies according to the different phases of a policy cycle (see figure) and may include
identifying early warning signals and evaluating evidence during the phase of problem definition; illustrating potential consequences if a problem remains untreated, which can be useful in
the phase of agenda setting; or evaluating the feasibility of policy options that are compatible
with the phase of policy formulation.
Figure. The policy cycle is a “heuristic” framework that breaks down the policy-making
process into several phases (Source: EEA 2006)
(for example, determining whether an exercise should be more or less quantitative,
should involve high-level decision makers, and so on). There are no rules about how
all the components fit together; instead, members of the scenario team should familiarize themselves with different options and decide how to best meet their goals.
An additional important consideration that might affect the deliberations relates
to the question of what comes after the scenario exercise. All too often this is only
considered after a set of scenarios has already been developed and analyzed. However, this question is more usefully posed at the start of the scenario planning phase
in order to determine why and how scenarios will be used and therefore how they
might be developed effectively. Sometimes the purpose of the exercise is to come to a
better understanding of the problem under investigation, and no follow-up action is
needed. The same could apply if the core objective was to bring together a group of
stakeholders and initiate a common learning process. However, more often than not
a scenario exercise is not an end in itself but also a means to support other activities
168 | A Manual for Assessment Practitioners
and processes, and the eventual scenarios will be used in the context they were designed for (and sometimes beyond). (See section 5.5.)
5.2.3 How to define the scope of a scenario exercise
Once the general purpose and the specific goals of a scenario exercise have been
established, it is useful to develop an overview plan. It is often unclear to assessment
teams that have not undertaken scenario work previously how much of an investment of time and resources is necessary to complete a full exercise. This usually
depends on how complex the scenarios are to be, whether there will be a modeling
element to them, what the scenario time horizon will be, and how participative the
exercise will be. Discussions with participants over what a scenario and scenario
exercise is, what will be included in the scenario exercise, and what will not be included should be initiated at the beginning of the exercise in order to clarify what
the exercise can and will achieve. Scenario exercises are not usually linear, and they
require much iteration between steps to produce the desired outcomes.
Determining the scope of the exercise involves making decisions about a number
of issues:
How long will it take to develop the scenarios?
At what point in the larger assessment will the scenarios be developed?
How much does a full scenario exercise cost?
How to involve stakeholders?
How to involve experts?
Which issues to focus the scenario exercise on?
What time horizons should the scenarios address?
What spatial scale should the scenarios address?
What balance to strike between qualitative and quantitative analysis?
The duration of a scenario exercise may depend on how quantitative the analysis
will be and if participants rely on modeling or workshop discussions. If the time
available to complete an exercise is extremely short, planners might consider using
and/or adapting existing scenarios from other exercises (as described later in this
chapter). In general, several months are required for preparation and development
of a scenario exercise at the sub-global scale. If many stakeholders are involved,
much more time may be needed just to get everyone together. In the MA, the exercises ranged from one participatory workshop that lasted a weekend (but required
several weeks of preparation and follow-up work) to a series of short workshops
that spanned several months.
The timing of a scenario exercise largely depends on the information it depends
on and how outcomes will feed back into the rest of the assessment. While scenarios can be useful at any point in the larger assessment process, most sub-global
assessment teams in the MA initiated scenario exercises toward the end of their
larger integrated assessment process. This had the advantage of having the assessment of conditions and trends of ecosystem services provide a starting point for the
scenarios. However, many of these exercises subsequently found that it would have
been preferable to start them a bit earlier in order to inform the main assessment at
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 169
the same time as being informed by it. The benefits of holding a scenario exercise
early in the process include jump-starting a positive participatory process around a
creative and nonthreatening activity, the early identification of stakeholder concerns
and priorities, and the identification of uncertainties that require further assessment
by the core assessment team. One sub-global assessment under the MA process (in
Wisconsin, in the United States), for example, began the integrated assessment with
a scenario exercise. This allowed them to identify key uncertainties and dynamics in
their system with the participation of stakeholders, and these findings could then be
used to focus the larger assessment (Peterson et al. 2003).
The available budget will often dictate whether meetings and workshops can be
held and therefore how much stakeholder participation is possible. Scenario workshops commonly require a minimum of two sessions of one or two days each, but
several additional workshops would be better and allow for more in-depth discussion. Local participants may be willing to set aside a day at no cost (other than
meals and logistics), but a regional or global exercise may need to fund travel and
accommodation. Funding may also be needed to cover the time of experts involved
in analyzing the scenarios. Particularly in model-based scenarios, analysis can become resource intensive, as described later. Budgets will thus vary widely depending
on the context of the exercise.
The level of stakeholder involvement may have an impact on the timing of the
exercise and the complexity of the scenarios. How many stakeholders to involve,
and in what capacity, strongly depends on the goal of the scenario exercise. This is
discussed in more detail later in this section.
The level of expert involvement depends on what kind of scientific and local expertise is available to help develop the scenarios. Generally speaking, expert involvement may strengthen an exercise’s credibility—but it does not guarantee it. Experts
may be involved as core members of the scenario team, or they may be able to offer
feedback on the developed scenarios. Another important consideration in this context is whether experts need to be paid for their involvement—again this depends
largely on the exercise context.
The degree of complexity of the scenarios depends on how many themes and/
or ecosystem services and/or drivers of change and/or indicators will be explored.
Particular ecosystem services and drivers of change will already have been identified
as especially relevant to an assessment context during the assessment of conditions
and trends in ecosystem services and human well-being. (See Chapter 4.) This identification process is also part of developing a conceptual framework that contains
the main factors and services and their relationships. The conceptual framework can
then be used as a tool to guide the scenario development as it provides a unifying
way of thinking and conceptualizing many relevant issues.
The time horizon of a scenario should be based on what is a reasonable amount
of time for the main issues of concern to be explored or managed. “Time horizon”
here refers to the period of time over which the scenarios will be allowed to unfold.
While 30 years might be an appropriate timeframe for scenarios developed to explore land use and land planning, more time might be needed to examine climate
change impacts and policies. Time horizons also have political implications and cannot always be selected in advance of the initial exploration of issues with stakeholders about policy cycles and information needs. In the Millennium Ecosystem
170 | A Manual for Assessment Practitioners
Assessment, for example, scenarios were explored against the time horizon of
The geographical scale for a scenario exercise is not always easy to determine,
particularly if complex driving forces interact across geographical or organizational
scales and levels. (See Box 5.5 for additional guidance on the issue of scales.) In most
cases the geographical scale of the assessment (e.g., global or local) also determines
the scale for the scenarios. Nevertheless, scenarios developed at a specific scale can
also aim to address multiscale issues and relationships—especially given that recent
work on socioecological systems stresses the need for understanding processes at
multiple scales and in particular their interaction across scales. Further information
on how to link scenarios across scales can be found in section 5.3.
The primary mode of analysis—that is, quantitative or qualitative—will greatly
depend on whether sufficient data and modeling tools are available, whether the
budget allows for extensive quantitative analyses, and what kind of scenario outcome will be most suitable for the scenario exercise’s purpose. In a scientific context,
for example, scenarios may benefit from quantifying as many relevant parameters
as possible. However, if no one is available to develop adequate models to support a
scenario analysis, then a more qualitative approach will be the only option. In an educational or informational context, a scenario exercise’s outcomes may be conveyed
more easily using narratives and qualitative analysis, making detailed quantification
superfluous or even counterproductive. Thus one approach toward answering this
question for a specific exercise is to read published accounts of previous scenario
exercises and compare the described process and outcomes with your own process
and information needs.
There are additional questions that the assessment team may need to consider
depending on the particularities of the scenario exercise. In some cases the team will
wish to explore specific international environment or development targets; in other
cases the focus will be on alternative development paths at local and regional scales,
examining trends in a few important ecosystem services only. These ideas have to be
reconciled with logistical limitations and the specific needs of important end users
of the scenarios. The assessment team can hold a scoping meeting with key scenario
end users to identify or refine a list of important ecosystem services and drivers of
change and to decide on the most relevant spatial and temporal scales. The output
would be a detailed overview plan of the scenario-building process that will serve as
a flexible blueprint for the exercise.
5.2.4 How to establish an authorizing environment and project team for a
scenario exercise
A scenario exercise’s authorizing environment refers to the level of support offered
for the scenario process and products by stakeholders and funding agencies. Setting
up an authorizing environment for an ecosystem assessment in general is covered in
Chapter 2, and many of the considerations highlighted there apply also to scenario
exercises. Developing a strong authorizing environment revolves around building
mechanisms to ensure the credibility, legitimacy, and relevance of the scenario exercise and its outcomes—which depends on how the exercise is run and how stakeholders are treated. It is therefore beneficial for a scenario exercise if stakeholders
participate in or are consulted throughout the process to ensure a continued match
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 171
Box 5.5. Why worry about scale?
One issue that most scenario exercises in ecosystem assessments have to deal with is how to
treat processes that play out at multiple scales. Scale refers here to the physical dimensions,
in either space or time, of phenomena or observations, whereas level is used to describe the
discrete levels of organization, such as individuals, households, ecosystems, or agroecological zone. Components of any complex system are structured hierarchically in space and time
across scales and levels—and hierarchy theory suggests that the best way to deal with problems in a multiscale complex system is to understand how the elements of the system behave
at a single time–space level.
Geographical scale simply refers to the boundary of the case study and—when models are used—the spatial resolution of the grid cells. These do not always overlap with the
boundaries of the socioeconomic or organizational levels (i.e., the functional scale of relevant
processes). Particularly in multiscale scenario exercises, selecting the appropriate temporal
scale is very important. There is a general tendency for processes to become slower as the
geographical scale expands. As a result, global assessments often consider a temporal extent
of 2050 to 2100, especially when climate change is one of the processes of interest. In contrast, local studies often consider a time horizon of 10 years or even shorter, being in line with
the policy cycle.
More often than not, within any case study important drivers and processes cover multiple
temporal and geographical scales. It is therefore important to consider them accordingly when
developing scenarios. The simplest way to consider multiple scales is by executing a single
scale assessment and including the most important drivers from higher scales as external drivers. The essential difference to a multiscale assessment is that drivers at only one scale are
dynamic. In other words, if the majority of all-important drivers are active at one scale—and
exogenous drivers can be considered constant over space and time—a single scale assessment can be sufficient. Conversely, if the aim is to address drivers from different scales, while
explicitly including changes in and between those drivers, a multiscale study is appropriate.
between their needs and the exercise outcomes. A few issues particular to setting
up an authorizing environment for a scenario exercise—beyond those indicated in
Chapter 2—also need to be taken into account.
How an appropriate authorizing environment is established will again very
much depend on the context of a scenario exercise. If the context is research and scientific exploration, a structured and transparent scientific approach will be needed
that includes identifying levels of uncertainty associated with future trends, based on
what is known from the literature about ecosystem services and human well-being.
For decision support, the relevant decision makers should be involved in the process
from the beginning and recognize its legitimacy. In many cases, it is very useful to
have stakeholders involved in describing the decision-making context central to the
scenarios. For information, communication, and learning processes, a broad range
of stakeholders may be included in the process and be given a chance to contribute
meaningfully to either the storyline development or the interpretation of the scenarios. A lot of time usually needs to be allocated to discussions among participants
of a scenario exercise (e.g., stakeholders and scientists) to promote learning and
ownership of the process and outcomes.
172 | A Manual for Assessment Practitioners
With the assessment team and key stakeholders, it will be useful for the process
and for the end users to put together a short document outlining what steps will be
taken toward establishing an appropriate authorizing environment that will lead to
a credible, legitimate, and relevant scenario process and outcomes. In setting up an
authorizing environment for a scenario exercise, the following might be considered:
• Identify which organization or group of people is convening the scenario
exercise. Are they trusted by the stakeholders who are involved? Is there a
trusted, neutral facilitator running the exercise?
• Identify participants that represent the groups that the exercise is designed to
reach. What will be their role in the exercise?
• Identify other people who may be affected by the changes being explored in the
scenario exercise. What is their role in the exercise?
• Identify people who understand the system in question and who could help
build credible scenarios. Are these people well respected by the relevant decision
makers and scenario end users?
• Identify people from different disciplines and sectors who could review the scenarios and/or offer informal feedback on the process and outcomes.
• If possible, make sure decision makers are included in the actual exercise. If the
decision makers do not have the time to participate in workshops, might they
participate in another manner? Consider interviews, as well as regular briefings
about the process.
• Are the scenarios being built around the expertise of the scientists involved or
around the needs of the decision makers? This balance may need to be adjusted
if the former is being promoted (unless this is the main focus of an exploratory
exercise). Note that the needs of the decision makers may change as the process
develops and more about the system is learned.
Planning, organizing, and facilitating a scenario exercise is a lot of work and requires a group effort, which commonly a core scenario team will be tasked with.
Therefore, an important early step when preparing for an exercise is to establish
such a team and assign responsibilities and tasks within it. The core team should include a coordinator who provides leadership and ensures the smooth running of the
whole exercise. If the scenario exercise is part of a larger ecosystem assessment, it is
advisable to include some members of the larger team alongside the experts and key
stakeholders who will work only on the scenario. Building scenarios benefits from
the inclusion of people representing different disciplines, different spheres of knowledge, and different societal roles, but the composition and balance of the team will
depend on the goals of the exercise. As scenarios within ecosystem assessments will
be broadly focused around the links between ecosystem services and human wellbeing, it is important to aim for a broad representation of scientific disciplines in the
core team, including both natural and social sciences. The scenario exercise will be
organized and driven by this core scenario team, often inviting a broader group of
stakeholder representatives to participate in the scenario development process. The
degree to which this latter group participates is discussed in section 5.2.5.
The scenario team’s organizational set up can range from very inclusive to formal and strategic, and this too depends on the purpose of the exercise. If the goal
of the exercise is outreach, a broad and inclusive scenario team may be desirable.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 173
If the goal is scientific exploration, a smaller group of creative people from a range
of scientific disciplines may be preferable. If the goal is decision support, again
the group might be smaller, but it could include the key stakeholders who will
inform the development of the scenarios according to their needs. Also, the focus
of the scenario exercise will affect who might be on the scenario team. Scientists
and practitioners who have a deep understanding of the different components of a
system and how they are interconnected may effectively inform complex and nuanced narratives about possible futures of a region. Linked, multiscale scenarios
addressing ecological dynamics may require the participation of scientists with
specific technical skills. Assembling the scenario team as early as possible in the
larger assessment process facilitates coordination of the process and information
flows between the scenario exercise and the rest of the ecosystem assessment work
(Evans et al. 2006).
Scenario exercises that rely heavily on stakeholder input or on intensive interactions between scientists and stakeholders benefit tremendously from including a
facilitator as a member of the scenario team. The facilitator is tasked with moderating meetings and encouraging effective participation within the exercise. Generally,
a facilitator should be someone who has run a scenario exercise before or who has
taken the time to learn from other people’s experiences and developed (and perhaps
tested) a plan for the exercise. In some cases, it may be important for the facilitator to
be able to communicate with people in their own language and style, while in other
cases a facilitator from a more “neutral” background may be preferred. Being very
familiar with the steps in a scenario exercise is of primary importance. Evans et al.
(2006) provide many tips on facilitation within diverse groups of participants. Some
key abilities in a facilitator include being able to paraphrase and summarize participants’ inputs to clarify main points, encouraging participation through comments
and body language, having strategies for dealing with difficult participants, being
able to move the process along without sacrificing meaningful discussions, capitalizing on the diversity within a group by encouraging discussions about differences in
perspectives and worldviews—and, most of all, being an unbiased moderator.
Other tasks within the core scenario team include:
• Keeping track of communication between the team and the stakeholder groups
and making sure that the team is meeting the expectations of the end user;
• Planning logistics, running workshops, and putting together background reading
• Taking notes at meetings, discussions, and workshops; and
• Developing illustrations, vignettes, and graphs to summarize important findings
of the exercise for communication purposes.
5.2.5 How to manage participation throughout a scenario exercise
The many benefits of ensuring effective participation in scenario development processes are increasingly recognized, especially in scenario exercises geared toward
supporting decision making or strategic communication. While scenario exercises
have often been viewed as simply being a procedure for developing a product, they
are now seen and used as a process that aims to involve the potential users of the
scenarios (see section 5.5).
174 | A Manual for Assessment Practitioners
An important reason for involving stakeholders in scenario development is to
enhance the legitimacy and the potential impact of scenarios. This can be a crucial factor in the usefulness of scenarios to support public decision making. At the
same time, stakeholder participation can help tap into the expertise and creativity
of stakeholders or experts. Involving stakeholders can guide emergent (social) learning processes within public, research, or policy communities (Alcamo and Henrichs
2008). Also, bringing together people working in different domains, many outside
science, within a scenario exercise is being recognized as central to understanding
the dynamics governing complex systems more completely. Scenario exercises with
broad stakeholder involvement can thus function as a way of pulling together often
very different knowledge sources and epistemologies.
A tested working model combines a core scenario team that organizes the exercise (usually composed of experts in the area of analysis and scenario practitioners)
with a broader stakeholder group that contributes to development of the scenarios
(this latter group is sometimes called a scenario panel). The broader stakeholder
group is selected by identifying stakeholders who have an interest in the topic of
the exercise and/or who could be affected by its outcome and then selecting a group
representing the various stakeholders to participate in the exercise. Together, the
core scenario team and the broader stakeholder group generally hold a lot of knowledge about the socioecological system, but individuals will vary in the information
they have about the system and will have different agendas and biases that will
inevitably lead to some conflict within the group. The core scenario team is tasked,
therefore, with planning how they might incorporate the different forms of knowledge from the stakeholder groups in order to achieve a desired outcome from the
The rules of participation are best set up from the beginning. Talking openmindedly about the future is difficult for some groups, as people have different ideas
about how much control they have over the future. Participants should therefore
know exactly how and when they will be contributing and how much influence they
will have over the storylines in comparison with the core scenario team. A talented
facilitator is essential to managing this balance and retaining the interest and trust of
all participants, as described earlier. Ensuring that the scenario process benefits from
the input of a relevant cross-section of society increases the likelihood that the scenarios will have buy in from the appropriate actors, affecting the potential impact
of the scenario exercise.
Also, the role of stakeholders in the scenario exercise should be discussed and
clarified before involving stakeholders to avoid misunderstandings and risking the
legitimacy of the process. Guidelines may be drawn up that should be shared with
the stakeholders in order to promote transparency and trust. Participation in scenario exercises can be organized in different ways and may involve helping to define
the system parameters of interest, participating in the development of scenarios,
using the end products, or offering guidance on the process and contents of the exercise (Volkery et al. 2008). Simply put, stakeholders may either be consulted or can
lead or co-lead a scenario development process (see Table 5.2).
Participation is both time consuming and sometimes expensive, and therefore
a balance needs to be struck between having effective and useful participation and
staying within the budget and time constraints of the exercise. If there is no budget
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 175
Table 5.2. Explaining different degrees of stakeholder involvement and the corresponding roles of
stakeholders and core scenario team
Degree of
Stakeholders team
The core scenario team develops the
content and process with inputs
from stakeholders via consultations
and iterative reviews of the products
Content lead, The core scenario team shares
responsibility for the process with
stakeholders, but leads the shaping
of the scenario content
Co-lead on
Co-lead on
The core scenario team shares the
development of the process and
content with the stakeholders fully,
often with the help of an outside
The core scenario team supports the
process of scenario development,
but the stakeholders are completely
responsible for both the process and
content of the exercise; an outside
facilitator may be used, or in some
cases the core scenario team may
for hosting scenario workshops, less costly means of engaging stakeholders may be
preferable. For example, several reviews of the scenarios by various stakeholders or
consultations with key end users might be planned instead. Boxes 5.6 and 5.7 give
examples of stakeholder participation approaches in two scenario exercises.
Again, the goals of the exercise guide how the different forms of knowledge
might be incorporated and how stakeholders might contribute. In the context of
developing scenarios for scientific exploration and research, the core scenario team
will have a strong say in the storylines and scenario components to ensure credibility. Inviting experts from many disciplines to review the scenarios will help balance
the inputs from participating disciplines and strengthen credibility further. An exercise aimed at scientific exploration might nevertheless try to encourage meaningful
contributions from a range of stakeholders in order to capture new ideas, knowledge, and unknown dynamics within the system. Conversely, if the scenarios are
developed to support decision processes, it is important to leave control of the storylines with the stakeholder panel in order to ensure legitimacy. The decision makers
often have a broader understanding of at least the social components of the system
and can help to design the system more realistically. An outreach exercise to communicate the scenarios to a broader audience will aim to have as open and inclusive
a process as possible in order to achieve buy in from stakeholders and encourage
broad collaborative learning.
176 | A Manual for Assessment Practitioners
Box 5.6. The PRELUDE scenario exercise
Over the course of a year, the European Environment Agency (EEA) organized three workshops
to develop the PRELUDE scenarios, focused on the future of European land use and management (EEA 2007). Each workshop lasted three days. Experienced professional facilitators conducted the sessions with stakeholders to arrive at the final storylines. About 30 stakeholders
were involved in the overall process, and although their travel and accommodation were paid
for, they were not compensated for their time.
In the beginning, it was not clear whether it would be possible to maintain strong engagement with a large group of stakeholders with limited time in such an intensive process. The
relative success of this endeavor can be explained by several factors:
• Full responsibility for drafting the storylines was given to the participants, and thus they
developed an ownership of the process that made them return to subsequent meetings.
• The external facilitation of the process underlined that there was no hidden agenda of the
sponsoring organization or any other involved organization. Observing this, the participants engaged in open and lively debate.
• Different animation exercises such as short movies were used to produce an inspiring
working atmosphere that stimulated idea exchange and creativity among participants.
The scenarios developed by the stakeholders were strengthened and given credibility through
reviews and quantitative modeling conducted by experts in relevant fields. The first workshop
focused on identifying key uncertainties, driving forces, and the underlying scenario logics, as
well as considering potential land use–related environmental impacts. After the workshop the
draft scenarios were analyzed and reviewed by the EEA project team and a scenario analysis
support group composed of land-use experts and modelers, who were also present at the
workshop. The draft scenarios were then quantified using spatially explicit data from land use
simulation models.
The objective of the second workshop was essentially to revise the first round of model results, check for inconsistencies, and refine the scenario storylines in view of the modeling data.
Interaction between modelers and stakeholders resulted in the translation of the qualitative
narratives into numerical trends. These numbers were further calibrated based on modeling
data from existing relevant exercises.
The third and final workshop had three objectives: a final review of the five scenarios, a
review of the environmental impacts of the scenarios, and a process to build consensus among
participants concerning the final PRELUDE results, main products, and future dissemination
activities. The outcome of the exercise is available at and is described in EEA 2007.
Balancing stakeholder involvement and expert knowledge within a time-bounded
exercise is tricky. The more the stakeholders are encouraged to participate, the more
complex and lengthy the process becomes, and time constraints must be managed
strategically. In all cases, careful facilitation is important for managing power imbalances, language differences, and expectations among participants. No participants
should ever feel that their contributions are being ignored.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 177
Box 5.7. The Southern Africa Millennium Ecosystem Assessment (SAfMA) scenario
The scenario exercises conducted within SAfMA took place at multiple scales. SAfMA-regional
was an assessment of the African continent below the equator, SAfMA-Gariep was a water
basin level assessment, and SAfMA-livelihoods was conducted within several villages in South
Africa. At the regional and water basin scales, the scenario exercises were conducted entirely by the core assessment teams. SAfMA-regional synthesized the work of several sets of
detailed scenarios that had been developed previously. The core team of SAfMA-Gariep developed original scenarios for the region. Both SAfMA-regional and SAfMA-Gariep then used
their respective advisory boards to review the scenarios. The advisory boards were composed
of representatives of key stakeholder groups in the region, and consultations with them were
used to develop subsequent iterations on the scenarios.
At the local level, SAfMA-livelihoods invested in more time-intensive participatory approaches to developing scenarios. About 20 local participants were selected by their respective
communities to be intensively involved in the scenario exercise based on the idea of so-called
user groups. Each user group was defined by its main activity (e.g., livestock owners, fuelwood collectors, farmers, etc.), and members were selected by the community based on their
participation in that activity, their willingness to participate, and their level of expertise. Each
of the groups had to have a range of ages, from young people to elderly. Gender balance was
achieved naturally because of the prevailing division of labor (i.e., fuelwood is usually collected
by women; livestock is kept by men, etc.).
These user groups were involved over six months in exploring changes in the activities
they were involved in, the resources related to these, and the drivers or underlying causes
of those changes. These trends and observations were compared with the core assessment
team’s understanding of drivers that were taking place at broader scales during the same period. A group of actors used the whole set of information to make up storylines about the future
based on this understanding of the past. The user groups then viewed these plays, changed
them to better suit the reality in their community, and eventually the scenarios were performed
for the entire community.
5.3 How to develop scenarios
Section’s take-home messages
• Scenario development processes should be tailor made to meet a scenario exercise’s
specific needs. A selection of standard approaches exists, but these should always be
adapted to the respective context and the scenario exercise’s focal issue.
• Usually, scenario development consists of a number of stages that can be carried out
in an iterative manner: identification of main concerns, discussion of key drivers and
uncertainties, selection of underlying scenario logics, description of scenario assumptions,
and analysis of the scenario implications.
• Scenarios can be depicted using either qualitative (such as stories or pictures) or
quantitative (such as numbers and graphs) means—or combinations of these two.
Translating qualitative descriptions into quantitative assumptions is not a trivial task,
178 | A Manual for Assessment Practitioners
5.3.1 Different approaches to developing scenarios
A host of different approaches can be and have been used to develop scenarios, very
much depending on who builds the scenarios and for what purpose. The approaches
to scenario development that might be used as part of an ecosystem assessment do
not differ principally from scenario processes in other contexts—except that the
primary focus here is on exploring how interactions between ecosystems and socioeconomic activities may unfold in the future. Generally speaking, approaches to
scenario development can be roughly grouped into three types: deductive, inductive,
and incremental. In this manual, we focus on introducing the key stages in a deductive type in order to illustrate the basic stages of a scenario exercise. For completeness, however, and to highlight key characteristics, we first introduce the inductive
and incremental approaches.
Inductive approaches to scenario development require similar steps as a deductive approach but apply a different method for identifying key uncertainties and developing scenario logics. Rather than systematically discussing and deducing driving
forces, this variant of scenario development starts off by describing individual events
or plot elements and then spins larger stories around these seeds. By starting from
episodic plot elements, developers can build a scenario that will have future consequences that call for some strategic decisions in the present. The process is somewhat less systematic than the deductive approach and calls for a degree of creativity
and imagination that may be difficult to structure and coordinate, but it may result
in more diverse and unusual scenarios (see Ogilvy and Schwartz 1998).
Incremental approaches build on expanding and questioning a “reference scenario” (or “business-as-usual scenario” or “official future”). A reference scenario is
the future that scenario developers or decision makers really believe, either explicitly
or implicitly, will occur. This is usually a plausible and relatively nonthreatening
scenario, featuring no surprising changes to the current environment and continued
stable growth. Then, to contrast this picture, alternatives are explored by first identifying key threats to this pathway and then varying the driving forces that appear
most influential. Based on these, different interactions between key driving forces
might produce unexpected outcomes and allow new scenarios to be built to contrast
the original future outlook (see Ogilvy and Schwartz 1998).
Deductive approaches have four main stages: a first stage geared toward identifying main concerns about future developments; a second stage focusing on discussing the main uncertainties, driving forces, factors, and actors that might be expected
to shape future trends and their interactions—and thus identifying the underlying
scenario logics; a third stage during which the actual scenarios are developed; and
a fourth stage in which the scenarios and their implications are analyzed. Each of
these stages involves a number of steps and generates different types of information that can be relevant at different points in decision-making processes. Similar
approaches to scenario development are also described by Jaeger et al. (2007) and
Schwartz (1991).
As noted above, this manual focuses on introducing the key stages in a deductive
scenario development. Figure 5.3 and Table 5.3 provide an overview of the basic
stages and the information generated in each step:
• How to identify a focal issue (Stage 1)
• How to discuss drivers and scenario logics (Stage 2)
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 179
• How to describe scenario assumptions and scenario storylines (Stage 3)
• How to analyze scenarios—that is, the implications of scenario assumptions for
the focal issue (Stage 4).
This approach allows a focus on exploring the most uncertain and most important
driving forces in a systematic manner and is overall quite straightforward and easy
to follow. However, note that while the approach is described here in a more or less
linear way, there are likely to be iterations between the various steps. Furthermore,
it is important to keep in mind that this approach to scenario development has the
drawback of being rather rigorous and may not be the most appropriate in all circumstances—it very much depends on the purpose of the scenario exercise. Indeed,
when embarking on a full-scale scenario exercise, the process should be tailored to
the specific situation and context.
The remainder of this section and the next one focus on the outcomes that each
scenario development stage needs to achieve in order to carry out a complete scenario exercise. However, they do not attempt to provide a universal step-by-step
approach nor do they attempt to provide a set of authoritative guidelines to scenario
development. For guidance on this, see Wollenberg et al. (2000), Evans et al. (2006),
Jaeger et al. (2007), or Alcamo and Henrichs (2008). Also, this chapter does not
aim to address in any detail the process and related (facilitation) techniques that
might be used to arrive at the respective outcomes described in each stage—instead
it stresses that involving trained facilitators to help guide the scenario development
process is invaluable (see also section 5.2).
Figure 5.3. Key stages in a deductive-type scenario development process
180 | A Manual for Assessment Practitioners
Table 5.3. Steps in a deductive-type scenario development process and relevance to decision
Steps *
Stage 1: How to
identify focal
1) discuss historical areas
and developments that
led to present situation
2) identify main concerns
for the future
3) identify focal questions
(main problems) to
be addressed by the
Stage 2: How
to discuss
and main
4) list main drivers and
uncertainties that will
change the future
5) identify possible driver
trajectories, thresholds,
and uncertainty about
6) identify main interactions
between drivers
Stage 3: How
to describe
and scenario
7) develop first drafts of
scenario storylines
8) translate storylines
into model inputs and
execute a modeling
exercise (optional)
9) finalize scenarios based
on critical assessment
of storylines (qualitative)
and modeling
(quantitative) results, as
well as on stakeholder
Type of information
1) analysis of current
problems and their
roots, based on
stakeholder analysis
2) analysis of key
questions for the
3) clear understanding
of main assumptions
for the future of the
investigated system
4) analysis of main
drivers shaping
the future and their
5) voicing of different
view points on
drivers’ trajectories
and their importance
6) understanding of
system’s interactions,
development of a
system perspective
7) creative ideas about
the future and
emerging changes
8) challenges for
assumptions on
drivers’ interactions,
consistency checks
9) grounding of
qualitative knowledge
through modeling
Relevance to
– identifying issues
– framing issues
– identifying
stakeholders to be
engaged in decision
– framing issues
– prioritizing
– informing policy
– identifying decision
– evaluating policy
– selecting policy
– designing
monitoring systems
5.3.2 How to identify a focal issue (Stage 1)
When setting up a scenario exercise, the general scope will have been defined, often
implicitly or explicitly determining the principal issues that should be addressed in
a scenario exercise (see section 5.2.2). Within an ecosystem assessment the main
issue would typically be exploring the future of ecosystems and human well-being
in a particular region. However, the issue could also be analyzing the prospects for
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 181
Table 5.3. continued
Stage 4: How
to analyze
Steps *
10) conduct analysis across
the scenarios set
11) discuss scenarios
analysis’s results for
various stakeholder
12) write up and
disseminate scenario
Type of information
10) assessment of
trade-offs and
synergies of various
management options
11) information
to different
stakeholders on
differing view points
12) awareness of
emerging issues for
the future
Relevance to
– identifying policy
– evaluating policy
– developing
strategies for policy
implementation and
* Although the steps are described in a linear way, in practice there is much iteration among them.
** For more details on the policy cycle, see section 5.2 and Chapter 2.
Source: adapted from WRI 2008.
biodiversity and habitat loss, elaborating the trade-offs between different types of
ecosystem services depending on different development paths, or similar issues.
It has proved to be helpful to clarify the questions that a set of scenarios should
attempt to address further and to establish a so-called focal issue or focal question.
For this, the main concerns and questions stakeholders might have concerning the
future need to be identified. This can be done based on literature review, but it should
ideally also involve a more iterative process, including interviews with stakeholders
and decision makers. Based on the resulting understanding of the perceptions of the
nature of prospective challenges, the impacts and dynamics of past developments
(see below), and the prevalent expectations about future trends, a focal question for
the scenario exercise can be constructed. It is desirable to make this focal question as
objective and unambiguous as possible (Scearce et al. 2004) and preferably to link it
to concrete choices, decisions, or strategic considerations at stake. Thus, rather than
asking in general terms “What does the future hold for our ecosystem?” the focal
question might be “Does the way we use ecosystem services in our region put future
food security at risk?” or “Can the Millennium Development Goals be achieved,
short term and long term?”
Typically, the answer to a focal question constructed in this manner will be “It
depends!” One of the key aims of any scenario exercise is to distill what the answer
depends on and what the related implications are. One of the main advantages of
centering a scenario exercise on a concrete and unambiguous focal question is that it
creates a common and transparent platform for the process. Particularly in the context of complex systems, it is easy for scenario developers to get lost in the myriad
of interconnections and possible angles to an issue—a concrete focal question here
helps to manage the degree of complexity somewhat (without neglecting the need
to place the discussion around a specific focal question within its broader ecosystem
182 | A Manual for Assessment Practitioners
and socioeconomic context). Consequently, focal questions guide the further scenario development and help identify which specific topics need to be included in the
actual scenarios and which might be omitted—thus grounding the scenarios. Based
on the focal question, developers will also have to discuss the time frame for the
scenarios, balancing the short-term time horizon of many planning processes with
the slow, long-term nature of ecosystem changes.
As noted, an early step in the process can include a look into the past to familiarize participants with changing development patterns and explore the wide range
of factors that contributed to the current situation and problems. Identifying the
main actors responsible for certain developments can be part of this analysis, which
will later help to discuss the possible role of individual actors in the future and how
the relationship among actors could change in different scenarios. Talking about
past developments from different perspectives can also help create a common understanding of the present shared by all those who set out to construct the scenarios.
Especially if the group developing the scenarios is composed of experts and stakeholders with varying expertise, this type of retrospective discussion is useful in order
to establish a common starting point for the look into the future. Often it is illuminating to look as far back into the past as the group sets out to look into the future:
This can help people realize how volatile systems can be and help scenario developers to get out of their current mind-set and take a more long-term perspective.
5.3.3 How to discuss drivers and select scenario logics (Stage 2)
The second stage extends the discussion to include an analysis of drivers of change
(and if the scenario exercise is part of a broader ecosystem assessment, this discussion can build on the drivers already identified—see Chapter 4). Here the focus
should be on the main drivers of change expected to play a major role in the future.
For this, developers can distinguish between the drivers that directly affect the concerns identified in Stage 1 along with the underlying ones that seem more removed
and indirect. Often it is the interaction of these indirect drivers that shape direct
drivers’ trajectories in the future, but they can be difficult to assess over shorter time
frames. Exploring the web of drivers and how they are linked and influence each
other is not an easy task. This discussion is likely to be shaped by different perspectives of scenario builders and differences in knowledge on each topic. Discussing the
interactions between drivers can also help identify possible thresholds that will not
just change the course of one driver but affect the functioning of the whole system.
The possible trends and trajectories for each driver should be discussed, such
as the expected brackets of population numbers over the scenarios’ time horizon or
possible economic growth rates in the future. It is also important to identify how
different participants perceive these trends and how certain they are they will play
out in the future and why. This will help uncover participants’ main assumptions
about the future. The discussion of the certainty of how these trends will play out in
the future is crucial, as one possibility is to develop the scenarios around the most
uncertain drivers, exploring their possible different pathways in the future.
The discussion of direct and indirect driving forces can help identify the main
uncertainties for the future, as any threshold points or bifurcation points identified
can later help develop the actual scenario storylines. One important decision has to
be made when working with this approach to identifying uncertainties: There can be
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 183
uncertainty about a driver itself, its trend over the course of the future, or its interaction with other drivers that is the result of a lack of information (i.e., ignorance).
In addition, there can also be uncertainty surrounding the impact that a driver and
changes in its trend can have on the future. Clarifying which kind of uncertainty is
more important for a particular exercise is helpful to avoid confusion among the
scenario developers.
To prepare for a discussion of the underlying scenario logics and then build the
actual scenario storylines, participants may find it helpful to rank all the drivers
with respect to their importance for the identified problems and the degree of uncertainty (see Figure 5.4), document the reasons for the ranking, and discuss why and
how the drivers interact. Reflecting on both types of uncertainty—what is known
about a driver and about its impact—can be a useful tool for deciding on the critical
uncertainties (see Jaeger et al. 2007 for additional information on how to identify
critical uncertainties).
Based on this, the underlying scenario logics can be developed. One approach
used in the past is to frame the scenarios based on two critical uncertain drivers that
seem to influence most or all of the others (this has been used in a number of scenario exercises such as IPCC 2000 and MA 2005a; see Figure 5.5). Reaching agreement on these key uncertainties—usually the most important and most uncertain
ones—is not an easy exercise, but once agreement is reached a matrix depicting the
two opposing extremes of each uncertainty can be constructed. By combining the
extreme assumptions of each of the two axes, four scenario logics emerge that can
be used to develop the main stories of the scenarios.
It should be noted again that other approaches can be applied to establish scenario logics and that the “axes approach” is particular to the deductive approach
Figure 5.4. How to identify critical uncertainties.
Note: CU 1 and CU 2 denote critical uncertainties.
Source: Jaeger et al. 2007
184 | A Manual for Assessment Practitioners
Figure 5.5. Schematic illustration of SRES scenarios. The four scenario “families” are
illustrated, very simplistically, as branches of a two-dimensional tree. In reality, the four
scenario families share a space of a much higher dimensionality, given the numerous
assumptions needed to define any given scenario.
Source: IPCC 2000
described here. Other approaches may build on exploring “emblematic events” (see,
for example, Ogilvy and Schwartz 1998), using decision tree type approaches (see,
for example, Kahane 1992), or simply expanding on existing scenarios (see Box
5.8).There is no single best practice to arrive at scenario logics, and which to use
depends considerably on the context of the scenario exercise and the dynamics and
schools of thought prevalent in the group that develops the scenarios. Here, in particular, facilitators experienced in guiding scenario development process may prove
to be invaluable (see section 5.2.4).
With respect to the number of scenarios to develop, there should not be too
many scenarios, as it will be difficult for the audience to keep track of each storyline
and see the differences between them (generally more than five scenarios are difficult
to handle and comprehend). Developing an even number of scenarios avoids problems with the audience focusing on one scenario that is seen as covering a middle
position and thus seemingly being the most likely future. The approach described
here leads to four scenarios, which allows some variety while following a rigorous
structure. Other exercises may use two scenarios (for example, a best-case and a
worst-case scenario) or three scenarios (for example, a reference scenario framed by
two alternatives). Still others may aim to highlight the many uncertainties around an
issue and use five or more scenarios.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 185
Box 5.8. Building on existing scenarios
Over the last decade a number of scenario exercises were carried out with an environmental
focus, such as IPCC (2000), MA (2005a), and UNEP (2002, 2007). When carrying out a new
scenario exercise, a first question is whether the new scenarios should in one way or another
build on previous ones. And if yes, what should be taken from previous exercises that can either have the same or a very different focus as the new scenarios to be developed.
The answer to such questions depends on a number of factors, such as what the previous
scenarios were intended for, what kind of information they produced, and how well they were
documented. As a general rule, the more buy in the scenario exercise aims to achieve and the
more the exercise is done to help with concrete decisions, the better it is for a scenario exercise to devise its own set of logics and storylines fit to the focal issue at hand.
Nevertheless, existing scenarios can provide a useful source of information and inspiration. The information of previous exercises about specific drivers, for example, may help outline future trajectories in new scenarios. More caution is necessary when simply “borrowing”
full scenario logics, however, as a key value in any exercise is to discuss the specific uncertainties to its focal issue and to construct scenarios around these.
In some studies, existing sets of scenarios have been used to frame more detailed thematic
analyses. The ATEAM project (Rounsevell et al. 2005) offers an example: The IPCC SRES scenario logics (IPCC 2000) provided a backdrop for regional quantification of land use changes,
using appropriate models to further “enrich” these global scenarios at the European scale.
5.3.4 How to describe scenario assumptions and scenario storylines
(Stage 3)
The third stage involves developing a set of stories about the future about how various drivers might interact and unfold in different ways based on the scenario logics.
Each scenario’s logic inherently entails a specific set of assumptions about the drivers and how they develop over the chosen time horizon. In order words, the scenario
logic determines the way the events in each scenario play out.
In order to develop illustrative scenario storylines, developers should draw on
whatever tools are available to stimulate creative thinking that generates interesting
and even provocative descriptions of the future while ensuring that these remain
plausible and consistent with the understanding of socioeconomic and environmental processes. Storylines are the basic qualitative descriptions of each scenario. They
may, for example, be developed as a series of fictitious events that illustrate how
the world might develop over the time horizon and how these events influence the
decisions of different stakeholders. This can be told in the form of stylized facts
about the future, “letters from the future,” short stories that play out in the assumed
future world, fictitious future newspaper articles, essays about future prospects, or
illustrations of future developments—there is no limit to participants’ creativity.
However, it is important to ensure that the scenario storylines relate back to the
original focal issue identified in Stage 1. Also, the value of each storyline increases
if it remains somewhat comparable to the other storylines developed, and features
at least some similar elements. This allows a better understanding of the contrasting
pathways the scenarios depict and how individual assumptions shape future trends.
186 | A Manual for Assessment Practitioners
Furthermore, it may be useful to relate the storylines back to how main actors may
act under the respective scenarios. If, for example, an analysis of the key factors and
actors has been carried out earlier in the assessment, scenario developers can use this
information to add additional elements to the storylines and explore different ways
for these actors to behave.
Quantifying the main trends in the scenario storylines can also enhance the
descriptions of future pathways and may help ensure internal consistency of the
respective scenarios. A useful tool to enrich scenario storylines with quantitative
information is to run simulation models to quantify future trends of drivers—if
time and resources permit and if the issues at hand lend themselves to modeling
approaches. Thus assumptions on key driving forces (such as population, economic
growth, consumption patterns, lifestyle choices) may be quantified, and possible outcomes for ecosystems and their services (such as food production, climate change,
or water availability) can be calculated. In such cases, each model run needs to be
based on specific assumptions about drivers’ trajectories and their interactions that
correspond to the respective scenario storylines. In complex models, however, simulations can take a few months to complete (see also section 5.4).
The translation of qualitative descriptions into quantitative assumptions, however, is not trivial. It is a particular challenge to ensure that this happens in a transparent manner—and all too often the qualitative description and the quantitative
analysis are separate processes. An approach used in several previous assessments
was to develop tables containing the main characteristics of the storylines and
graphic storyline summaries (for example, with upward or downward arrows to
indicate important trends; see MA 2005a). This formed a basis for arriving at model
input assumptions for the quantitative part of the scenario exercise. Elsewhere, key
input parameters were “semi-quantified” using simple ranking or scaling methods
(see Box 5.9). And in some cases it may be useful to build on existing scenarios and
their quantifications, as mentioned earlier. As noted, this translation of the qualitative descriptions into quantitative information (that can then provide input to
further quantitative analysis, see section 5.4) can be done by individual modeling
groups, but it is probably best done in close interaction between modeling groups
and storyline developers.
Several iterations between qualitative and quantitative techniques help achieve
consistent scenarios. But just using either only qualitative methods or only modeling
approaches can also generate interesting plausible futures that can stir up discussion.
The approach best chosen depends very much on the purpose of the scenario’s exercise and the available resources (see also reflections on the balance between qualitative and quantitative information in sections 5.2.2 and 5.4.1). Developers also need
to consider the multiple geographical scales that the scenarios can or should address
(see Box 5.10).
The final stage in a scenario exercise involves analyzing the implications of the
scenarios that have been developed for informing decisions taken today or in the
near future (i.e., Stage 4). For this the scenarios have to be “locked in” at one point
in time—that is, the stories will not change any more after the iterative process. This
is needed to ensure some degree of consistency and comparability in the analysis.
Indeed, much of the value of the scenario exercise lies in being able to compare different outcomes, as described in more detail in section 5.4.
Box 5.9. Translating storylines into model input (some examples)
The Millennium Ecosystem Assessment (MA 2005a) provides an example of how scenario storylines can be translated into model input. The four global-level scenarios developed in the
MA differ in economic growth and international collaboration (see table). One stylized fact, for
example, postulated that high income levels are associated with low fertility levels and low
mortality levels. Based on the description of income levels and international distribution, estimates of fertility and mortality in different parts of the world were made. These then were used
as model input, so that population projections could be derived (see figure).
Assumptions underlying MA population trends
Scenario 1:
Scenario 2:
Order From
Scenario 3:
Adapting Mosaic
Scenario 4:
D: low
I: medium
D: high
I: low
D: high
I: low until 2010,
deviate to medium
by 2050
D: medium
I: medium
D: low
I: low
D: high
I: high
D: high
I: high until 2010,
deviate to medium
by 2050
D: medium
I: medium
D = developing countries; I = industrial countries
Box 5.9 Figure. Population trends in the MA.
Source: MA 2005a
Box 5.9. (continued)
In the PRELUDE project (EEA 2007), a scenario exercise was carried out to stimulate
strategic discussions of Europe’s land use policies. Considerable attention was paid to transparency in the development of five scenarios that represent different perspectives of Europe’s
future. Scenarios were constructed around both qualitative and quantitative information and
were refined and revised several times. As an important step, key driving forces were quantified roughly using a simple scale from 1 to 10 during the scenario development process, in
order to give modeling teams an indication of the relative trends in each scenario.
It is important to note that the opinion of a limited group of scenario developers may play
a key role in this interpretation process. One way to reduce the risk of an unwanted determination of outcomes by scenario developers has been the explicit use of worldviews. This approach has, for instance, been successfully applied in the TARGETS project to make scenario
developers aware of consistent interpretations of facts different from their own worldview. As a
result, rather broad assumptions were made for scenario inputs—and in some cases, models
were even adjusted to better reflect the worldviews (Rotmans and de Vries 1997).
Box 5.10. Linking scenarios across multiple geographical scales
Scenario exercises need to consider how to treat the geographical scale(s) of the issues that
are investigated as part of the assessment. Increasingly, these exercises address the question
of how to build a multiscale scenario (i.e., various geographical scales and cross-scale interactions are explicitly addressed in a single scenario exercise) or multiple scale scenarios (i.e.,
scenarios are built at different scales independently and then linked with each other).
Linking scenarios across scales has a number of advantages for an ecosystem assessment, such as ensuring consistency of methods and results across scales, enabling the analysis of feedbacks across scales, and enhancing the potential of communication between key
stakeholders at different scales. But it should also be noted that it may entail some risks, such
as achieving arbitrary consistency (or “over-consistency”) of results and thus losing relevance
at specific scales. Also, multiscale assessments usually require large investments of time and
Generally speaking, scenario exercises can be linked across geographical scales in two
principal ways: through the scenario development process or the scenario elements (Zurek and
Henrichs 2007). During an assessment exercise that spans several geographical scales, separate scenario exercises can be carried out at different scales—and these may or may not be
coupled across different regions and geographical scales. Five levels can be distinguished:
• Joint scenario development processes (i.e., scenarios developed at different geographical
scales in a joint scenario exercise, with the same group of scenario developers).
• Parallel scenario development processes (i.e., different groups of scenario developers
building scenarios at different scales but in more or less parallel processes in terms of
focal question, conceptual frameworks, scenario development approach, or information
Box 5.10. (continued)
• Iterative scenario development processes (i.e., developing draft scenarios at one scale that
provide a starting point for scenario development at another scale, which then provides
input and feedback for revision of scenarios at the original scale).
• Consecutive scenario development processes (i.e., a set of scenarios first developed and
finalized at one geographical level and then scaled to another geographical level).
• Independent scenario development processes (i.e., separate exercise carried out at two
or more geographical scales—which may or may not inform each other in an informal
Alternatively, scenarios developed at different geographical scales can be linked by using the
same scenario elements. The degree of similarity of scenarios across different scales can vary
and largely depends on how they are developed. Zurek and Henrichs (2007) describe five different ways of linking scenario elements across scales:
• The closest link between scenarios across scales is achieved when scenarios are equivalent or congruent across scales and fully share their scenario logics, key assumptions, and
• If scenarios are consistent across scale, they share main scenario assumptions, driving
forces, and trends, but these may play out differently with regard to the scenario implications and outcomes.
• Coherent scenarios follow the same scenario logics across scales—in other words, the
scenarios “match.” This does not preclude substantial differences with regard to how the
scenarios play out in the selection of important driving forces, their major trends, and/or
scenario outcomes.
• Comparable scenarios may be constructed to be largely independent at different scales,
connected mainly by the issue they address and possibly addressing the same focal issue.
• Scenarios may be independent and thus complementary across scales—yet this does
not preclude selected information from scenarios at one scale feeding into scenarios at
The degree to which scenario elements will be linked in different scenarios will also depend
to a large extent on the type of scenario process applied (see Table). Thus, at the outset of a
multiscale exercise the developers should decide what degree of linkage between scenarios
is desirable (i.e., how closely the scenario elements should be linked) and design the process
accordingly. An example of linking scenarios across scale is the work done by the subglobal
groups in the MA. Details on how the process worked for each group can be found in the MA
subglobal volume (MA 2005b).
190 | A Manual for Assessment Practitioners
Box 5.10. (continued)
Relationship between scenario development processes and scenario outcomes across
geographical scales. While any process can lead to any degree of linkage, some are more
likely than others.
across scales
likely, if a
for S1 & S2
S1 & S2
S1 & S2
sessions or
likely, if
S1 defines
input / data
for S2
unlikely, no
between S1
& S2
very likely,
very likely,
very likely,
unlikely, only
across scales
as S1 & S2
developed by
same group
but only if
both S1 & S2
explicit aim
for it
if S1 or S2
if S1 sets
for S2
if S1 & S2
are de facto
across scales
likely, if S1
likely, if
very likely, if
unlikely, only
if S1 & S2
& S2 share
S1 & S2
follow same
S1 provides
starting point
if S1 & S2
are de facto
point, and
for S2
possible, if
S1 & S2 are
very likely,
S1 & S2
if S1 &
possible, if
S1 and S2
in parallel de
the same
S2 aim to
if different
focus S1 &
S2, same
use similar
as S1 & S2
likely, if S1
& S2 are
this implies
possible, if
S1 provides
very likely,
S1 & S2
parallel yet
info only
for S2
similar issues
across scales
across scales
Note: S1 and S2 denote two separate scenario exercises at two different geographical scales.
Source: Zurek and Henrichs, 2007.
5.4 How to analyze scenarios
Section’s take-home messages
• To analyze scenarios, either qualitative or quantitative approaches or a combination of
the two can be used. Each has strengths and weaknesses, and using a combination can
harvest the advantages of both, but it requires substantial amounts of time and resources.
• In the context of ecosystem assessments, a broad range of modeling tools exists at
various scales that can support quantitative analyses—but it is difficult to find a single
model that can adequately capture all the dynamics depicted in a scenario.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 191
• An analysis across a set of scenarios—for example, with respect to similar or differing
trends—can facilitate discussions about possible future trends and their plausibility, the
impact of different response options, or the robustness of different response options
under different assumptions.
5.4.1 Analyzing scenarios—combining qualitative and quantitative
The analysis of the potential implications of scenario assumptions is often core to
environmental scenario exercises: much of the value of the exercise lies in being able
to compare different outcomes. Such comparisons can reveal unanticipated results
and provide different stakeholder groups with insights about the outcomes of the
future pathways they may have advocated. Scenarios can be analyzed using a range
of different tools—both qualitative and quantitative. Section 5.3 highlighted the role
of and interplay between qualitative and quantitative information when developing
scenarios. This section explores this notion in further detail, shows how combining
these different types of information can be used in the analysis of scenario implications, and discusses the role of analytical tools in scenario analyses in more detail.
As noted, the differentiation between qualitative and quantitative scenarios has
proved to be useful. But this distinction is not always clear-cut. Many reference
(or business-as-usual) scenarios, for example, focus on projections of expected future changes without emphasizing explicitly the underlying assumptions (see, for
example, IEA 2007, FAO 2002, EEA 2005). Conversely, qualitative scenarios about
possible future developments are seldom void of all quantification. Thus, as most
scenarios implicitly combine qualitative and quantitative information to some extent, this differentiation is mainly useful for highlighting the principal focus of a
given set of scenarios.
The specific roles of qualitative information within scenarios include the
• To capture the multiplicity of perspectives of actors, and the significance that
those actors assign to events. This allows subjective insights from stakeholders
to be taken into consideration. The necessity of subjectivity (in understanding
social issues) is due to the recognition that there might be several different
alternative perspectives of reality, all of which may be “valid” and should be
explored. Models are not capable of doing this.
• To enlarge the range of outcomes of a scenario analysis. Models, for example,
cannot address all issues at all geographical scales at the same time. In particular, local developments are usually very dependent on close relations between
actors that cannot be quantified by models. Qualitative approaches can provide
insights into these aspects and therefore can broaden the scope of the scenario
analysis by picking up issues that models cannot capture.
• To deal with shocks and disruptive scenarios. Usually simulation models
are constructed to provide insight into gradual changes in systems, since the
numerical approach can play out its strengths in capturing this type of change.
In reality, however, societal changes are usually most dominant when extreme
events occur. Where models lack numerical insights in capturing these extreme
events, qualitative approaches can dwell upon their potential impact.
192 | A Manual for Assessment Practitioners
The specific roles of quantitative information within scenarios—for example, based
on modeling output—include the following:
• To “enrich” qualitative scenarios by showing trends and dynamics not
anticipated in the storylines. Models may provide another way to explore the
consequences of changing one set of parameters for other parts of the system. In
this way, models can provide insight into complex relations of a system that are
difficult to grasp. Clearly, models are restricted to aspects that can adequately be
translated into numerical concepts.
• To check the consistency and plausibility within scenarios. Models include various constraints on how model parameters may develop. In a climate model, for
instance, the rate at which greenhouse gas concentrations can be reduced is constrained by knowledge on the removal rates of these gases from the atmosphere.
In that context, models can be used to test the consistency of initial storyline
descriptions. It should be noted that if models provide different insights than the
original storyline, the model is not necessarily right (as models are themselves
only one possible simplification of reality). Confronting “qualitative expectations” and model outcomes might lead to a very useful mutual learning exercise.
• To provide relevant numerical information. If uncertainty about the future
allows quantitative assessments, in many cases numerical information can be
more convincing than purely qualitative descriptions. For example, for decision
makers in the field of climate policy it is useful to get some idea of the costs
involved in various climate policy proposals; for decision makers in the area
of bioenergy it is important to know how much land is likely to be used in the
context of various proposals.
Recently, scenarios that explicitly aim to combine qualitative and quantification
information—for example, the “story-and-simulation” approach (see also EEA
2001)—have gained prominence and have been seen as particularly useful in international environmental assessments (see Box 5.11; see also IPCC 2000, Cosgrove
and Rijsberman 2000, Alcamo et al. 2000, UNEP 2002, MA 2005a, UNEP 2007).
Such approaches make best use of the respective strengths of qualitative and quantitative information and should allow for interaction. For example, the assumptions
captured in qualitative storylines may be adjusted on the basis of model outcomes.
At the same time, model inputs (or even the models themselves) can be adjusted
to the respective scenario storylines. Preferably several interactions are organized
between the storyline development and the model-based analysis during a scenario
exercise in order to increase the consistency of the two types of information and
ensure cross-fertilization of both processes (see section 5.3.4).
The degree of quantification is an important choice (see section 5.2.2). While
using models provides some clear advantages, it comes at a cost (in terms of time
and money, but also in terms of reduced flexibility in which issues can be explored
and how). And for certain issues, models might not be available or appropriate. The
degree of quantification also depends on the characteristics of the scenarios that are
developed. In scenarios that deliberately explore the implications of singular events
(such as shocks or surprises) rather than gradual trends, modeling approaches are
harder to implement, as most models have been calibrated against overall long-term
trends and struggle to deal with discontinuities.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 193
Box 5.11. Some examples of the role of modeling in developing scenarios
In the Millennium Ecosystem Assessment, qualitative and quantitative scenario elements
played an equal role (MA 2005a), providing some examples of the specific role of modeling.
• Some scenarios were designed to include climate policies while others did not. Climate
policies will lead to significant changes in the energy system (e.g., less use of fossil fuels
with carbon capture and storage and more use of renewable energy). These changes will
have numerous indirect impacts. In the MA, energy models were used to illustrate the
trends in air pollutant emissions in relation to climate policy—something that cannot be
easily done without model analysis.
• Land use trends are determined by factors such as population change, dietary changes,
and crop yields. These trends may partly occur independently, but they also interact:
increasing land prices as a result of scarcity is major driver behind yield increase. In the
MA, it was found by quantitative modeling that offsetting trends in driving forces led to a
much smaller range in land use outcomes that originally expected. These offsetting trends
include the facts that high economic growth scenarios (with a related strong trend toward
meat-intensive diet patterns) coincided with low population trend, and endogenous drivers
for yield (land prices) acted as a mitigating factor. Subsequent discussions led to revisions
in both the original storylines (weaker differences in land use) and modeling.
• For the MA scenarios, hypotheses were formulated on the consequences of the different
scenarios for biodiversity. For instance, it was expected that biodiversity loss in a Global
Orchestration world may be larger than in a Technogarden world. Still, without further
quantification these statements are not more than hypotheses—which may be ignored by
certain target audiences. Therefore a simple biodiversity model was developed to integrate
trends in factors such as land use, climate, and nitrogen deposition and to illustrate the
expected consequences for biodiversity of the different scenarios.
There are a few substantial risks, however, in using a combined “story-andsimulation” approach:
• Scenario development with models requires substantial resources and time; as
a result, modeling results tend to come in late—reducing the opportunity of
• One approach may dominate over the other; this obviously reduces the purpose
of doing both activities.
• The two outcomes may be inconsistent; Parson et al. (2007), for example,
criticizes the Millennium Ecosystem Assessment (MA 2005a) for inconsistencies
between storyline and quantification. (It should be noted, however, that it was
a deliberate choice to present these inconsistencies in the MA: inconsistencies
were first confronted and discussed and those that could not be resolved
were presented equally, so as not to introduce bias between qualitative and
quantitative analysis.)
Thus, while there are great benefits in combining qualitative and quantitative information within a scenario exercise, this might not always be feasible or desirable
—and it depends on the scenario exercise’s context and goals. In such cases the
194 | A Manual for Assessment Practitioners
strengths of the different types of information should be weighed carefully against
each other.
5.4.2 How to analyze the implications of scenarios regarding ecosystems and
human well-being
There is no single recipe for analyzing the implications of scenarios. In order to illustrate the principal steps of such an analysis, they are elaborated here as a continuation of the deductive scenario development introduced above—assuming that the
main assumptions underpinning the scenarios have been quantified (as indicated in
section 5.3.4) and that a “story-and-simulation” type process to scenario analysis is
The implications of assumptions made about future developments under each
scenario can be analyzed using both qualitative approaches and simulation models.
The term model or modeling tool is used here to refer to tools that describe in a
formal manner the relationship between different elements that together determine
the behavior of a larger system. In other words, models are geared toward linking
assumptions on driving forces to the possible implications these may have in a quantitative manner. Note that this section focuses on the use of such modeling tools to
support scenario analyses—while recognizing that a host of nonquantitative tools
also exist to analyze scenario implications in a more qualitative manner.
At the core of any ecosystem-related scenario analysis is the exploration of the
future trends of key driving forces, their interactions, and their implications for ecosystems and human well-being. In this context, it is useful to distinguish between
different types of driving forces—direct drivers (such as land use change, climate
change, or air pollution) and indirect drivers (such as demographic, economic, or
technology development, as well as human behavior or institutional factors) (see
Chapter 4). For many of these drivers of ecosystem change a variety of detailed
models exist at most geographical scales. Such models can be incorporated into an
assessment. Examples of issues for which simulation tools that have been used in
past assessments at a global scale presented by the Millennium Ecosystem Assessment (MA 2005a) or the Global Environment Outlook (UNEP 2002, UNEP 2007)
include demographic trends, economic development and associated changes in demand for physical goods, emissions, food demand and production, climate change,
land use change and land cover, and water resources. For many other issues, however, modeling tools seem to be scarcer (and arguably cannot exist, due to the high
degree of indeterminism in socioeconomic developments). This, for example, refers
to sociopolitical drivers, educational drivers, and cultural drivers as well as trends in
science and technology.
With regard to the interaction between drivers, currently available quantitative
approaches can be somewhat limited—and qualitative methods are often needed
to complete a scenario analysis. For many of these linkages, “stylized facts” can
be drafted that might be commonly accepted but are not part of formal models.
In some cases, this is because models simply do not capture these links, as most
models only capture one specific driver. In other cases, the stylized facts are still too
weak to use in a quantitative way in models. Examples of such stylized facts include assumptions used in the MA scenarios that “fertility rates are low under high
economic growth scenarios” (based on the fact that current fertility levels are low
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 195
in high-income countries) and that “technology development is high under globalization scenarios.” Both these stylized facts have been included in the scenarios by
introducing them into the scenario storylines. For the models, they simply constitute
exogenous assumptions.
In the case of assessment of complex systems such as those related to ecosystem
services, it is unlikely that one single model can be used to describe the full system.
Although some highly aggregated models have been developed that specifically try
to focus on the overall picture (for example, GUMBO [Bouwmans et al. 2002] or
IFs [Hughes 1999]), such models often lack the detail that is needed for a policyrelevant assessment. Therefore, in recent ecosystem assessments a suite of models
has been used. Usually, the different models are harmonized by using the same driving forces and by linking the output of one model to the input of another model. In
some cases the modeling linkage is even more sophisticated, by allowing different
models to iterate the model outcomes. In this way, the benefits from the specific (expert) knowledge included in each model can be harvested.
At the same time, however, the potential for inconsistencies grows when scenario
exercises use multiple models and attempt to harmonize them, particularly when
some key quantities are externally specified for some models and calculated within
others. These inconsistencies can be minimized, however, for example by iteration
across models (see, for example, Box 5.12) or by choosing model links at the point
of limited interaction (energy models and emissions, for example, are relatively independent of changes in climate). Nevertheless, it should be noted that attempts to
avoid such inconsistencies by standardizing model outputs can entail more serious
risks by obscuring interpretation of results and precluding use of model variation to
illuminate uncertainty. Also, attempts to connect qualitative and quantitative aspects
of scenarios have been particularly challenging for pursuit of consistency. Different
narrative scenarios often reflect different assumptions about how the world works,
which correspond more closely to different model structures than does parameter
variation. Better integrating the two approaches will require developing ways to
connect narrative scenarios to model structures rather than merely targeting values
for a few variables that models are then asked to reproduce.
Analyzing response options also plays a key role in most scenario assessments.
Sometimes these options are to some degree incorporated into the scenario assumptions—but in most cases it is far more useful to analyze possible response options in
a separate step against the backdrop of different scenarios (and thus test the robustness of various options). Response options can be analyzed in a range of different
manners (see also Chapter 6).
One option to analyze the effectiveness of different response options is to use a
model to analyze various options against a single criterion (such as cost minimization or meeting specific targets). An example of this is the standard practice in climate mitigation analysis: by attaching a price to greenhouse gas emissions, measures
that reduce emissions are induced in the model. Analysts report the measures that
are chosen and the associated costs. The analysis can be done starting from different
nonpolicy scenarios in order to map out the consequences of uncertainties in driving
forces (see, for instance, van Vuuren et al. 2007).
An alternative approach is to use scenario storylines to assess whether a certain
measure fits into the set of assumptions that depicts future world settings. For example, in the EU-Ruralis project (Westhoek et al. 2006) different agricultural policies
196 | A Manual for Assessment Practitioners
Box 5.12. Coupling models: experiences from the MA and EU-Ruralis
In the Millennium Ecosystem Assessment (MA 2005a), quite a large set of models were coupled to describe the broad range of changes in drivers of ecosystem services. They were “softlinked” in the sense that output files from one model were used as inputs to others (see figure).
The time interval of data exchanged between models was usually one year. The model linkages included factors such as food production, climate, irrigated areas and water consumption, electricity use and livestock, river discharge, land use change, and nitrogen deposition.
Taking the example of food production, information was forwarded from the agro-economic
model IMPACT to the integrated assessment model IMAGE to compute land use and land
cover changes at grid level that were consistent with the agricultural production computed in
In the EU-Ruralis project (see Westhoek et al. 2006) a modeling framework was constructed to improve the linkage between an agro-economic model (GTAP) and IMAGE. Iterations were done between GTAP and IMAGE to increase the consistency of the information on
agricultural production, availability of land, and climate change. After the iteration, a consistent
set of modeling results was obtained from GTAP and IMAGE, allowing analyses on agricultural
consumption, production and trade (from GTAP), and land use change, greenhouse gas emissions, and climate change (from IMAGE).
Linkages between models are seldom straightforward and usually require the upscaling
(aggregation) or downscaling of various data. The IMPACT model, for instance, uses a more
detailed regional disaggregation level than the IMAGE model. In almost all cases, however,
regional scaling was possible by combining regions (upscaling) or by assuming proportional
changes (downscaling). The models were not recalibrated on the basis of the new input parameters provided by the other models, but in most cases the models had been calibrated
using comparable international databases. The new linkages therefore did not lead to major
inconsistencies in assumptions between the models.
Box 5.12 Figure.
between models
in the MA.
Source: MA
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 197
are added on top of the climate change scenarios laid out by the IPCC (IPCC 2000)
in a manner consistent with the underpinning scenario logics. Agricultural policies
in this example are differentiated between full liberalization of the Common Agricultural Policy of Europe in a globalized and economy-focused world (in the A1
scenario), a complete continuation of current policies under a regionalized and economy-focused world (in the A2 scenario), and even an increase in agri-environmental
payments in a regionalized and environment-focused world (in the B2-scenario).
Obviously, the starting point (the so-called nonintervention scenario) plays a major role in response analysis. For instance, in analyzing the climate response assumption, the technology assumptions in the nonintervention scenario play a major role.
Slow technology development—and thus high emissions—may imply that certain targets are unachievable. As a result, these nonintervention scenario assumptions often
remain an important part of discussion (see, for example, Pielke et al. 2008).
5.4.3 How to treat uncertainty in scenario analysis
While scenario exercises themselves form a tool of uncertainty analysis by exploring
different possible futures, likelihoods and uncertainties can also play a role within
scenarios (addressing, for example, questions such as “how certain is it that a certain assumption made in a scenario will come true?”). A model-supported scenario
analysis can usually only grasp a few key uncertainties by varying among scenarios.
It should be noted that complex narrative scenarios pose special problems in representing and communicating uncertainty, given the large number of posed interactions across parameters.
At the same time, many users of scenarios would like to be presented with some
indication of the possible range of outcomes, or even the likelihood or probabilities
associated with future trends. Most assessments make a strong case trying to prevent
such interpretation. Constructing scenarios offers a way of thinking about the future
in a structured manner, and the aim is to provide a range of outcomes in which the
probability is not the most important aspect. In the MA (MA 2005a) and IPCC
(IPCC 2000) scenarios, for example, it is emphasized that no probabilities could
be attached to any of the constructed scenarios. This has often been interpreted as
equal probability—but in fact this interpretation is incorrect.
However, current scenario practice leaves the question unresolved of whether
and how to best include probabilistic information (Groves and Lempert 2007). It
has been reasoned that assigning probabilities to the likelihood of different scenarios would be misleading and would undermine the credibility—not least due to
the overwhelming influence of societal choice (and associated “deep uncertainty”)
(IPCC 2000). Elsewhere, particularly in the context of climate change impacts, it
has been argued that policy analysts and decision makers need probability estimates
to better assess risks and decide how to respond to them (Schneider 2001, Schneider
2002, Webster et al. 2002). Several studies have applied the contrasting probabilistic
approach to emission scenarios (Webster et al. 2002, Webster et al. 2003, Richels et
al. 2004). An important critique formulated against this approach is that attempts
to assign subjective probabilities in a situation of ignorance forms a dismissal of
uncertainty in favor of spuriously constructed expert opinion (Grübler and Nakicenovic 2001, Grübler et al. 2006).
Instead, scenario analysis often explicitly aims to consider futures that span a
range of possible outcomes—making probabilities less relevant. Still, there seems to
198 | A Manual for Assessment Practitioners
be a natural inclination to interpret the range of scenarios as an indication of possible outcomes. For instance, the Millennium Ecosystem Assessment outcomes (MA
2005a) for loss of natural area range from a 20% loss (under one set of assumptions)
to a 10% gain (under another set). This can easily be interpreted as an indication
that the most likely development is a change in area between –20% and +10%. In
most cases such interpretation is incorrect, as the existence of one low scenario does
not exclude the possibility of even lower ones. However, given the need to explore
the range of possible outcomes, it is worthwhile to consider scenario development
against the backdrop of extreme cases. And most assessments in fact do so.
It may be worthwhile to try to combine the strength of scenario analyses with
probabilistic approaches. In both the IPCC SRES report (IPCC 2000) and the MA
(MA 2005a), for example, the assessment first discusses the full range of possible
outcomes for different scenario drivers—and, in case of the IPPC, also scenario outcomes. Next, the assumptions made for the different drivers are related to the possible range of outcomes. For example, a scenario that emphasizes high population
growth would use an estimate close to the upper range, while a scenario that emphasizes low population growth would use an estimate close to the lower range. As a result, the full scenario range would somehow represent the (uncertainty) range found
in the literature. To some degree, a similar situation exists with respect to scenario
outcomes. If scenarios would all represent a small range of possible outcomes for
some key variables, they would fail to communicate the possible range of outcomes.
For these variables, therefore, a fuller scenario range is most helpful.
5.4.4 How to analyze across a set of scenarios
There are a number of ways to compare outcomes of different scenarios (see Table
5.4). In general, lessons can be drawn from focusing on either the similarities or
the differences in trends across the scenarios. These can be connected to the policy
choices made along the different pathways. Lessons for decision making can also be
drawn by comparing the risks taken and the benefits gained by different groups of
society or by mapping out the trade-offs in each scenario.
Insights emerge from questioning the assumptions made within one story or
pathway and comparing its outcomes with another possible pathway into the future. This analysis can clarify what is known and what is uncertain about the future.
It also sheds light on unexpected results of a particular pathway. In other words, the
scenario analysis can reveal uncertainties and help decision makers avoid the unintended consequences that often plague policy processes.
Most assessments produce a set of scenarios in order to explore possible developments under different sets of assumptions. Here, the set of scenarios can be used
to explore possible future outcomes—and such outcomes may include similar trends
and differing trends, as well as offsetting trends.
In the case of differing trends, scenarios can depict possible different future situations and thus communicate the consequences of different assumptions or trends.
For example, greenhouse gas emissions under the Millennium Ecosystem Assessment’s Global Orchestration scenario are very different from those under the Technogarden scenario. This emphasizes how assumptions about the future dynamics
with regard to economic growth, material consumption, and technology development are of critical importance for future trends in greenhouse gas emissions. In
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 199
Table 5.4. Options for Comparing Scenarios
Options for
comparing across
the scenarios
Look for future
developments that
are the same in all
Example from the Millennium Ecosystem Assessment
Same trend of rising world population up to 2050 in all
scenarios, then stabilization; exact population numbers in 2050
Global forest area declines up to 2050 in all scenarios: velocity
of trends differs.
Look for uncertain
future developments
that differ across
Number of malnourished children in 2050 differs widely among
Identify trade-offs
described in the
Risk of trading off long-term environmental sustainability for
fast improvement in human systems (Global Orchestration).
Quality and quantity of available water resources by 2050 differ
widely among regions and across scenarios.
Risk of trading off solutions to global environmental problems
(requiring global cooperation) for improving local environments
(focusing on local solutions only) (Adapting Mosaic).
Risk of trading off biodiversity conservation for food security
(Global Orchestration).
Identify policy options
that make sense in all
Major investments in public goods and poverty reduction,
together with elimination of harmful trade barriers and
Widespread use of adaptive ecosystem management and
investment in education.
Significant investments in technologies to use ecosystem
services more efficiently, along with widespread inclusion of
ecosystem services in markets.
Source: WRI 2008.
their combination, the two scenarios are indicative of both high emission and low
emission scenarios published elsewhere.
Similar trends for outcome variables—if the outcomes of very different assumptions about underlying drivers are similar or identical—might be just as indicative.
Such similar trends may occur for very different reasons, including similar consequences of different trends, offsetting trends, and delays. In the IPCC SRES scenario
set (IPCC 2000), for instance, the issue of similar consequences of different trends
200 | A Manual for Assessment Practitioners
is illustrated by the B1 and A1T scenarios. Both scenarios depict similar greenhouse
gas emissions—but for very different reasons. In the B1 scenario, low emissions are
caused by a preference for renewable energy and environmentally friendly lifestyles
(in combination with favorable technology assumptions). In A1T, however, the only
driver is the fast development of alternative, carbon dioxide neutral technologies.
The issue of offsetting trends plays some role in the outcomes of the Millennium
Ecosystem Assessment scenarios for land use. In most agroeconomic models, a major
driver for technology development for yield improvement is land scarcity. As a result,
the strongest pressure to improve yields exists for a scenario with (initially) high land
scarcity, while the lowest pressure exists for a scenario with (initially) low land scarcity.
In other words, drivers that result in diverging land use (such as population or diets)
are partly offset by this “equilibrium-seeking” mechanism. This is an important reason that the outcomes regarding land use in these scenarios are much closer together
than the original population, income, and lifestyle assumptions would suggest. Similarly, the impact of inertia can be observed in the trends in global mean temperature
increase in the same scenarios: in 2050, despite very different greenhouse gas emission trends, the range of temperature outcomes is still relatively small. The main
reason for this is the inertia in the climate system, which does not allow global mean
temperature to differ much across scenarios in the short term.
Another way of using scenarios in the analysis of response options is the socalled robustness analysis. Here, the focus is not on finding optimal responses based
on a single scenario but on identifying response options that will work against a
range of different scenarios and, as a result, will be more robust regardless of which
scenario is assumed to unfold (see, for example, Lempert et al. 2004).
It should also be noted that the analysis needs to be fit for its purpose with respect to its scale. Scenarios are mostly designed to explore long-term futures using
models that focus mostly on long-term dynamics. As a result, long-term scenarios
in most cases have only limited variation in the short term. This, however, does not
imply that short-term ranges in output variables are small. This range, however,
will result from very different dynamics. For instance, while in the long term energy
prices are mostly driven by technology development and depletion, in the short term
factors like underinvestment, short-term variation in economic growth, and speculation may play an important role.
It is worth noting that not all model results can be used in the same way to present the outcomes of a scenario analysis. While some results may be presented with
relatively high confidence and can serve as backdrop to detailed quantitative analyses,
other outcomes should be seen as purely indicative and lend themselves primarily for
comparison purposes. A good example of an indicative indicator is land use. Clearly,
land use maps should not be used for predictive purposes, given the large uncertainties
that exist regarding future land use. Therefore, land use maps should be communicated in a way that justice is done to the uncertainties that surround the results. In the
EU-Ruralis project (Westhoek et al. 2006, Eickhout et al. 2007), for example, land use
maps are used to pinpoint areas where abandonment occurs in all scenarios. In this
case it is assumed that when land abandonment occurs in all four scenarios, an area is
susceptible to land abandonment irrespective of the scenario assumption.
This process of comparing and analyzing the scenarios then leads into a discussion on response options to meet future challenges. How the scenarios can enhance the (policy) debate on how to tackle the identified problems and along which
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 201
pathways to proceed depends to a certain extent on who the scenario developers
were, on the authorizing environment, and how the scenarios are written up and
disseminated. For example, scenarios built for the purpose of decision making or
strategy development can be more effective if the key stakeholders taking the decision are included in the scenario development process, so as to trigger a discussion
and a reassessment of views and perspectives on the decision to be taken. To stimulate a wider debate on issues at stake, the dissemination of the scenarios can also
be carried out not just via reports or books, but also through such other means as
theater plays or cartoons about the scenarios, movies about them, or other creative,
artistic ways to illustrate the stories, as described in the next section.
5.5 How to use and communicate scenarios
Section’s take-home messages
• Scenario exercises do not end after developing and analyzing a set of scenarios; their
use and active communication is as important as the scenario development process and
needs to be planned well ahead.
• Effective use and communication of scenarios depends on their respective contexts and,
in particular, on whether the scenario process applied was appropriate, the type and
format of scenarios were suitable, an adequate communication strategy exists, and the
scale of analysis fits the context.
• Key features of effective scenarios depend on their main purpose: in order to support
scientific exploration and research, credibility and transparency are essential; for education
and public information, scenarios need to be relevant and accessible; to be effective as
decision support and in strategic planning, scenario exercises need to address relevant
current concerns in a legitimate and challenging manner, which is why the focus is on the
process as much as on the scenarios themselves.
5.5.1 The main dimensions of scenarios use
Developing and analyzing well-crafted scenarios is a first step toward getting a broader,
informed overview on present and future developments, available options for action
and their potential effectiveness and robustness. Translating findings from a scenario
exercise into effective action is the next step. However, experience shows that this is
not a simple or straightforward process (Wilson 2000). The context dependency of
using scenarios makes it difficult to establish a standard recipe for success.
Indeed when the aim is to support decision making or trigger learning processes,
it is difficult to determine and measure success. Is it enough, for example, if scenarios trigger lively discussions for a day? Or should they lead to longer-term changes
in behavior, which are hard to measure? Can it already be seen as a success if policy
makers start to think about longer-term consequences and if scenarios help to better manage conflicts between policy makers and key societal stakeholders? Or has a
scenario exercise only been successful if it really has an impact on decision-making
processes? If so, what is the appropriate time perspective for a successful uptake of
a set of scenarios in the context of decision making?
202 | A Manual for Assessment Practitioners
The ambiguity of answers to these and similar questions underlines the importance of defining clearly the purpose and goals of a scenario exercise at its outset—as
stressed in section 5.2—and to scrutinize the important factors determining successful use. A number of influential factors in this regard should be carefully checked.
A first factor is the appropriateness of scenario development process. Whether
scenarios can be used successfully is to some extent predetermined by their development process. If, for example, scenarios are to be used in support of scientific
exploration or research but are not perceived to be sufficiently credible and fail to
conform to criteria of good scientific practice, achieving success will be difficult.
The perception of saliency, credibility, and legitimacy (as discussed in Chapter 1
and section 5.1) is thus an important influence on the success of scenario exercises.
(Sections 5.5.2 to 5.5.4 discuss in greater detail which of these characteristics are of
particular importance in each context of scenario exercises and which components
thus need particular attention.)
A second factor is the suitability of the type and format of scenarios to their
context and intended use. A normative scenario exercise, for example, is not well
suited when trying to understand the implications of certain decisions depending on
possible future evolutions. Conversely, this type of scenario is well suited to helping
start a broader discussion about the prospects of initiating long-term policy developments to reach a certain target or the work toward changes to current structures
desired by societal groups.
A third factor is the match between context and communication strategy. Here
an analysis of the context within which a set of scenarios will be used is essential:
What is the key target audience, and what kind of communication strategy and
presentation style is suitable for it? A public audience, for example, needs a different presentation format than a scientific audience: triggering lively discussions with
a diverse set of societal stakeholders requires a compelling, attractive presentation
of alternative storylines and their consequences, whereas it is not too important to
present methodological details. The opposite might be true for a scientific audience.
Furthermore, the time requirements might differ: Initializing a strategic conversation
among different stakeholders needs at least a day to present all the scenarios in a
circuit and give participants time to get into the modus of thinking in alternatives.
Identifying key target audiences and defining the suitable communication strategy requires a good analysis of the overall context and corresponding institutional
landscapes: What are the outstanding questions, or problematic issues, the target
audience is facing? Is the topic already well established on the agenda, or is it emerging? Is the target audience long-standing actor networks with well-established belief systems or loose, open networks? What are the time resources and the level of
knowledge? Any communication strategy should include clear criteria that visualize
successful achievements. This will help to monitor and evaluate process. The criteria
should allow for flexibility to adapt in the process but should set minimum standards (such as, for example, engaging relevant policy makers in discussions about
the full set of scenarios in a workshop or meeting). Visualizing a successful outcome
helps people understand the aim of the exercise.
A fourth factor is the fit of (geographical) scale. Whether scenarios can be put to
effective use is also often a question of addressing the appropriate scale and whether
the scenarios address and display trends at the right level—that is, the level of interest and relevance to the intended audience or political decision makers. For example,
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 203
national government officials are first and foremost interested in learning about the
implications for their own country under the respective scenarios.
The remainder of this section discusses in greater detail the conditions of using
scenarios for scientific exploration and research, for education and information, and
for decision support and strategic planning.
5.5.2 How to use scenarios for scientific exploration and research
Within scientific assessments, scenarios often provide a structure for the analysis of
alternative future developments. In addition, scenario-based approaches that combine storyline development with modeling may give indications for those inputs that
cannot be explicitly calculated within the respective models but that are nevertheless
necessary for representing full accounts of plausible future developments (i.e., they
specify exogenous inputs to a model).
The products of a scenario exercise—the actual scenarios—can also be used as
an input into other assessments (Parson et al. 2007). Indeed, to date this appears to
have been the dominant use of scenarios in activities related to international ecosystem assessments (see, for example, the global MA scenarios [MA 2005a] or the
IPCC SRES scenarios [IPCC 2000], both of which have been used in numerous projects as a source of input by subsequent assessments such as Reidsma et al. [2006] or
Rounsevell et al. [2005]).
In addition, scenarios can also be used as input into other research activities
or as a platform to organize inter- and intradisciplinary discussion and exchange.
Scientists external to the scenario development process might be asked to comment
on the scenarios, to highlight gaps in the analysis, and to indicate needs for further
improvement. In the end, this can contribute to establishing an agenda of long-term
research needs with regard to ecosystem assessment.
Appropriateness of scenario development process
If scenarios are to be successfully used for scientific exploration and research, they
first and foremost need to be perceived as credible according to standards of good
scientific practice. Being considered as legitimate and salient is, of course, helpful
but not as important. A note of caution is warranted, however, with regard to the
credibility of scenario exercises. To some extent all scenarios are speculative and will
never fully comply with standards of good scientific practice (in particular, there is
considerable discussion about whether scenarios are reproducible or can be validated). Thus scenarios will never be based on scientific knowledge alone but always
need to blend different forms of expertise and judgment. Therefore transparency
regarding the process, the underlying reasoning assumptions, and the methodology
used—but also plausibility with regard to defining assumptions, choosing data, and
applying models—are of critical importance in order to ensure credibility (see Parson et al. 2007 for this discussion).
Suitability of scenario type and scenario format
In the context of scientific assessments, scenario exercises need to comply with the
overall format requirements of the wider assessment process it is supposed to feed
204 | A Manual for Assessment Practitioners
into (see also Chapters 1 and 2 as well as Box 5.2). A qualitative scenario exercise
will hardly provide relevant information into an otherwise purely quantitative assessment. Past assessments have found the combination of qualitative and quantitative information to be useful. Furthermore, the assumptions and methods used need
to be transparently documented and to be congruent with the overall basis assumptions of the assessment exercise.
Match between type of context and communication strategy
As scenarios provide a structure for future analysis or can be used as input into
other assessments, the key target groups in this context are often other scientists.
Analyzing the political and institutional context to find the relevant target groups is
therefore not too important. But it is important to understand the factors that shape
scientific discourses in the relevant communities, even—or especially—if the aim is
to enrich the discourse or challenge dominant paradigms.
For a scientific audience, the communication should be geared toward the scenario outcomes, as well as the data and methods that have been used in order to
allow a transparent discussion of key assumptions and model applications. Publishing in peer-reviewed academic journals is often a prerequisite for information to be
acknowledged as scientifically sound and trustworthy. The related requirements—
especially with regard to transparency about the methods applied and documentation of underlying assumptions—need to be taken into account when designing
scenario exercises that often cut across disciplinary boundaries.
More and more frequently, the information gained from scenario exercises is
compiled in multimedia presentation tools (be it Web-based or a CD-ROM) that
allow accessible documentation of the wealth of information and methodological
assumptions as well as much of the underlying data. Examples here are the presentation tools for two recent scenario exercises focusing on land use in Europe—that is,
EU-Ruralis (see and PRELUDE (see
In most cases, such tools should not focus on the exact numbers but on the comparison of results between scenarios or regions. One way to do this is via “classical”
graphs or tables. Different presentation methods have been used and developed in
scenario analysis, including vague graphs, traffic lights, and spider diagrams. (Some
examples are provided in Figure 5.6). In a spider diagram, for example, different
scenarios or regions can be displayed in one figure, offering an analysis of the difference between those scenarios or regions. Usually the focus of such analysis is not on
the exact numbers but on the relative differences.
Fit of scale
As described in section 5.3, various methods help link scenarios across different
scales. While a congruence of scale is of course helpful for successful use of the
scenario’s results in an overarching assessment or another study, it is not a precondition. It is more important to have a transparent outline of the key assumptions used,
input data, and methodological approaches at the respective geographical scale to
determine the relevance of the findings for other research purposes. If the scenarios
are to be used to inform long-term research programming, however, the issue of “fit
of scale” becomes more relevant.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 205
Figure 5.6. Different presentations of scenario output from EU-Ruralis study, OECD
Environmental Outlook, and GEO3.
Sources: Rienks 2008; OECD 2008; UNEP 2002
5.5.3 How to use scenarios for education and information
Scenarios are often used to provide policy makers, stakeholders, and the general
public with a structured overview of plausible future developments—that is, they offer an approach to making sense of the myriad of future-related information, helping to create a common language and a common platform for diverse communities
to discuss their respective perceptions of the future. Thus scenarios can contribute
to the framing of the key choices, risks, and opportunities that organizations or
societies face (Starr and Randall 2007). In other words, they have an informational
Quite often, this is combined with an educational function. Policy makers, stakeholders, and the general public often ignore or deny the possibility of far-reaching
change and stick to successful decision modes of the past. Scenarios can help break
such positions of denial by illustrating the plausibility of different trend developments or the consequences of certain decisions or inaction. Normative scenarios
can also be used—and have been used—to frame wider public discussions about
desirable futures that contribute to processes of fundamental social change, as, for
206 | A Manual for Assessment Practitioners
example, the Mont-Fleur scenarios in South Africa (see Kahane 1992). Yet another
purpose is to build capacities for scenario development among public administration, public interest organizations, or corporate stakeholders. The educational function of scenarios has steadily gained in importance over the last few years (see van
Notten 2005).
Involving stakeholders into the process of developing scenarios increases the
educational value and collaborative learning aspects of scenario exercises (see section 5.2). However, the opportunity to directly participate in the process of scenario
development is naturally limited, especially in the case of a broader public on a
national or international scale. The way the scenarios are presented then becomes
crucial. Effective engagement rather than one-way communication is important; participants need to experience the scenarios in a more interactive format. The related
time and resource requirements should not be underestimated. Furthermore, the
scenarios must be described in sufficient detail. This does not necessarily include a
detailed, rich narrative, although this can be helpful. But it always needs a convincing and illustrative representation of the main strands of development (or scenario
mega-trends) and their causes.
Appropriateness of scenario development process
In order to initialize collaborative learning processes and discussions about longterm trends and the consequences of trend development, scenarios first and foremost need to be salient—that is, they must be relevant enough to engage the key
target audience in a discussion about either alternative or desirable futures. It is
also important that the scenarios are perceived to be as legitimate and credible, but
if they fail to present an interesting story they will not succeed in reaching their
intended audiences. It is therefore important to ensure that scenarios present appealing, thought-provoking, compelling yet coherent storylines that relate back to a
clear focal question. This is especially the case for normative scenario exercises that
aim to trigger a wider discussion about desirable futures. Such processes needed to
involve a larger number of stakeholders to provide convincing illustration that the
scenarios present a wider array of societal opinions (see also section 5.2.5 on involving stakeholders).
Suitability of scenario type and scenario format
Generally speaking, long-term contrasting scenarios have a greater potential to challenge and inspire the mind-sets of policy makers, stakeholders, and the general public than projections or reference scenarios do. Also, in most contexts it seems to be
easier to capture the imagination of audiences by presenting compelling qualitative
information than detailed quantitative analysis.
Exploratory scenarios that cover a broader framework of social, technological,
economic, environmental, and political driving forces in a consistent manner are
more suitable to challenge “mental maps” and to challenge participants to think
about “weak signals” or “early warnings” of change. If the focus of discussion is on
desirable or necessary actions—as opposed to feasible or most likely actions—then
normative scenarios seem to be better suited than exploratory ones, as the latter
tend to direct participants toward focusing on feasible action.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 207
Match between type of context and communication strategy
In order to be relevant in the context of education and information, scenarios need
to be presented in an appealing, easy to understand, and communicative format.
The presentation also needs to reach a wide variety of backgrounds and interests. In
the past, various formats have been used in diverse outreach processes: the storylines
have been presented in form of newspaper articles (see, for example, Kahane 1992),
fictive letters from the future (see, for example, NIC 2004), short video teasers (see,
for example, EEA 2007), and interactive Web games or other forms of visualization
(see also Box 5.13). It is not advisable to operate a broader outreach action process
on the basis of a traditional project report only. Having said this, it is important not
to underestimate the time and resources it takes to organize a successful outreach
information and education process. And—once again—the process is as important
as the product.
Outreach activities can be structured along the categories of type of audience
and type of discussion desired. The type of audience can be rather broad and include representatives from different ministries and agencies, research, business, and
nongovernmental organizations or it can be rather narrow and consist of a specific
group of actors from one sector only. Depending on the structure of the target audience, the main questions of interest and the levels of knowledge vary considerably.
Furthermore, the broader the target audience is, the more diverse—and potentially
contradicting and controversial—the belief systems, policy goals, and institutional
routines tend to become. Accordingly, there needs to be more time for broader discussion, which can easily end in a stalemate and thus benefit from being guided by
experienced facilitation (see section 5.2.4).
Box 5.13. The PRELUDE 2 action scenario outreach process
The European Environment Agency organized specific outreach workshops to familiarize different audiences with the concept of scenario development and discuss the specific implications
of the PRELUDE land use scenarios for Europe. The workshops lasted a full day or two days.
Participants were split into working groups and led through presentations of each of the scenarios in a “scenario circuit.”
In this circuit, each scenario was presented in a different room, and discussions were
moderated by external facilitators. Presentations were supported by video-animations that
highlighted key characteristics of the scenarios in an attractive way. Participants discussed
and voted on all scenarios on simple radar charts (high, medium, low) with regard to three main
criteria: relevance, probability, and desirability. After a group had voted, the results of votes in
other groups were shown.
This approach provided a very effective, convincing illustration that individual scenarios
can be perceived in very different ways. Comments for the same scenario often ranged from
strong refusal (“this is not possible”) to strong agreement (“this is already happening”), and the
contradictions of these comments triggered learning processes. The skepticism that participants initially expressed about the overall approach seemed to decrease with the number of
scenarios ”visited” as well as with the time spent exploring these alternative futures (Volkery
et al. 2008).
208 | A Manual for Assessment Practitioners
Whether to aim for a broader or narrower audience depends on the overall
context conditions and the related main aim—that is, whether the aim is agenda
setting for an emerging issue or challenging mind-sets with regard to an established
issue. The first case calls for a broader outreach process. The latter case can benefit
from mixing diverse opinions and benefits (knowledge cross-fertilization), but if it is
necessary to gain confidence it can also be helpful to have the first round of scenario
discussions with a restricted target audience only. In all cases it is important to map
the overall policy context sufficiently: key policy events, including an analysis of
key questions arising in that context, need to be effectively linked into the scenario
presentation process.
Fit of scale
If scenarios are to be effective vis-à-vis their educational function, the “fit of scale”
is not the most important factor. Since the main aim is to engage a wider range of
stakeholders or the public in discussions about alternative futures and their uncertainties, it is more important to have compelling and consistent storylines that help
thinking through “what if” questions.
A “fit of scale” factor is more relevant in the context of the informational function of scenarios—that is, if, for example, local communities should be informed
about the consequences of nonaction or should be provided with a common platform for discussing key choices, risks, and opportunities for maintaining relevant
ecosystem services.
5.5.4 How to use scenarios for decision support and strategic planning
Scenarios can support decision-making processes by providing an analytical framework for finding the suitable, or robust, options with regard to a specific policy
target. Scenarios can function as “test beds” (see Wilson 2000):
• Sensitivity/risk assessments evaluate a specific strategic decision, when the
need for the decision is known beforehand. Using a series of descriptive and
judgmental steps, the process ends with a “go” or “no-go” decision. The series
implies identifying future conditions that would be needed to justify a “go”
decision, assessing these conditions in each scenario, analyzing how successful
the “go” decision would be in each scenario, and finally considering the need
to “hedge” or modify the decision to increase its robustness. This approach
requires a very clear decision focus, and it is easier to implement in the
corporate than the public policy world.
• Strategy evaluations function similarly, but here the scenarios are used as “test
beds” to evaluate the viability of a whole existing strategy or policy (robustness
testing). The strategy or policy is disaggregated into single thrusts (for example
“focus on increasing public awareness,” “focus on improving private liability
schemes,” or “focus on better monitoring”), which are checked for their
relevance and likely successes under the different conditions of the scenarios,
looking for opportunities that the current strategy or policy addresses or misses
and for threats and risks that have not been taken into account. Comparative
analysis should allow common success and failure conditions of the strategy to
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 209
be identified and henceforth the need for a strategy or policy change but also the
need for contingency planning.
However, the complexities and uncertainties of future development might require
the development of completely new policies and strategies or a complete organizational overhaul. Scenarios allow the rehearsal of new policies or strategies, analyzing
related risks and trade-offs and setting priorities for action. This practice intends to
surface long-term blind spots in any policy or strategy and to trigger related learning
and decision-making processes (GBN 2006). Strategy development can be done with
or without a planning focus (Wilson 2000):
• In strategy development with a planning focus, one scenario is chosen—after
analyzing all of them—as a “planning focus” scenario. This is normally the
one regarded as most probable by high-level decision makers. It provides the
basis for formulating a new strategy or policy, which is subsequently tested for
resilience against the other scenarios. Scenario comparison is used to determine
the need for modifying the strategy, “hedging,” and contingency planning.
Although the limitations of a “planning focus” approach are obvious, this
approach can help take first steps toward scenario planning in an environment
that is more rooted in traditional planning.
• In strategy development without a planning focus, each scenario is analyzed
for strategic opportunities and threats that arise, which gives a maximum
feasible range of choices. Comparison across the scenarios helps to identify the
strategic options that are most robust—that perform reasonably well in any of
the scenarios and work best in comparison to all other options considered (see
Groves and Lempert (2007) for a detailed description of a quantitative approach
to support such analyses).
Scenario exercises are thus effective tools to support better decision making for an
uncertain future. The recent literature stresses the process dimension of this endeavor. However, scenario planning is not only a tool to support policy planning,
it can also trigger new ways of thinking in policy and organizational development.
Scenario planning can be run as singular exercises, which make better sense for
existing strategy processes. But it adds real value to decision support if it is institutionalized on a more permanent basis through the creation of a level playing field
for a “strategic conversation” among key actors. This implies moving from a more
ad-hoc style of developing scenarios toward a permanent process of tracking change
in the external environment (van der Heijden 2004).
Furthermore, applying scenarios to decision support and strategic planning is
not simply an add-on activity to a scenario exercise. Instead, it needs to be well
planned and well staffed—and preferably already be planned alongside the scenario
exercise itself. There needs to be a genuine interest in the organization commissioning or undertaking the scenario work in exploring uncertainties in a systematic
way, as well as the necessary buy in from high-level management (see discussion on
authorizing environments in section 5.2.2). Also, it is essential to understand the
organizational context and be able to facilitate complex discussions with clear-cut
questions. A good process management is thus equally important as, if not more
important than, well-crafted analytically rich scenarios.
210 | A Manual for Assessment Practitioners
It is important to pay attention to some other prominent pitfalls when using scenarios in the context of decision making and strategic planning (Parson et al. 2007,
Starr and Randall 2007):
• The challenge of action paralysis: While policy makers might very well be aware
of the need to think in alternatives due to future uncertainties and welcome the
scenarios, they might be overwhelmed by the magnitude of options and abstain
from taking any decisions.
• The challenge of action cosmetics: there is a danger that policy makers instrumentalize scenarios for legitimizing action that has already been taken. The scenarios look nice and attractive, and they symbolize that long-term thinking and
preparing for the future is being taken serious, but their actual content is weak
and they do not really contribute to the decision-making process (van Notten
[2005] calls this the “hollow diamond” effect).
• The challenge of action overshoot: scenarios can misguide political decisionmaking processes if they are used tactically by policy makers, interest groups, or
scientists to highlight problems or foster solutions that are not, but that appear
to be, relevant and adequate because they are illuminated by a new, fancy-looking decision support tool.
Appropriateness of scenario development process
In order to receive buy in and support from high-level political decision makers and
key stakeholders who have not been part of the development process, scenarios need
to be perceived of as legitimate but also salient. The question of legitimacy (“who
has been involved in developing the scenarios?” and “how have the scenarios been
developed and analyzed?”) becomes important to policy makers. This is even more
the case for contested or emerging topics with a probability of high impacts. In such
cases, the process needs to clearly indicate how different actors have been involved.
Furthermore, in a decision-making context scenarios need to show a true added
value—that is, they need to be relevant to the issue at hand and provide a sufficient,
nonstereotypical variety of scenario sets. Accordingly, the conditions of and reasons
for framing the problem, shaping options, and taking decisions need to be spelled
out in a transparent manner. In this context, it is probably less relevant whether the
scenarios fully comply with standards of good scientific enquiry.
Suitability of scenario type and scenario format
The interplay between format and use of scenarios is rather important in a decisionmaking context. For instance, scenarios that address in great detail very specific
questions (especially in terms of scientific exploration) are not useful for generic
decision support. Conversely, scenarios that are rather broad and aim at capturing a
multitude of stressors are likely to be too vague in order to have an efficient impact
on concrete political decision-making processes. Scenarios useful in the context of
decision support should inform about so-called predetermined elements on the one
hand (those forces that can be anticipated with certainty as they are already evolving in their early stages). They frame a strategic corridor in which strategy and
policy formulation can take place (Schwartz 1991). Meanwhile, scenarios also need
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 211
to highlight key uncertainties that could change the validity of key assumptions underpinning strategies or policies. They need to be taken into account to find strategies or policies that are robust enough to function across a wider range of plausible
futures (Dewar 1993). Useful scenarios need to allow for policy discretion—that
is, if the policies are clearly framed and analyzed throughout the different scenarios, there is no opportunity to experiment with different options, and the scenarios
might get rejected as prescriptive.
Also, in a decision-making context, it is important to distinguish between decision unit and decision context in a rigorous manner (or at least try to do so)—that
is, to separate the actions an individual decision unit is taking from the actions that
are out of the immediate reach and frame the wider decision-making context (see
section 5.1). Unfortunately, the evaluative scenario literature on this is rather slim
and offers so far only limited insight into the effectiveness of different scenario types
for decision support. But it is safe to say that exploratory, alternative scenarios are
more helpful to support concrete policy formulation. By contrast, normative scenarios are more helpful in the context of agenda setting and issue framing, with a
view to defining a long-term vision and illustrating the magnitude of the challenge of
moving from the current status to the future desired status.
Match between type of context and communication strategy
As noted earlier, it is crucial to understand who the key target audiences of a scenario exercise are. Government and public administration are not monolithic blocks,
and the expectations toward the scenarios can vary much according to the level of
decision making. Experts on the working level usually prefer a more technical presentation than high-level decision makers do, and while this seems to be a trivial
statement, it is often exactly at this stage where many scenario outreach processes
fail. Moreover, it is important to set the expectations right from the beginning and
to stress that scenario exercises are tools to support decision making. They do not
replace the need to take decisions.
Often, decision makers in government (especially at higher levels) may seem
somewhat biased toward forecasts or reference scenarios forecasts and might be
overwhelmed by sets of scenarios that seem confusing rather than helpful in taking
decisions effectively (“I don’t want choices, I want answers!”). Providing a list of
examples of failed single-point forecasts in the past and a clear-cut illustration of the
way scenarios could be used—rather than presenting a battery of detailed findings—
can help overcome these rather cultural barriers (Wilson 2000).
If scenarios are to guide strategic planning, it is helpful to work with a list of
clear-cut questions that get political decision makers to identify the consequences
of action or nonaction and respective opportunities for policy making, especially
with a focus on their own organization in each scenario, although these might seem
unattractive from today’s perspective. It is relevant to determine who should be targeted and also when. Any action should be guided by a mapping of key policy
events, or milestones of decision making, and the related decision-making needs. If
the respected policy field and its key actors are rather closed and dominated by wellestablished belief systems, the primary task of using scenarios is then to overcome
agency resistance to change, which requires a rather broad, open presentation with
built-in persuasion effects. But scenarios might also be used to rehearse strategies
212 | A Manual for Assessment Practitioners
on how to break opposition by key societal stakeholders and develop alliances with
new stakeholders.
Fit of scale
A fit of scale is highly relevant for scenarios to be useful in decision support and
strategic planning. The implications of different strategic decisions must be clearly
visible and understandable on the right scale. Otherwise, the scenarios will be regarded as nice, even visionary work that is of little added value to concrete processes
of decision making (see Box 5.14).
Box 5.14. Using scenarios “out of context”
Scenarios are not bound to the original thematic context and purpose they have been developed
for. Especially under conditions of time or financial resource constraint that do not allow for developing new scenarios, the results from other exercises can be used “out of context.” Due to
the increasing economic, social, and ecological interlinkages across regions and scales, a set of
scenarios that contains a qualitative description of alternative futures based on a set of different,
yet coherent assumptions about driving forces normally offer enough opportunities to link the
scenarios to their own organizational context and to explore uncertainties of long-term trends.
Using scenarios out of context works best with rather broad thematic scenarios in settings
that aim at general discussions—thus especially for public information and learning, but also
when more indirect forms of decision support are required, such as, for example, clarifying an
issue’s importance, shaking up habitual thinking, or stimulating creativity (Parson et al. 2007).
Direct forms of decision support, such as informing specific decisions, normally require, on the
contrary, tailor-made scenarios that reflect the relevant drivers, uncertainties, and information
needs of the organization.
However, if scenarios are indeed used out of context it is necessary to make sure that
all the relevant information is available that is needed to fully understand the scenarios. It is
not enough to simply take a report from the shelf, as most often this does not offer sufficient
information on the choice and meaning of assumptions, details of analysis, and interpretation
of results. Furthermore, if information is borrowed from different scenario sets, a key requirement is to ensure a basic compatibility of scenario assumptions to avoid inconsistencies and
contradictions. Especially in larger outreach settings, participants are bound to scrutinize the
scenarios in detail, and a failure to provide clear answers can easily undermine the legitimacy
of the process.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 213
Quick Reference Charts
214 | A Manual for Assessment Practitioners
Related resources
• see sections 5.1.3, 5.2.2
• see chapters 1 and 2
• see sections 5.1.3, 5.2.3
and 5.4.1
• see section 5.2.1
• see sections 5.1.2, 5.2.3
• see sections 5.2.4 and 5.2.5
• see chapter 2
Related resources
• see sections 5.2.2, 5.3.2
• see chapter 2
• see section 5.3.1
• see section 5.3.3
• see chapter 4
• see Jaeger et al., 2007
• see Schwartz 1991
• see section 5.3.4
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 215
Related resources
• see sections 5.1.3, 5.2.2
• see sections 5.4.1, 5.4.2, 5.4.4 (what and how)
• see EEA, 2001
• see section 5.4.2
• see MA 2005a
• see UNEP 2002
• see section 5.4.3
• see MA 2005a
• see UNEP 2002
Related resources
established?), and information
• see sections 5.1.3, 5.2.2
• see chapters 1 and 2
• see sections 5.2.1, 5.5.2,
5.5.3, 5.5.4
• see sections 5.5.2, 5.5.3,
• see chapters 1 and 2
• see chapters 1 and 2
216 | A Manual for Assessment Practitioners
Alcamo, J., and T. Henrichs. 2008. Towards guidelines for environmental scenario analysis. In
Environmental futures: The practice of environmental scenario analysis, ed. J. Alcamo. Oxford:
Alcamo, J., R. Leemans, and E. Kreileman, eds. 1998. Global change scenarios of the 21st century:
Results from the IMAGE 2.1 model. Oxford: Elsevier.
Alcamo, J., T. Henrichs, and T. Rösch. 2000. World water in 2025—Global modelling and scenario
analysis for the World Water Commission on Water for the 21st Century. Report A0002, Kassel,
Germany: Center for Environmental Systems Research, University of Kassel.
Armstrong, J., ed. 2001. Principles of forecasting: a handbook for researchers and practitioners.
Norwell, MA: Kluwer Academic Publishers.
Börjeson, L., M. Höjer, K. Dreborg, T. Ekvall, and G. Finnveden. 2006. Scenario types and techniques: Towards a users guide. Futures 38 (7): 723–39.
Bouwmans, R., R. Costanza, J. Farley, M. A. Wilson, R. Portela, J. Rotmans, F. Villa, and M.
Grasso. 2002. Modeling the dynamics of the integrated earth system and the value of global
ecosystem services using the GUMBO model. Ecological Economics 41:529–60.
Carpenter, S., E. Bennett, and G. Peterson. 2006. Scenarios for ecosystem services: An overview.
Ecology and Society 11 (1): 29.
Cosgrove, W., and F. Rijsberman. 2000. World water vision: Making water everybody’s business.
London: Earthscan Publications.
Dewar, J. 1993. The importance of “wild card” scenarios. Santa Monica, CA: RAND.
Ducot, C., and H. Lubben. 1980. A typology for scenarios. Futures 12 (1): 15–57.
EEA (European Environment Agency). 2001. Scenarios as tools for international environmental
assessments. Environmental issues series No 24. Copenhagen: EEA.
EEA. 2005. European environment outlook. Report 2/2005. Copenhagen: EEA.
EEA. 2006. Urban sprawl in Europe. Briefing 04-2006. Copenhagen: EEA.
EEA. 2007. Land use scenarios for Europe: Modelling at the European scale. Technical report No
09/2007. Copenhagen: EEA.
Eickhout, B., H. van Meijl, A. Tabeau, and T. van Rheenen. 2007. Economic and ecological consequences of four European land use scenarios. Land Use Policy 24 (3): 562–75.
Evans, K., S. J. Velarde, R. Prieto, S. N. Rao, S. Sertzen, K. Davila, P. Cronkleton, and W. de Jong.
2006. Field guide to the future: Four ways for communities to think ahead. Nairobi: Center for
International Forestry Research, ASB, and World Agroforestry Centre.
FAO (Food and Agriculture Organization). 2002. World agriculture: Towards 2015/2030—An
FAO perspective. London: Earthscan Publications.
Gallopin, G., and F. Rijsberman. 2000. Three global water scenarios. International Journal of
Water 1 (1): 16–40.
Gallopin, G., A. Hammond, P. Raskin, and R. Swart. 1997. Branch points: global scenarios and
human choice. Stockholm: Stockholm Environment Institute.
GBN (Global Business Network). 2006. Scenarios, strategy, and action. Guide & workbook. San
Francisco: GBN.
Gordon, T., and H. Hayward. 1968. Initial experiments with the cross-impact matrix method of
forecasting, Futures 1:100–16.
Groves, D. G., and R. J. Lempert. 2007. A new analytic method for finding policy-relevant scenarios. Global Environmental Change 17:78–85.
Grübler, A., and N. Nakicenovic. 2001. Identifying dangers in an uncertain climate. Nature
Grübler, A., B. O’Neill, and D. van Vuuren. 2006. Avoiding hazards of best-guess climate scenarios.
Nature 440 (7085):740.
Helmer, O. 1983. Looking forward: A guide to futures research. London: SAGE Publications.
Hughes, B. 1999. International futures: choices in the face of uncertainty. Oxford: Westview
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 217
IEA (International Energy Agency). 2007. World energy outlook. Paris: IEA.
IPCC (Intergovernmental Panel on Climate Change). 1992. Climate change 1992: Supplementary
report to the IPCC Scientific Assessment. Cambridge, U.K.: Cambridge University Press.
IPCC. 2000. Special report on emission scenarios. Cambridge, U.K.: Cambridge University Press.
IPCC. 2001. Climate change 2001—Synthesis report. Cambridge, U.K.: Cambridge University
IPCC. 2007. Climate change 2007—IPCC fourth assessment report. Cambridge, U.K.: Cambridge
University Press.
Jaeger, J., D. Rothman, C. Anastasi, S. Kartha, and P. van Notten. 2007. Training Module 6:
Scenario development and analysis. In GEO resource book: A training manual on integrated
environmental assessment and reporting. Nairobi and Winnipeg, Canada: U.N. Environment
Programme and International Institute for Sustainable Development.
Jungk, R., and N. Müllert. 1987. Future workshops: How to create desirable rutures. London:
Institute for Social Inventions.
Kahane, A., ed. 1992. The Mont Fleur scenarios—What will South Africa be like in 2002? GBN
Deeper News 7(1). Emeryville, CA: Global Business Network.
Kahn, H. 1960. On thermonuclear war. Oxford: Oxford University Press.
Kahn, H., and A. Wiener. 1967. The year 2000. New York: Macmillan.
Kok, K., R. Biggs, and M. Zurek. 2007. Methods for developing multiscale participatory scenarios:
Insights from Southern Africa and Europe. Ecology and Society 13 (1): 8.
Leemhuis, J. 1985. Using scenarios to develop strategies. Long Range Planning, 18 (2): 30–37.
Lempert, R., N. Nakicenovic, D. Sarewitz, and M. Schlesinger. 2004. Characterizing climate-change
uncertainties for decision makers. Climatic Change 65: 1–9.
MA (Millennium Ecosystem Assessment). 2005a. Ecosystems and human well-being: Scenarios.
Washington DC: Island Press.
MA. 2005b. Ecosystems and human well-being: Multiscale assessments. Washington DC: Island
Meadows, D. H., D. L. Meadows, J. Randers, and W. W. Behrens 1972. The limits to growth. New
York: Universe Books.
Mesarovic, M. and E. Pestel. 1974. Mankind at the turning point. New York: Dutton.
Mitchell, R., W. Clark, D. W. Cash, and N. Dickinson, eds. 2006. Global environmental assessments: Information, institutions, and influence. Cambridge, MA: The MIT Press.
NIC (National Intelligence Council). 2004. Mapping the global future. Report of the National Intelligence Council’s 2020 project. Washington, DC: U.S. Government Printing Office.
OECD (Organisation of Economic Co-operation and Development). 2008. OECD’s environmental
outlook to 2030. Paris: OECD.
Ogilvy, J., and P. Schwartz. 1998. Plotting your scenarios. In Learning from the Future, ed. L. Fahey
and R. Randell. New York: John Wiley & Sons.
Parson, E., V. Burkett, K. Fisher-Vanden, D. Keith, L. Mearns, H. Pitcher, C. Rosenzweig, and M.
Webster. 2007. Global-change scenarios: Their development and use. Report for the U.S. Climate Change Science Program and the Subcommittee on Global Change Research.
Pereira, E., C. Queiroz, H. Pereira, and L. Vicente. 2005. Ecosystem services and human well-being:
A participatory study in a mountain community in Portugal. Ecology and Society 10 (2): 14.
Pielke, R., T. Wigley, and C. Green. 2008. Dangerous assumptions. Nature 452: 531–32.
Raskin, P. 2005. Global scenarios: Background review for the Millennium Ecosystem Assessment.
Ecosystems 8 (2): 133–42.
Raskin, P., G. Gallopin, P. Gutman, A. Hammond, and R.Swart. 1998. Bending the curve: Towards
global sustainability. Stockholm: Stockholm Environment Institute.
Raskin, P., T. Banuri, G. Gallopin, P. Gutman, A. Hammond, R. Kates, and R. Swart, R. 2002.
Great transitions: The promise and lure of times ahead. Boston: SEI-B / Tellus Institute.
Reidsma, P., T. Tekelenburg, M. van den Berg, and R. Alkemade. 2006. Impacts of land use change
on biodiversity: An assessment of agricultural biodiversity in the European Union. Agriculture,
Ecosystems and Environment 114:86–102.
218 | A Manual for Assessment Practitioners
Richels, R. G., A. S. Manne, and T. M. L. Wigley. 2004. Moving beyond concentrations: the challange of limiting temperature change. AEI–Brooking Joint Center for regulatory studies. Washington DC: Brookings Institution.
Rienks, W. A. 2008. The future of rural Europe. An anthology based on the results of the Eururalis
2.0 scenario study. Wageningen, The Netherlands., Wageningen University Research and Netherlands Environmental Assessment Agency.
Ringland, G. 2002. Scenarios in public policy. Chichester, U.K.: John Wiley & Sons.
Rothman, D. 2008. Environmental scenarios—A survey. In Environmental futures: The practice of
environmental scenario analysis. Ed. J. Alcamo. Oxford: Elsevier.
Rotmans, J., and B. de Vries, eds. 1997. Perspectives on global futures: the TARGETS approach.
Cambridge, U.K.: Cambridge University Press.
Rounsevell, M. D. A., F. Ewert, I. Reginster, R. Leemans, and T. R. Carter. 2005. Future scenarios
of European agriculture land use. II. Projecting changes in cropland and grassland. Agriculture,
Ecosystems & Environment 107:117–35.
Scearce, D., K. Fulton, and the Global Business Network community. 2004. What if? The art of
scenario thinking for nonprofits. Emeryville, CA: Global Business Network.
Schneider, S. H. 2001. What is “Dangerous” Climate Change? Nature 411:17–19.
Schneider, S. H. 2002. Can we estimate the likelihood of climatic changes at 2100? An editorial
comment. Climatic Change 52:441–51.
Schwartz, P. 1991. The Art of the long view. New York: Currency Doubleday.
Starr, J., and D. Randall. 2007. Growth scenarios: Tools to resolve leaders’ denial and paralysis.
Strategy and Leadership 35 (2): 56–59.
Stewart, J. 1993. Future state visioning—A powerful leadership process. Long Range Planning 26
(6): 89–98.
Toth, F., E. Hizsnyik, and W. Clark. 1989. Scenarios of socioeconomic development for studies of
global environmental change: A critical review. IIASA Report RR-89-4. Laxenburg, Austria:
International Institute of Applied Systems Analysis.
UNEP (United Nations Environment Programme). 2002. Global environment outlook 3. London:
Earthscan Publications.
UNEP. 2007. Global environment outlook 4. London: Earthscan Publications.
van der Heijden, K. 2004. Scenarios: the art of strategic conversation. New York: John Wiley & Sons.
van Notten, P. 2005. Writing on the wall: Scenario development in times of discontinuity,, Florida.
van Notten, P., J. Rotmans, M. van Asselt, and D. Rothman. 2003. An updated scenario typology.
Futures 35:423–43.
van Vuuren, D., M. den Elzen, P. Lucas, B. Eickhout, B. Strengers, B. van Ruijven, S. Wonink, and
R. van Houdt. 2007. Stabilizing greenhouse gas concentrations at low levels: An assessment of
reduction strategies and costs. Climatic Change, 81:119–59.
van Vuuren, D., et al., 2008. Forthcoming.
Volkery, A., T. Henrichs, Y. Hoogeveen, and T. Ribeiro, T. 2008. Your vision or my model? Lessons from participatory land use scenario development on a European scale. Systemic Action &
Practice Research, forthcoming.
Wack, P. 1985. The gentle art of reperceiving. Harvard Business Review (article in two parts).
September–October:73–89; November–December:139–50.
WBCSD (World Business Council for Sustainable Development). 1997. Exploring sustainable development: WBCSD global scenarios 2000–2050 summary brochure. Geneva: WBCSD.
Webster, M. D., M. H. Babiker, M. Mayer, J. M. Reilly, J. Harnisch, R. Hyman, M. C. Sarofim,
and C. Wang, C. 2002. Uncertainty in emissions projections for climate models. Atmospheric
Environment. 36 (22): 3659–70.
Webster, M. D., C. Forest, J. M.Reilly, M. Babiker, D. Kickligher, M. Mayer, R. Prinn et al. 2003.
Uncertainty analysis of climate change and policy response. Climatic Change 61:295–320.
Scenario Development and Analysis for Forward-looking Ecosystem Assessments | 219
Westhoek, H., M. van der Berg, and J. Bakkes. 2006. Scenario development to explore the future of
Europe’s rural areas. Agriculture, Ecosystems & Environment 114 (1): 7–20.
Wilkinson, A., and E. Eidinow. 2008. Evolving practices in environmental scenarios: a new scenario
typology. Environmental Research Letters 3.
Wilson, I. 2000. From scenario thinking to strategic action. Technological Forecasting and Social
Change 65 (1): 23–29.
Wollenberg, E., D. Edmunds, and L. Buck. 2000. Anticipating change: Scenarios as a tool for adaptive forest management: A guide. Bogor, Indonesia: Center for International Forestry Research.
WRI (World Resources Institute). 2008. Ecosystem services—A guide for decision makers. Washington, D.C.: WRI.
Zurek, M., and T. Henrichs. 2007. Linking scenarios across geographical scales in international
environmental assessments. Technological Forecasting and Social Change 74 (8): 1282–95.
Assessing Intervention Strategies
R. David Simpson and Bhaskar Vira
What is this chapter about?
This chapter looks at the effectiveness of strategies that respond to the degradation of ecosystems that provide services to society. First we consider who might develop strategies for
responding to ecological degradation and what considerations and constraints might shape
the strategies they can choose. The choice among strategies inevitably involves trade-offs
between competing objectives. The chapter looks at some techniques practitioners may use
in evaluating such trade-offs. Then we consider several strategies that have already been used
to address the degradation of ecosystems. This discussion is intended to assist practitioners
who are considering different types of intervention options by highlighting the advantages as
well as the pitfalls of various alternatives. The final main section discusses how the results of
assessment procedures may be used to inform decision-making processes, as well as how
the attributes of decision-making processes may affect the motivations for, and uses of, assessment procedures.
6.1 How to choose from a menu of possible strategies
Section’s take-home messages
• It is important to ensure congruence between the diagnosis of a problem, the capacity of
agents to respond to the problem, and the scale at which an intervention is implemented.
• For interventions to be effective, knowledge about the problem is fundamental, but it
needs to be supported by appropriate institutions and social contexts to ensure that
instrumental interventions achieve their desired objectives.
• If knowledge is incomplete, institutions are imperfect, or social conditions are
inappropriate, practitioners may not be in a position to implement desirable strategies
without addressing these fundamental constraints.
• The five arenas of action that actors can potentially be involved in are provision of
knowledge, reform of institutions and governance, societal and behavioral innovation, use
of markets and incentives, and development of improved technologies.
222 | A Manual for Assessment Practitioners
• Identification of constraints helps to focus attention on the enabling conditions that are
necessary for an intervention to work.
In deciding on the best response, if any, to ecological degradation at a particular
time and place, several interrelated questions must be addressed:
• What is the ecosystem change/loss in human well-being that needs to be
addressed, and why?
• Who will respond?
• Which strategies will they choose?
• How will these strategies be structured?
• What will their effects be on both ecosystems and human well-being?
These key questions are addressed in this chapter, starting in this section with the
first three questions.
6.1.1 How to determine who can and should respond
Whether one wishes to evaluate the success of a measure that has been implemented
in the past or the likely efficacy of one that is now being contemplated, a critical
question is “Does this strategy propose an effective measure for dealing with the
problem at hand?” To answer this question, one must first understand what the
problem is. For our purposes, the important point is to highlight that problem identification and definition are usually the first stage of a planned intervention or evaluation strategy. Furthermore, decision makers usually frame problems based on their
own perceptions, knowledge, and understanding of the drivers of change (Adams
et al. 2003). It is possible that different decision makers will focus on different elements of a perceived empirical reality; this framing of the problem is also likely to
inform the range of solutions that are explored.
Strategies that respond to ecological degradation and the consequent loss of human well-being will be futile if those planning them do not have the capability to
implement them. In this sense, “capability” may refer to the understanding needed
to address a complex problem, the political, organizational, or economic power to
overcome resistance, or a combination of the two. A good place to begin to consider
these matters is with a discussion of “scale.”
Local events can have global implications. Even when physical effects cannot
be demonstrated to occur over great distances or from one time to another, local
actions can still have consequences for the global community. Someone may derive
no material benefit from the continuing survival of rhinoceroses in Africa or polar
bears in the Arctic, for example, but may still care enough about preventing their extinction to be willing to contribute to their survival. Equally, a number of processes
that occur at local scales may have profound local implications, but they need not
necessarily cascade to other scales of impact; localized contamination of lands due
to industrial effluent may not necessarily affect a very large area. Such local issues
are in no way less important, but the level at which the problem is perceived, and
the strategies to respond to the problem, are likely to be operating at a very different
Assessing Intervention Strategies | 223
scale to an issue that commands regional, national, or global attention. Hence, the
question of scale (of impact and response) is critical.
While local actions may have global implications, altering local incentives is sometimes sufficient to solve problems of global importance. Consider, for example, overfishing. While the depletion of fisheries may have global implications for ecosystem
function, an effective intervention may in some instances involve only convincing local
fishers that it is in their interest to conserve the resource on which they depend.
Thus, in evaluating the effectiveness of strategies for conservation and preservation of ecosystem services, a critical issue is whether such strategies have been
implemented on a spatial and temporal scale sufficiently broad as to lead to effective
action. This is a matter not only of geography and time but of political boundaries
and knowledge as well.
Bohensky and Lynam (2005) provide a useful framework for thinking about
these issues, which we will borrow here. They note that effective strategies arise
when impact scope, awareness, and power converge. Impact scope is the spatial
and temporal scale at which the consequences of ecosystem change are felt. It is
necessarily a matter of degree and interpretation, and the scope of analysis may
vary depending on the aspect of the problem being considered. Awareness is simply
knowing that ecosystem change is occurring and driving changes in well-being. Different actors may cognitively frame these changes very differently, thereby shaping
their perceptions of the problem. Power is the ability to effect change. “Power”
means that affected people are able to either address the problem themselves or to
provide sufficient incentives to others to change their behavior.
It is easy to identify situations in which impact scope, awareness, and power
overlap as well as others in which they do not. Timber concessions and oil palm
plantations in the tropics, for example, often result in deforestation in areas in
which local people depend on ecosystem services such as water purification, erosion
protection, and habitat for harvested wildlife. Local people are often aware of the
benefits they obtain from existing forests, but they may lack the power to prevent
land clearing. If urban or international interests stand to benefit from deforestation,
and these interests are politically more powerful than community-based or indigenous groups, no effective response may be possible.
These possibilities are illustrated in Figure 6.1. The top figure illustrates the situation in which a local community is aware of the effects of deforestation but does
not have the power to resist it. In the bottom figure we envisage a situation in which
two things have happened. First, the power of the local community has increased,
as represented by an increase in the size of the most darkly shaded ellipse. Second,
the local community has acquired a greater ability to affect outcomes, as represented
by a shift in the most darkly shaded ellipse to overlap the “awareness” and “impact
scope” ellipses more. This may have resulted from both an increase in its overall
authority and the transfer to local groups of some authority that formerly lay with
external actors, for instance through the introduction of decentralized and devolved
forms of forest governance.
This way of thinking highlights the need to have the knowledge, institutional, and
social prerequisites for effective strategies in place and to ensure that these converge in
terms of impact scope, awareness, and power. The next question concerns identifying
which strategies are most appropriate and the feasibility of the chosen options.
224 | A Manual for Assessment Practitioners
Figure 6.1a. Impact scope, awareness,
and power.
Source: Adapted from Bohensky and
Lynam (2004)
Figure 6.1b. Responses can be made
more effective if power is increased and/
or transferred within the impact scope.
Source: Adapted from Bohensky and
Lynam (2004)
6.1.2 How to decide which strategies are appropriate
The Millennium Ecosystem Assessment (MA) identifies five types of problems that
might result in the degradation of ecosystems and motivate strategies (MA 2005a):
• Inappropriate institutional and governance arrangements, including the presence
of corruption and weak systems of regulation and accountability;
• Market failures and the misalignment of economic incentives;
• Social and behavioral factors, including the lack of political and economic
power of some groups (such as poor people, women, and indigenous peoples)
that are particularly dependent on ecosystem services or harmed by their
• Underinvestment in the development and diffusion of technologies that could
increase the efficient use of ecosystem services and could reduce the harmful impacts of various drivers of ecosystem change; and
• Insufficient knowledge (as well as the poor use of existing knowledge)
concerning ecosystem services and management, policy, technological,
behavioral, and institutional strategies that could enhance benefits from these
services while conserving resources.
With such a wide range of problems, it is only to be expected that deciding when
and where to apply pressure for strategies will sometimes be difficult. In general, the
choices of who should respond and how depend on context and the capabilities of
Assessing Intervention Strategies | 225
To see which actors might best be involved in which contexts, consider the five
arenas of actions suggested. Summarizing briefly, they are provision of knowledge,
reform of institutions and governance, societal and behavioral innovation, the use of
markets and incentives, and the development of improved technologies.
Figure 6.2 presents a schematic representation of these potential actions. Focus
for now on the left side of the figure. The far-left column presents levels or “tiers” of
potential actions. The first tier is fundamental. An adequate knowledge base for action is a prerequisite for taking any meaningful action. Without sufficient knowledge
of biological and social facts it would be impossible to structure effective strategies.
Below fundamental knowledge, two categories of actions are listed that will, in turn,
enable more instrumental strategies. These are the categories labeled “institutions
and governance” and “society and behavior.” We characterize the function of these
strategies as “enablement,” as actions in these realms establish the preconditions for
enacting instrumental strategies. This last category, “instruments” describes specific
actions taken to achieve particular ends, which include the deployment of technological interventions and/or the use of markets and incentives.
To give some examples, an instrumental (third tier) response might be to pass
regulations intended to restrict deforestation. Such regulations would not be effective without second-tier prerequisites, such as a legal system capable of enforcing
them, or a “culture of compliance” in which citizens generally accept the decisions
of their government as fair and just—and abide by them. Another instrumental response to ecosystem change might be to subsidize research on more benign agricultural technologies. Again, such subsidies would not be effective absent certain social
and institutional preconditions, such as the existence of a university system embodying the required scientific expertise or social norms under which the best qualified
individuals can pursue careers in research, regardless of background, economic status, ethnicity, or gender.
Figure 6.2 presents a sort of cascade, in which knowledge creates the preconditions for institutional and social strategies, which in turn create the preconditions
for specific instrumental strategies. Thus a sequential procedure would involve identifying whether there is appropriate foundational knowledge, checking whether the
appropriate enabling conditions are in place, and then identifying if relevant and
appropriate instruments can be deployed.
There are, however, feedbacks. For example, technological improvements can
lead to the generation of knowledge that will guide other strategies, as well as for
the reform of institutions and social relations. We will suppose, however, that the
foundational and enabling prerequisites must be in place if instrumental strategies
are to be effective.
Note that the far-right column in Figure 6.2 is labeled “initiating actors,” to
distinguish between actors who might participate in particular approaches or be affected by them and those who would typically initiate them. This is in keeping with
the general idea that, in assessing strategies, it is important to know who must be
given an incentive to act in order for the strategy to be taken up. That is, who must
make the policy choices required to get the action under way?
The identification of initiating actors is based on political and economic considerations. The essential idea here is that each type of action represents a kind of public
good: something that, when provided by one actor, is available to many. Knowledge
has this character: when knowledge is in the public domain, it is available for the
226 | A Manual for Assessment Practitioners
Figure 6.2. Responses and actors.
benefit of all. A similar statement might be made concerning good governance: reforms that improve the functioning of government provide benefits to all of the
governed—that is, to everyone. The ecosystem services of concern here are also generally public goods: enhancements made by one person benefit many. There is, of
course, an important caveat here, since while some types of collective goods benefit
many, they do allow for the exclusion of certain groups, and this may not always be
equitable or fair. We should not assume that all ecosystem services are equally accessible to all groups in society.
Conventional wisdom has it that public goods must be provided by public authorities. No one individual or group has the incentive to provide a public good, as
they will bear all the costs but appropriate only some of its benefits (i.e., others will
ride free on the private provision of a public good). Therefore, in many instances we
identify one or another level of government as the most likely and appropriate locus
for initiating an intervention strategy.
Our focus on public entities as initiating agents is not intended to minimize the
role of other actors. Once a strategy is set in motion by one type of actor, some or
every other type might play useful roles. For example, local governments may enact environmental regulations, but the laws’ efficacy depends on compliance among
Assessing Intervention Strategies | 227
community and business actors. Moreover, private-sector and civil society actors
interact with public-sector actors to set the social agenda. It is therefore a simplification to treat public entities as initiating agents, when public priorities are generated
by a complex interaction of societal interests. Practitioners may want to enlist business or community constituencies in order to generate support for particular strategies. This section focuses, however, mostly on those actors who can directly initiate
The acquisition of basic knowledge has been and is likely to continue to be
funded most prominently by national governments. While there are compelling arguments to be made that the spillovers of knowledge are global, international cooperation in research financing remains relatively rare. Private industry has also, on
occasion, been a major funder of basic research, and that may continue. Nongovernmental organizations (NGOs) and community-based and indigenous organizations
may also play a role in this area, although knowledge that is embedded within such
organizations is often not more widely used until it has been formalized and published, usually by outsiders (who act as promoters or propagators of this knowledge
and in some cases illegitimately appropriate it for their own selfish ends). Research
organizations are often very important nodes in the knowledge process, but they
are usually responsive to external needs for new products or processes rather than
initiators of the information acquisition process. “Blue skies” research, when it occurs, is also usually supported by government or private foundations, and it is rare
for institutions to have resources that enable an investment in such fundamental
research without external support.
In the second tier, national and local governments and community-based and
indigenous organizations are identified as the most likely initiators of reforms in
governance and social relations. The need for action can arise at any of several
levels: internationally, for example, to address greenhouse gas emissions and global
warming; at the level of the nation or region, to address major pollutants or major
ecosystem disruption; or at the level of local pollution or habitat issues. While the
general conclusion of extensive literature on this issue (see, e.g., Olson 1965, Ostrom 1990, Baland and Platteau 1996) is that effective management is more likely
in smaller and more homogeneous groups, incorporating the needs of all affected
parties (often referred to as stakeholders) in decision making is a prerequisite for
dealing with ecological issues at any level.
In addition to government and community organizations, NGOs have on occasion provided the impetus for institutional and social reforms. Research organizations are also listed as an initiating agent, as the findings of social scientists
and other researchers have sometimes spurred action. While international action
would be appropriate in many instances, issues of sovereignty have often led international efforts at influencing domestic institutional reform to be less successful or
less ambitious.
In the third tier, that of instrumental strategies, action is generally initiated by
national or local governments. This is almost a tautology in the case of regulations, which are imposed and enforced by government. International agreements
have been pivotal in many strategies, such as the Montreal and Kyoto Protocols
and the Convention on Biological Diversity. These international agreements have
often contained provisions for particular instruments, such as trading in emissions
or credit for reductions under the Clean Development Mechanism of the Kyoto
228 | A Manual for Assessment Practitioners
Protocol. In some instances community-based or indigenous organizations adopt
specific instrumental strategies, such as restrictions on harvesting or resource use.
NGOs and research organizations have also been active in some instrumental strategies, particularly in developing, promoting, and distributing technology.
The distinction between strategies in the second and third tiers of Figure 6.2 is
generating the conditions to enable more specific strategies and adopting specific
instruments tailored to particular ecological circumstances. Each response has its
merits and drawbacks. The virtue of a more direct response—passing regulation to
reduce deforestation, for example, or subsidizing research on sustainable agricultural technologies—is that it focuses on the problem of interest rather than matters
that may, from an ecological perspective, be more peripheral. In other words, if an
instrumental approach “works,” the problem will be “solved.”
The potential drawback is that such direct approaches presume certain institutional and social infrastructure. For instance, regulations restricting deforestation presume a competent regulatory authority, fair and consistent enforcement,
and private actors who will be dissuaded from illegal action by the threat of sanction. Similarly, subsidies to research presume a research sector capable of producing
Would it be more effective to respond to wasteful degradation of ecosystem services with institutional or social and behavioral interventions rather than regulatory
and technological ones? There are pluses and minuses. On the positive side, such
strategies are intended to create the preconditions for the subsequent application
of direct strategies. On the negative side, institutional and social strategies may dissipate the effect of resources intended to be devoted to the preservation of imperiled ecosystems. Consider, for example, a couple of strategies reviewed in the MA:
“property right changes” and “recognition of gender issues” (MA 2005b). Although
few people would dispute that assigning unambiguous ownership of resources and
assuring equality of opportunity are socially desirable goals, neither response was
rated fully “effective” by the Millennium Ecosystem Assessment. One reason cited
was that neither response in and of itself assures the enhancement of the targeted
ecosystem services or improvements in human well-being. If these strategies are to
be effective, it is necessary to ensure that those assigned property rights in natural
ecosystems attach greater value to preserving ecosystems than to converting them
to other uses (see Box 6.1). Similarly, if recognition of women’s rights and interests
is to be an effective response to ecological degradation, it must be ensured that the
women empowered as a result will choose preservation over other uses (see Box
6.2). If these assumptions are not met, then these strategies—desirable though they
are for a number of other societal reasons—cannot be assessed as effective for ecosystem service preservation.
In deciding between the institutional and behavioral approaches of the second
tier and the instrumental strategies of the third tier, potential actors will want to devote special attention to the alignment of awareness with power. It may well be that
institutional or social reform is essential for the achievement of ecological goals, but
some actors devoted to the achievement of ecological goals will lack the capacity
and access to implement such reforms.
Figure 6.3 summarizes a decision process for determining appropriate intervention strategies. It involves answering three questions: “Is the problem well understood?” “Are institutions and governance adequate?” “Are social conditions
Box 6.1. Farm forestry in India
In the mid-1970s, policymakers across the world perceived an impending fuelwood crisis
caused by rapid increases in population (especially in rural areas in poor countries) accompanied by a decline in area under forest cover. The Indian government responded to this crisis
by introducing social forestry programs on a large scale. One element of these programs was
to provide incentives for small and marginal farmers to cultivate fuelwood trees on their own
lands, for their own use, thereby relieving pressure on remaining natural forests. This “farm
forestry” program was promoted through the provision of subsidized seedlings and technical
support by government forest departments.
Subsequent evaluations suggested that the farm forestry program was adopted by farmers, but typically by large farmers who perceived trees as a valuable cash crop (as raw material
for the paper and matchwood industries), not by small and marginal farmers (for whom growing
trees was not the best option on their limited landholdings). The desired outcome, which was
to be a decline in pressure on natural forests for rural fuelwood needs, was not achieved, although there was a substantial increase in tree cover (but typically of species that had a market
value). The intervention was unsuccessful in alleviating pressure on natural forests as a source
of rural energy supply, so it failed to deliver benefits for either the natural ecosystem or the wellbeing of poor rural households.
Box 6.2. Water, work, and women in rural India
In the Indian state of Gujarat, women spend long hours of backbreaking labor gathering and
carrying water to their homes. The quality and reliability of water supply could be made considerably better by instituting ecological and engineering interventions. Traditionally, however,
because women did not have social standing or political power, these improvements often
were not realized.
In 1986 the State Water Board invited the Self-Employed Women’s Association (SEWA)
in Gujarat to help improve the water supply in rural areas. SEWA initiated the Water, Work and
Women program. Local groups known as pani samitis (“water users’ committees”) cooperate
to maintain and manage local water sources. The program has both required and motivated
social change. To give one example, some local men opposed their wives’ or daughters’ involvement in the program because it required women from different castes to interact. One of
the many consequences of the institutional and social reform that relaxed such restrictions has
been the improvement of water supply.
Although SEWA and its Water, Work and Women program have achieved great results both
in providing water to communities and advancing the rights of women, practitioners concerned
with the ecological consequences of such institutional reforms should think broadly about such
matters. Watershed protection and restoration—the preservation of ecological services—is
just one of several activities conducted by the pani samitis. They also line ponds with plastic
to prevent salinization, install pumping and piping, and make other improvements. Moreover,
among the consequences reported is an increase from one to three crops planted per year in
areas with improved water supply. While such developments testify to the tremendous social
and economic benefits of the program, practitioners focused specifically on improving natural
ecosystems and their resilience may want to consider the broader effects of such institutional
Source: Kapoor 2003, Panda 2007.
230 | A Manual for Assessment Practitioners
Figure 6.3. A decision process to choose types of intervention strategies.
(especially participation and equity) acceptable?” If these questions can be answered
affirmatively, practitioners may go on to crafting specific instrumental strategies. If
they cannot be answered in the affirmative, practitioners must decide if they have the
competence and power to intervene at a more fundamental level, or if they should
instead focus their efforts elsewhere.
The implicit starting point of the analysis is the perception that there is a problem: ecosystem services are being lost, with consequent impacts on human well-being. Then the first logical question is, “Is enough known about the problem to devise
a response to it?” There have been innumerable instances in which the answer to
that question was “No” (see Box 6.3).
But if problems are well enough understood so as to know what strategies would
be appropriate in theory, the next question to ask is, “Can effective strategies be adopted in practice?” Implementing strategies typically requires some social apparatus
with which to distribute rewards or mete out punishments. This could be done in a
formal and centralized fashion, as might be the case with payments made from an
agricultural or environmental ministry for the provision of ecosystem services from
forest or farmlands, or in a less formal, decentralized way, as in the various structures traditional communities have developed to assure compliance with community
norms for land and resource use. In the absence of some such governing institutions,
however, as might occur in frontier areas or in regions enmeshed in conflict, it may
be difficult or impossible to motivate more ecologically benign practices.
Closely related to the notion of adequate institutions and governance is a concern with social conditions, especially issues relating to participation and equity. It is
Assessing Intervention Strategies | 231
Box 6.3. The importance of understanding the problem: The case of fire
To many people the appropriate response to fire in ecosystems designated for protection has
been to prevent its occurrence and to extinguish fire once it has been detected. In a number
of places, however, the wisdom of such policies has come into question. Lucas et al. (2008)
Since the 1970s environmental managers in Canaima National Park, Venezuela have attempted to eliminate local savanna burning practices due to a widely held belief that it
causes a gradual reduction in park forest cover. Apart from attempting to teach the native Pemon new fire burning techniques, for decades resource managers have tried to
introduce new farming alternatives in order to reduce the reliance in traditional burning
practices, but managers have had little success. A recent study, however, has shown that
far from being a destructive activity, Pemon savanna burning is in fact an important tool to
control large destructive fires. The failure of resource managers to understand the purpose
of Pemon fire burning practices and value local environmental knowledge has been a key
source of conflict. These failures are responsible for contributing to important alterations
in local fire burning regimes which increase rather than reduce the occurrence of large
scale destructive fires in the park.
Similar histories of fire suppression followed by a renewed appreciation for traditional fire management policies have been observed in many other areas.
possible that institutions exist through which local people can be compelled against
their will to protect more of the ecosystems around them than they would otherwise
choose to. If the focus is on human well-being and its relationship to ecosystem services, however, steps to preserve the latter should not unjustly undercut the former.
As the adverb “unjustly” suggests, determining when this might occur often involves
the exercise of subjective judgment. Preserving ecosystems that provide benefits to a
community generally requires that some members of the community forgo uses that
would be more privately rewarding but less beneficial to the community as a whole.
It seems reasonable to suppose that people who have been traditionally engaged in
uses that are now to be restricted, or who have few if any alternative means of livelihood, should be compensated. Such choices often require that “legitimate” claims
for compensation be distinguished from others, a process that may require drawing
somewhat arbitrary lines between claimants (Brandon 2002).
If practitioners understand the problem being faced and can work with capable
institutions to implement strategies in an equitable fashion, they should begin to
apply the criteria we will discuss in Section 6.2 to choose among the instrumental
strategies described in Section 6.3. But if they cannot answer the questions posed
in Figure 6.3 affirmatively, they may face difficult choices. For people in some positions it is natural and appropriate to conduct essential research, devise and improve
institutions of governance, and work to enhance participation and equity. For those
whose expertise and capabilities lie in these areas, the enhancement of ecosystem service benefits may only be a fortuitous by-product of their pursuit of the primary goal
of research or governance. For those whose fundamental interest lies in preserving
232 | A Manual for Assessment Practitioners
and enhancing ecosystem services, however, the identification of institutional impediments to achieving their goal may constitute a major roadblock.
The Bohensky and Lynam paradigm of impact scope, awareness, and power
may be useful in charting a course. Practitioners may have become aware that problems within a certain scope of impact are detracting from human well-being, but
they must then ensure that they, or a viable partner, have the power required to
bring about change. Each of the “No” branches in Figure 6.3 directs the reader to a
response to overcome the problem: “Do research.” “Build capacity.” “Reform society.” These are not easy tasks. This chapter cannot address the myriad possibilities
and challenges involved in these deliberations, as the focus is more on instrumental
strategies and the consideration of trade-offs in their adoption. Practitioners should
carefully consider their own objectives, capabilities, and demands before seeking to
address ecosystem service loss through institutional interventions. A fundamental
question arises in assessing the efficacy of any response: “Is this where we should
be putting our time, effort, and money, or would those resources be more effective
6.1.3 How to identify binding constraints
In developing guidance on intellectual, institutional, and societal prerequisites for
various intervention strategies, it is useful to think about constraints and control
variables. The latter are quantities that can be changed continuously over some
range in pursuit of better outcomes. Governments could, for example, vary the payment made to local landowners for maintaining landscapes that provide ecosystem
services. Constraints are limitations on the values the control variables can take.
Landowners cannot receive less compensation for providing ecosystem services than
they would receive by converting the land to alternative use, for example, as they
would then pursue the alternative.
Although constraints are rarely so clear-cut as this—even in the example given,
landowners could be compelled to preserve ecosystems even if they might otherwise
choose not to—practitioners face a pragmatic choice of what instruments to treat as
control variables and which considerations to treat as binding constraints. In other
words, they must decide what factors they should treat as constraints on their range
of action and which they may vary in trying to achieve their ends.
It is often possible to anticipate potential hurdles and to assess whether these
are binding constraints that could potentially jeopardize the response or instead
could be avoided by judicious interventions that ensure the success of the chosen
approach. Table 6.1, based on the MA, suggests that such an analysis will need to
assess binding constraints across four domains: political, institutional, economic,
and social.
Table 6.1 argues that decision makers need to be realistic about the extent to
which their interventions can fit within the socioeconomic, political, and cultural
parameters of the society in which they are being proposed. And they should not
attempt to implement strategies that are not congruent with the inflexible circumstances of the intervention context.
On the other hand, where conditions are conducive, changing some of these
circumstances might be necessary as a precondition for the implementation of a
chosen response. In this sense, the constraint is not binding, but identification of the
Assessing Intervention Strategies | 233
Table 6.1. Binding constraints—considerations that might rule out a potential response
Can all potential stakeholders be
Not likely to be a binding
constraint, unless neglected
stakeholders mobilize political
Is the political context supportive?
If not, could be a binding
Can the political context be changed? If not, a binding constraint
Is there adequate capacity for
governance at an appropriate
If not, can adequate capacity for
governance be created?
If not, could be a binding
Is the outcome cost effective?
Could be a binding constraint if
funds are limited
Could be a binding constraint if
there are numerous competing
demands on the resource
Could be a binding constraint,
if this is a high priority or
if those disadvantaged by
the response can effectively
oppose it
Not likely to be a binding
constraint unless consensus is
an explicit objective
Are there secure and well-defined
property rights?
Is the outcome equitable, in a
distributional/material sense?
Does the outcome violate the cultural
norms of particular groups?
If not, a binding constraint
Source: MA 2005a.
constraint allows the decision maker to create enabling conditions for the chosen
response to work more effectively. Box 6.4 presents an example in which an economic constraint that might otherwise have been binding was eased by combining
economic and social objectives.
6.2 How to resolve trade-offs
Section’s take-home messages
• Most intervention strategies involve winners and losers. Practitioners have to make
judgments that balance competing interests when evaluating intervention options.
• Multi-criteria analysis allows different criteria to be considered in making a judgment about
the desirability of an action, and it is also useful when trying to incorporate the views of
diverse actors on an intervention strategy.
234 | A Manual for Assessment Practitioners
Box 6.4. Working for water
As is the case in many areas of developing countries, parts of South Africa suffer from a shortage of clean water. Such shortages are in many instances exacerbated by the presence of
alien species. Nonnative trees sink deep roots into the soil, extracting limited supplies of
The South African Subglobal Assessment reports on efforts to remove exotic species and
restore native ecosystems and their associated hydrology under South Africa’s “Working for
Water” program. The process of removing nonnative species is extremely labor intensive and
consequently very expensive. This expense might have been a binding constraint preventing
implementation of the Working for Water program and thus the restoration of native species
and the increased availability of water. But South Africa also faces serious social problems
arising from inadequate employment opportunities, particularly for traditionally disadvantaged
groups. By hiring such people to clear land of exotic vegetation, South Africa was able to restore ecosystem services and enhance job opportunities.
While “win-win” outcomes are certainly desirable, generalizing from such examples
should be done only with care. Two features of the Working for Water program may distinguish
it from some other public works efforts. First, the benefits arising from clearing land of nonnative species are largely public goods: they accrue to people other than those who own the land
being cleared. Second, there were severe problems of unemployment and underemployment
in the regions where the Working for Water program was initiated. It may have made good
economic and social sense in South Africa to pay more to employ workers to clear land than
the market value of the land once it was cleared. Before emulating such a program, however,
practitioners in other areas should ensure that the activities they contemplate serve similar
public purposes.
Source: SAfMA 2004.
• Cost-benefit analysis reduces different outcomes to a common unit of measurement, but
it is controversial because of several problematic methodological assumptions. Despite
this, it remains a pragmatic and popular technique that provides helpful information for
decision making.
• If a full cost-benefit analysis is not possible, cost-effectiveness analysis can be used to
assess whether a planned intervention strategy can be achieved at a reasonable and
acceptable cost.
The assessment of most actions that might be contemplated to address the degradation of ecosystem services will entail an analysis of trade-offs. Almost any public
policy choice creates winners and losers (see Box 6.5). To improve one measure of
well-being, there typically must be some deterioration in another measure. To make
one person better off, typically some action is taken that will make another person
worse off. How should society make such choices? A variety of solutions have been
Various criteria might be considered in making choices about the environment:
the welfare of the people who favor a change from the status quo, the welfare of
those who oppose it, and in some formulations the welfare of other organisms or
Box 6.5. Visualizing trade-offs
The subglobal assessments conducted by the Alternatives to Slash-and-Burn (ASB) program
have adopted a useful way of visualizing trade-offs. They assemble matrices whose cells reflect the “payoffs,” in physical or monetary terms, to different stakeholders from different policies. The following table provides an example from the forest margins of Sumatra.
Beneficiary or interest
Global community
Returns to
Returns to
rating (0–1)
land ($/ha)
Natural forest
Land Use
based forest
Oil palm
Upland rice/
bush fallow
degrading to
Source: ASB 2003.
Not surprisingly, the ecosystem services that provide the broadest global benefits—carbon sequestration and species diversity—yield the lowest payoff to the national governments
and local people whose actions determine the state of affected ecosystems.
Such a representation of trade-offs can serve a couple of purposes. The first is simply that
it makes clear, and in some instances stark, that there are, in fact, trade-offs to be made. Second, such representations may be the basis of “what if” exercises. For example, if Sumatran
forest authorities could earn €25 per ton of carbon sequestered on the land they control, their
incentives and choices would likely be very different.
It is important to be careful in doing such exercises, however. The ASB project suggests
that doubling the return of nontimber forest products might actually reduce the area of forest
maintained. Why do they get this counterintuitive result? Because local people might invest the
proceeds of such sales in clearing more land.
Source: ASB 2003.
236 | A Manual for Assessment Practitioners
entities affected by the proposed change. Once practitioners accept that there is in
fact a trade-off to be resolved among these different criteria, they have implicitly
acknowledged that an analysis weighing multiple criteria is required. The question,
then, is how to do it.
The simplest approach is to establish a hierarchy among criteria. One way to resolve confrontations between competing goals is to decide that one takes precedence
over another. This leads to problems, however, if the starting position is that certain
considerations are always more important than others. What can be done when
facing competing imperatives? In practice, what is typically done is not to privilege
any particular interest with absolute priority but rather to assign weights to different considerations. This section considers three alternatives that seek to balance
competing interests: multi-criteria analysis (MCA), cost-benefit analysis (CBA), and
cost-effectiveness analysis (CEA).
6.2.1 Multi-criteria analysis
Multi-criteria analysis refers to a procedure for aggregating different criteria into
a single index to be used in making a complex decision. The technique attempts to
integrate different elements of a decision, recognizing that these might not always
be easily comparable. One way of doing this is through a process that either ranks
or rates the different outcomes according to the desired criteria and then compares
them using a formula. This formula could take any of several forms. For instance, it
could be a simple checklist that would list several desired features. The “formula”
in this case is “assign one point for each criterion met, and add up the points.” Such
a procedure is unlikely to prove robust, however. An option may fail to meet one
criterion, while exceeding others in more-important measures.
MCA is a useful technique for aggregating the views of different actors on a particular issue. It lends itself particularly well to team-based decision making, where
the preferences of different team members can all be integrated into the final decision without necessarily privileging the opinion of one person. This avoids the risk
of more “dictatorial” techniques, in which those conducting the assessment impose
their own sense of what is important in determining the well-being of the people
and ecosystems being evaluated. In this sense, MCA encourages more participatory
decision making.
In sophisticated uses of multicriteria analysis, it is common to define aggregate
measures as weighted sums of individual numerical measures (see the discussion of
the Human Wellbeing Index in Chapter 4). However, multicriteria analyses in which
different attributes are weighted and summed to form a total are, at best, approximations of more accurate measures of performance that would be based on more
complicated, nonlinear multivariate functions. This is not to say that simpler MCA
procedures are ill advised or wrong; rather, only that any such procedure is only as
reliable as the assumptions that underlie it.
6.2.2 Cost-benefit analysis
Cost-benefit analysis is a specific form of trade-off analysis that reduces the different outcomes to a common unit of measurement. Typically, certain quantities
Assessing Intervention Strategies | 237
(“benefits”) are multiplied by positive weights (“prices”), while certain other quantities (“costs”) are multiplied by negative weights (also “prices”), and then all of
these weighted quantities are summed to derive measures of the net benefits of the
project or program under consideration. If net benefits are less than zero, they are
“net costs.”
CBA differs from many other forms of trade-off analysis in that the weights
placed on quantities in the analysis are inferred from observed data rather than
being assigned based on moral principles or a priori judgments. However, this “objective” attribute may be both a virtue and a defect of the procedure. It implicitly
assumes that additional money income (or loss) is of equal moral worth regardless
of the initial wealth of the recipient. This problematic assumption may be relaxed,
but only at the cost of introducing other complications.
In the actual use of CBA, several issues arise; all are controversial:
• What can be done to infer the prices of commodities that are not traded in
markets? This is the issue of nonmarket valuation and has been the subject of
many articles and texts (see, e.g., Freeman 2002; see also Box 6.6).
• How should nonmarginal changes be measured? In economic theory, prices are
related to marginal well-being: the improvement in welfare realized in response
to a small change in the provision of a good or service. Cost-benefit analyses can
be done in consideration of large changes (in fact, it only makes sense to do a
CBA in consideration of nontrivial changes), but it typically involves the use of
advanced statistical and/or computational techniques.
• How can measurements be aggregated across people? The economic theory
underlying CBA describes how one person’s welfare may change when
circumstances change, but there is no good way to compare one person’s welfare
with another’s. Conventional cost-benefit analysts typically compare monetary
estimates of what different people would pay to adopt a certain policy.
However, this procedure relies on the implicit assumption that money in the
hands of one person is of equal moral worth to an extra unit of currency in the
hands of any other. This assumption is deeply problematic.
Another very controversial aspect of cost-benefit analysis concerns discounting. This
is the practice of assigning a lower value to costs or benefits that would accrue at
some point in the future than would be assigned to the same costs or benefits that
would accrue immediately. The argument for adopting some form of discounting
relies on three elements:
• People have an intrinsic preference for “sooner” rather than “later.”
• Capital is productive. Benefits forgone today in order to increase production
next year will lead to a more than one-for-one increase in future production
over current consumption sacrificed.
• People grow wealthier over time. If current investment pays off in increased
future consumption, that consumption will be less valuable to its wealthier
(future) recipients than the same quantity of consumption today.
For these reasons, it is generally considered appropriate to discount costs and
benefits at a constant exponential rate when evaluating projects and programs.
238 | A Manual for Assessment Practitioners
Box 6.6. Measurement of nonmarket values
Cost-benefit analyses involving environmental improvements or ecosystem services are very
difficult because damages to the environment and ecosystem services typically do not have
market prices that can be used to measure the value to society. This is not necessarily because
these things are not valuable, but rather because they are public goods that accrue to many
people. Economists have developed a number of techniques of inferring values for goods that
are not traded in markets.
Revealed preference methods rely on observations of behavior in related markets. They
• Hedonic pricing, which involves making inferences concerning the value of component
attributes of a market good from the price of the good. For example, the price of a house
reflects ecosystem services from which it benefits. It often requires considerable subtlety
to disentangle the effects of different attributes, but sophisticated statistical techniques
have been developed to address these issues.
• Travel cost methods, in which peoples’ desire to travel to recreational or scenic sites is
assumed to be related to ecological attributes of the sites. As admission prices are often
not charged, the willingness to pay for the attributes of such sites may be inferred from the
“price” people pay, in time and travel expense, to visit them.
• Production function approaches, which involve treating ecosystem services as unpriced
inputs into the production of agricultural or manufactured products. The “price” of the
ecological input can then be imputed from its contribution to the production of marketed
commodities. Again, subtle statistical treatments are often required to disentangle the effects of inputs.
Stated preference methods have been developed to estimate the value of things that are enjoyed even when they are not consumed or used. People may care about the preservation
of wild species such as mountain gorillas or giant pandas, for example, even if they never
expect to buy products made from these animals, visit their habitat, or even see them in a
zoo. Under such circumstances, since there is no market transaction to which people’s value
of their existence may be related, the only way to estimate the worth people place on them is
to ask questions on surveys. Such “contingent valuation” or “conjoint analysis” exercises are
extremely controversial. Their expense and credibility should be considered carefully before
they are undertaken.
Conventional wisdom has it that appropriate discount rates range from between
approximately 3% and 10% per annum, with the lower rates more appropriately
applied to wealthier, more stable economies and rates toward the higher end of the
scale applied in the developing world.
With its philosophical foundations in considerable doubt, CBA is justified
largely on the pragmatic grounds that it is something that analysts can do, even if it
remains less clear why its results should be considered (Posner 2006). It is often best
treated as one (among many) inputs into decision making rather than the sole basis
for making policy choices. CBA can be a powerful and useful tool, and it certainly
has much to recommend it to the analysis of past and prospective decisions. But
there are some subtle points to be appreciated, and practitioners wishing to employ
Assessing Intervention Strategies | 239
cost-benefit analysis in the assessment of responses should bear certain caveats in
mind (see Box 6.7).
6.2.3 Cost-effectiveness analysis
The objective of cost-benefit analysis is to total all the costs and all the benefits of a
proposed action. If total benefits exceed total costs, the action should be carried out.
In many instances, however, it is far easier to estimate the costs of an action than to
quantify its benefits. Consider, for example, choices made to preserve endangered
species. The cost of land purchased as a wildlife reserve or income forgone from
prohibiting incompatible activities is often easily calculated. But what is the monetary value of the benefits of species preservation?
When the answers to such questions are especially difficult, economists often
settle for the less demanding alternative of performing a cost-effectiveness analysis.
This asks if a specific objective (preserving an endangered species, for example) is
accomplished at the lowest possible cost. The CEA may be sufficient to establish
that the costs of securing certain unquantifiable benefits are “reasonable.” And, of
course, conducting a CEA can help determine the least expensive way of achieving
Box 6.7. Common errors in economic analyses
While many of the ways in which programs and projects may be assessed are straightforward
and make common sense, common errors should be avoided:
• Marginal versus total values. Economic value is determined by how much an additional
amount of a thing is worth, not how much the thing is worth in total. If an ecosystem service is to be reduced but not eliminated, the loss to be estimated is the benefits forgone
as a consequence of the reduction.
• Value-added. The argument is often made that developing countries should capitalize on
their natural resources by engaging in “value-added processing” at home rather than simply being suppliers of raw materials. This confuses the sources of value. Value arises from
scarcity; if the resources a nation controls are not scarce, and hence economically valuable, the country cannot generate real profits by investing in processing facilities unless it
is particularly favored in processing. In many instances “value added” is simply recouping
costs incurred to engage in processing.
• Substitutes. If there are alternative ways to generate the goods or services of natural
ecosystems, the value of such goods and services cannot be greater than the cost of the
• Replacement costs. It follows from the above observation that if there are cheaper ways of
producing a good or service than replacing the system that currently provides it, the cost
of replacement will overstate the value of the good or service.
• Double-counting. There are often many ways of estimating economic values (see Box 6.6).
Calculating values by different methods is sometimes useful to check on against the other,
but it is important not to count the same value twice.
• Alternative metrics. While embodied energy, ecological footprints, and other physical
measures may be useful for some purposes, they generally cannot be used in economic
240 | A Manual for Assessment Practitioners
social goals (such as when decision makers are faced by a choice of alternative
6.3 How to confirm that the right strategy was chosen
Section’s take-home messages
• An intervention strategy can be judged to be effective if it achieves its ecological
objectives without unduly compromising other ecological or social objectives.
• A number of instrumental responses have been used, and each is associated with advantages as well as potential pitfalls.
• Community management works well for some local resources and may be either formal or
informal; however, it is not necessarily an equitable or egalitarian system, and it may not
always be easy to introduce through external intervention.
• Command-and-control regulation may be appropriate at larger scales, but it relies on effective enforcement and implementation.
• Market-based incentives include both tax- or fee-based systems and quota or cap-andtrade systems. These need to be carefully planned and implemented in order to ensure
• Payment for ecosystem services programs are innovative theoretical ways to reward
providers of valuable ecological services, but there are few working examples that truly
involve direct and contingent payment systems in practice.
• Sustainable use strategies are innovative ways to ensure that conservation remains
compatible with economic uses, but they require well-designed and credible certification
programs to be effective.
To evaluate particular response options, we must address three questions: First, by
what standard or standards are actions to be judged? Second, what types of responses ought to be considered? Third, how effective have such responses been or
might they be? This section discusses these questions in order.
6.3.1 How to define “effectiveness”
We adopt the Millennium Ecosystem Assessment’s definition of “effectiveness.” Defining “responses” as “human actions, including policies, strategies, and interventions, to address specific issues, needs, opportunities, or problems,” the MA defines
effectiveness as follows:
A response is considered to be effective when its assessment indicates that it has enhanced the particular ecosystem services (or, in the case of biodiversity, its conservation
and sustainable use) and contributed to human well-being without significant harm to
other ecosystem services or harmful impacts to other groups of people. A response is
considered promising either if it does not have a long track record to assess but appears
likely to succeed or if there are known means of modifying the response so that it can
become effective. A response is considered problematic if its historical use indicates
either that it has not met the goals related to service enhancement (or conservation and
Assessing Intervention Strategies | 241
sustainable use of biodiversity) or that it has caused significant harm to other ecosystem
services [MA 2005b:123; emphases added].
We adopt the MA’s view that effectiveness is generally to be measured against the
goal of “enhanc[ing] . . . particular ecosystem services.” We also note another element of the MA definition: an approach is effective if it achieves its ecological objective without unduly compromising other ecological or social objectives. This latter
criterion raises the issue of trade-offs between the achievement of desirable objectives, a topic already discussed in some detail.
6.3.2 How to identify response options
To identify the most effective responses to ecological degradation, it is important
to understand the range of possible options. This chapter focuses on “instrumental
responses”: actions taken specifically and explicitly to address deterioration in or
threats to ecosystem services. As noted earlier, a variety of other responses might
be undertaken to create the preconditions for implementation of instrumental responses. The efficacy of such enabling responses might be measured by the extent to
which they permit the successful implementation of instrumental responses.
This section briefly considers five broad categories of programs to address the
degradation of ecosystems services and biodiversity: community management, command-and-control regulation, market-based incentives (MBIs), payments for ecosystem services (PES), and sustainable use.
Community management
In some cases, public goods such as ecosystem services can be better managed if the
“public” who benefit from their provision manage the services cooperatively. Communal ownership often makes sense when critical components of an ecosystem are
not easily confined to a particular area, as in the case of many animals, or when it
is important to maintain large contiguous areas in order for services are to be maintained, as is often the case with forests, grasslands, or wetlands.
Successful common property or community management regimes tend to be localized; it may either be formal and recognized by the state or operate as informal,
traditional systems based on the (benign) neglect of the state. Access is allowed for
all members of the community, and nonmembers are excluded. Under such circumstances, an effective response may be for communities to adopt rules and norms
for managing the ecosystems that provide them with services. Pioneering work by
Elinor Ostrom (1990; see also Baland and Platteau 1996) provides guidance as to
the determinants of effective common property management responses.
These conditions arise largely from a consideration of repeated interactions between community members. One person’s maintenance of community infrastructure, or forbearance from the depletion of community resources, is more likely to
be rewarded by another’s reciprocation if both parties anticipate interacting over
many years. Such understandings are likely to be upset if new residents arrive in traditional communities, and this has been a common problem. It is also increasingly
acknowledged that communities are not simple homogeneous groups that work harmoniously to promote group objectives. What has emerged in the contemporary
242 | A Manual for Assessment Practitioners
literature is a more subtle and nuanced understanding of communities as complex,
dynamic, and characterized by internal differences and processes (Leach et al. 1999,
Agrawal and Gibson 1999). Box 6.8 provides an example of how the decline of
common property regimes can result in the degradation of ecological resources,
while the restoration of communal rights may revitalize affected ecosystems.
Among the policy options open to conservation practitioners and their funders
is to “encourage more effective common property management.” It is important to
ask whether outside actors can induce local people to do something that they have
not yet found in their interest to do for themselves. Responses intended to promote
more effective community property management can only be expected to improve
biological conditions if communities are unaware of the effects of their actions or if
individual decision makers do not take account of the reciprocal effects they have
on each other. A community that has already solved its resource management problems to its own satisfaction will not be affected by appeals to work together if they
are already working together. If a concern with global ecosystem services such as
carbon sequestration or biodiversity protection motivates the response, the community can be expected to react to such a concern only if their incentives to do so are
augmented by outside rewards.
Box 6.8. Van Panchayats in Kumaon, India
Communities in the Central Himalayas of northern India had been managing local forests for
centuries when the British took control of the country. The forest ecosystems managed by villagers provided a variety of ecosystem services: water flow regulation and erosion protection,
fodder for animals, fuel for cooking and heating, and construction materials. Elaborate systems
had evolved over time for the maintenance and management of such forests. The rights of
local villagers to forest products were closely regulated, and outsiders were prevented from
unauthorized access to forests.
During the nineteenth and twentieth centuries, demand for wood increased, and new forest regulations were adopted to promote planting and harvesting pine rather than oak species.
While the native oak provided the suite of ecosystem services just mentioned, pine was useful for timber and resins, but not for fodder or erosion control. Moreover, local villagers were
generally precluded from benefiting from the sale of pine timber. Consequently, local people
exercised less care in the management of forests, and the forests deteriorated. Moreover, new
regulations made the mistake of assuring “all bona fide inhabitants” of a region access to forests for traditional uses such as grazing. It was then impossible for inhabitants of one village
to exclude those of another from exploiting common resources or to punish those found to be
overusing resources or failing to contribute to their maintenance.
Forest degradation was partially reversed with the adoption of van panchayats (local management committees) in the early twentieth century. Under these institutions, local people were
again given rights to the proceeds of their forests and a greater hand in their management.
Equally importantly, they were again allowed to exclude unauthorized users from exploiting
their resources. Some of these institutions remain robust today, almost 80 years after their
Source: Somanathan 1991.
Assessing Intervention Strategies | 243
Command-and-control regulation
When public concern (by which we mean to encompass local, regional, national,
and global concerns) with the management of ecological resources reaches a critical
level, public action is taken. The enactment of statutes and regulations as a response
to ecological degradation is not necessarily a distinct approach from community
management but rather an extension of them using the mechanisms available to
larger polities. Larger communities have less of the direct person-to-person interaction on which traditional community management regimes depend for monitoring
individuals’ performance and dealing with those who violate community norms.
Larger communities do, however, have more organized regulatory apparatus and
police power to ensure compliance.
Hence, large modern communities have developed a variety of regulatory means
by which to restrict ecological degradation. The traditional ways to do this have
relied on the police power of the state to enforce command-and-control regulation.
The state commands that individuals and firms observe certain controls. There need
be nothing authoritarian or undemocratic about such measures. The majority may
recognize that it is in their common interest to prevent the degradation of their
environment, and they may introduce rules that enforce appropriate restrictions on
individual actions.
The form of state control most germane to the protection of ecosystem services
and biodiversity is the protected area. Establishing protected areas can be an effective response to the degradation of natural ecosystems. National parks and other
protected areas are, by and large, respected and maintained in the world’s wealthier
nations. In poorer countries, however, protected areas are sometimes derided as
“paper parks,” set aside in theory only. The success of prohibitions on the exploitation of protected areas will depend on a variety of factors. The most obvious is the
financial and political resources expended on preventing intrusion. Perhaps equally
important, however, is the legitimacy with which protected areas are regarded by
local populations. For both moral and pragmatic reasons, local people should be
compensated for their loss of access to protected areas.
Market-based incentives
There may be more effective ways to control ecological degradation than by prohibiting certain actions. It is not always necessary to require everyone to stop taking
actions that degrade ecosystems. If some can stop at a lower cost than can others, it
may make sense for them to bear a greater share of the cost of forgoing degrading
uses. Of course, the lower-cost entity will only assume a greater share of the cost if
it is compensated for doing so.
This is where market-based incentives are useful. Given that a certain amount
of ecological degradation must be borne in exchange for social benefits, there are
several ways to achieve the efficient allocation of degradation across actors. They
involve setting a “price” for degradation and allowing actors to choose how much
degradation to “buy” at that price. The price of degradation could be set with a tax
or a fee. Or a price will emerge if regulators place a limit on the total amount of degradation permitted and allow actors to trade among themselves the right to engage
in it (a cap-and-trade system).
244 | A Manual for Assessment Practitioners
There are two chief differences between the tax or fee-based systems and the
quota or cap-and-trade approach. The first concerns the distribution of the burden.
With taxes and fees, those who degrade the environment pay the state. With tradable permits, one polluter purchases from another the right to emit. The state can,
however, collect revenues in a tradable permit program by auctioning permits to
pollute rather than by giving them away for free.
The second key difference between taxes and tradable permits is that the effects
of the latter are more certain. In such a program, authorities set a fixed quantity of
allowable degradation, and they allow private parties to trade among themselves
to determine how much degradation each causes. With taxes, there is no predetermined limit on how much degradation can occur; this will be determined by the
economic circumstances of those who benefit from the degradation.
The most common application of MBIs has been to industrial pollution; the
two most widely cited tradable permit programs are the U.S. sulfur dioxide trading program and the European Union’s greenhouse gas Emissions Trading Scheme.
Industrial emissions have important consequences for ecosystem services—as demonstrated perhaps most prominently in the case of greenhouse gases—but often for
ecosystem services the source of concern is the depletion of particular biological
resources or the conversion of habitats providing the services. There have been experiments with market-based incentives in these contexts as well. Various nations
and jurisdictions have experimented with tradable development rights—the right to
convert land in one area in exchange for an obligation to protect it in another—and
the Business Biodiversity Offset Program (BBOP 2008) is an attempt to establish a
market in biodiversity protection. In Brazil, legislation has been in place for many
years requiring that a certain fraction of original forest be maintained in many areas; in recent years, steps have been taken to initiate trades in forest preservation
(see Box 6.9).
But a variety of factors must be considered in evaluating the efficacy of such
MBIs in general, and these considerations are often particularly important in the
context of ecosystem services:
• Market-based incentives are only as reliable as they are credible. Commitments
to preserve ecosystems in one area in exchange for valuable considerations
elsewhere must be credible. They must be monitored and enforced. Nations,
regions, or individuals who cannot provide evidence of compliance will either be
excluded from participation or will undercut the system.
• Market-based incentives must deal in a “common currency.” Desirable economic efficiency effects of MBIs are more likely to be realized if there is a larger
market over which compliance and, in the case of tradable permits, trading can
occur. On the other hand, however, wider possibilities for trade increase the
likelihood that such trade will exchange “apples for oranges.” Trading ratios for
disparate assets must be established if MBIs are to be used over broad areas.
• Market-based incentives must guard against “leakage.” One danger of MBIs is
that they will just displace degrading activities to other areas. Suppose that a
system of tradable permits is established by which landowners in a region are
prevented from deforesting more than a certain fraction of the land they own.
This will reduce the supply of forestland relative to the demand for timber and
Assessing Intervention Strategies | 245
Box 6.9. Tradable forest rights in Brazil
In many parts of Brazil, landowners are required to maintain 20% or more of their holdings in
native forest. The expense of doing so may vary dramatically between different areas. At the
same time, however, the ecological consequences of forest preservation may vary greatly between the areas of forest preserved or restored.
Chomitz et al. considered these issues in a study of the potential for tradable forest rights
in Brazil. There is a trade-off in expanding the area over which trading in forest rights is allowed
to occur. On one hand, the larger the area, the greater are the expected cost savings: it is more
likely that someone who can preserve forests at a very low opportunity cost will accept the
obligation to do so. On the other hand, the larger the area over which trades are allowed, the
more likely it is that disproportionate areas of forest will be traded.
In the state of Minas Gerais, the costs of complying with forest-maintenance regulations
could be reduced substantially if trades were allowed to occur over a wider region. Surprisingly, the study’s authors also conclude that the ecological benefits of a wider trading program
could be greater than one in which conservation or restoration must occur on-site. The reason
is that forest restoration in already degraded land would likely not be very effective. Moreover, if
some of the cost savings under the trading program could be directed to particularly sensitive
areas, the results might be still better.
The authors point out that Brazil is unusual, however, in that there already exists a quota
set by law on deforestation. They suggest that it might be more difficult to initiate a tradable
forest rights program entirely from scratch.
Source: Chomitz et al. 2004.
farmland. The price of land and timber will go up. Unless the same restrictions
can be imposed on all forestland in the area, other lands may be exploited,
weakening the impact of the conservation program. MBI programs must be
carefully planned to ensure that alternative, equally sensitive assets are not degraded when one area is designated for protection.
Payments for ecosystem services
The ultimate evolution of market-based incentives regarding the degradation of ecosystem services is a system under which payments are made for their preservation.
Such systems have emerged in recent decades in the policies of both wealthier nations and some developing countries. While the theoretical case for payments for
ecosystem services is impeccable, considerations often arise in practice that limit
their application or at least underscore the preconditions that must be met for their
The most important consideration is that ecosystem services remain, by and
large, public goods. Consequently, the problem of “free ridership” arises: any payment one person makes to protect ecosystem services benefits many people. For this
reason, conventional wisdom has it that public goods must be provided by public action: the state must compel individuals to pay their fair share via taxes, fees,
or similar measures. This is not to say that this process must be undemocratic or
246 | A Manual for Assessment Practitioners
unjust; to the contrary, the nations in which PES systems have caught on tend to
be progressive. The point is only that PES, like most other responses to ecological
degradation, requires that public action be mobilized to address the problem. Moreover, inasmuch as different ecosystem services affect different elements of the global
community, truly effective action to address global ecosystem services requires a
global commitment of resources. This is not necessarily impossible—in fact, proposals have been made recently for International Payments for Ecosystem Services
(OECD 2003, OECD 2004)—but it should be appreciated that such programs require an unusual level of international cooperation.
Ideally, PES programs would be initiated whenever the benefits provided by the
preservation of ecosystem services outweighed the costs of preserving them. In practice, however, PES programs have often arisen when political circumstances align
constituencies behind them. For example, the Conservation Reserve Program (CRP)
conducted by the U. S. Department of Agriculture has set aside over 15 million
hectares of farmland—almost 4% of the nation’s total agricultural land—for habitat preservation. The CRP has proved successful in preventing erosion, reducing
sedimentation, protecting wildlife, and providing other ecosystem services. Political
acceptance of the program is surely enhanced by the fact that it directs monetary
payments to a powerful constituency—U. S. farmers—that had already established
claims on public support via other agricultural price and income support programs.
One potential downside, then, of PES programs is that they may not necessarily
devote resources to the areas of greatest ecological merit but rather to those with a
combination of desirable ecological attributes and political interests.
A related concern is that PES programs face a trade-off between administrative complexity and ecological effectiveness similar to the one described for marketbased incentives. A program that simply announces payments of a certain amount
for land will enroll the least expensive land, which may or may not be valuable for
its ecological attributes. Inasmuch as inexpensive land tends to be located far from
concentrations of population and economic activity, however, it cannot be expected
to provide locally valuable ecosystem services. On the other hand, careful evaluation
of the ecological potential of parcels proposed for conservation may lead to better
selections. But there are two pitfalls here. First, such careful evaluation is expensive
in terms of administrative time and effort. Second, the more complex and arcane a
system for evaluating conservation land becomes, the more vulnerable it may become to manipulation for private gain or political advantage.
While PES programs ought to be initiated when the benefits of their enactment
would justify the expenses paid to property owners, in practice a somewhat different calculation is often made. For example, Costa Rica’s PES program holds out the
possibility of generating reimbursement for carbon sequestered in forested areas.
However, operating funds for the program are generated by a tax on fuels. So although there is a sense in which the conservation investments made under the PES
program offset carbon emissions from fossil fuel use, there is no direct quid pro quo
under which the beneficiaries of conservation compensate its providers.
This is not at all to say that PES programs are not effective or that they should
not be included among potential options for protecting biodiversity and ecosystems.
The point, rather, is that true international markets that involve direct and contingent payments for ecosystem services are only just beginning to emerge.
Assessing Intervention Strategies | 247
Sustainable use
A sea change in conservation strategy occurred in the last quarter of the twentieth century. For most of history, protected areas have been either maintained by
common consent of the communities in which they are located or established and
enforced by higher levels of government. In 1981, with the publication of the influential World Conservation Strategy: Living Resource Conservation for Sustainable
Development by the International Union for Conservation of Nature, the U.N. Environment Programme, and the World Wide Fund for Nature, the focus began to
shift (IUCN et al. 1991). A new emphasis on use, rather than simply conservation,
of resources developed. “Use” was to be nonconsumptive, or at least sustainable,
but under the new policy “natural areas” would be seen as productive economic assets rather than as reserves deliberately excluded from economic use.
The adjective “sustainable” implies constraints on the other uses that could be
made of natural areas in order to assure a livelihood for their human inhabitants.
Extensive deforestation and intensive large-scale agriculture, to say nothing of heavy
industry, are obviously incompatible with “sustainable” use. If it were felt necessary
to promote sustainable use as a conservation policy, it must have been because the
alternative was perceived to be the devotion of previously unspoiled habitats to activities that were not consistent with their preservation.
Of course, someone must also pay the costs of monitoring required to ensure
that certified products are, in fact, produced sustainably. This observation begs a
couple of questions similar to those considered earlier. First, if consumers are willing
to pay the extra cost of sustainably produced products, who should be responsible
for providing certification? Obviously, there would be a conflict of interest issue in
having every producer certify its own product, and this would carry over to having
an industry group certify the products of its constituent members. It is not clear
that such conflicts of interest necessarily prevent the operation of industry-funded
certification systems, however. While concern has often been expressed as to “Who
will certify the certifiers?” the value created by certification is only as great as its
credibility. Companies would have little incentive to incur costs for certification that
consumers did not trust.
It should also be remembered environmental certification is different in some
respects from other forms of product certification and may, as a consequence, prove
less effective. Consumers who purchase an electrical appliance whose performance
and safety are certified by Underwriters Laboratories are concerned that they do not
waste their money on a product that does not work or will endanger their family.
People who stay at a hotel certified by Green Seal are concerned that their lodging
choices do not have excessive negative consequences for the environment. These
incentives may be greatly dissipated by the fact that the impacts of individual decisions are diffused across so large a public. So while certification may be a response
to ecological degradation, it is unlikely to be effective without other accompanying measures. Broader measures, such as direct regulation of forestry, for example,
might in turn obviate the need for certification. In short, certification measures may
be useful and desirable responses in some contexts, but policy makers should consider carefully the interaction of these efforts with other responses, as well as their
relative effectiveness.
248 | A Manual for Assessment Practitioners
6.4 How to link assessment of responses to decision making
Section’s take-home messages
• Assessment of intervention strategies is a critical part of the decision-making and policylearning process.
• Attitudes toward organizational learning shape the overall value of an assessment process
and its impact on improving intervention strategies.
• There may be unexpected benefits from undertaking assessment processes, such as
positive impacts on organizational culture, improved relationships with key stakeholders,
and better communication outcomes, in addition to actual learning about the effectiveness of intervention strategies.
• Considerable methodological innovation has taken place in assessment processes,
and tools now include a sophisticated combination of both quantitative and qualitative
Both in an evaluative context (looking back at previously implemented strategies)
and in an anticipatory context (looking ahead at possible alternatives), the real
value of assessing alternate responses emerges once this gets fed into an iterative and
continuous process of decision making and policy learning. Figure 6.4 illustrates the
different stages at which assessment feeds in to the policy learning cycle.
An ex ante assessment is usually undertaken in order to understand the potential
impact of a response on a desired objective; it is important to use forecasting and
scenario planning approaches to generate the required information to predict possible impacts. This is important in the case of Strategic Environmental Assessments
(now a requirement within the European Economic Area), Environmental Impact
Assessments, Social Impact Assessments, and Regulatory Impact Assessments. Once
a policy has been implemented, an assessment looks back to evaluate the effect of
the policy on desired objectives; this might take place at the project level or at a
more macro-level, within the context of sectoral or national reviews.
In the policy learning cycle, there is an important distinction between monitoring, which can be seen as a continuous process of data collection to track changes in
specified indicators, and evaluation or assessment, which is a more reflective process
that seeks to understand the nature of change. In this sense, assessing responses
needs to go beyond a simple measurement exercise if it is to have any significant
impact on the overall effectiveness of policy processes. However, attitudes to organizational learning vary, depending on the institutional context within which assessment takes place; hence, it is not always necessary that new strategies evolve out of a
careful and empirically informed understanding of past experience (or of systematic
scenario planning). If an organizational culture does not reward reflexive decision
making, interventions may not necessarily be based on a careful assessment of past
Some of the unexpected benefits of assessment processes that have been reported
include process outcomes. The introduction of participatory monitoring and evaluation in some projects, for instance, might create an improved organizational culture
that might benefit program delivery and effectiveness, even if the “data” from such
an exercise are not all that valuable. The process of assessment might help create
Assessing Intervention Strategies | 249
Figure 6.4. The policy learning cycle.
Source: EEA 2005
capacity and nurture a culture of shared learning on policy design, implementation, and evaluation (EEA 2005). Here, processes that allow affected stakeholders
to engage with the assessment, as well as to shape its methods, might enhance the
legitimacy of the process and create more inclusive learning outcomes, with lasting
value beyond the specific context of the assessment.
In a learning organization or policy context, assessment is a critical part of the
process, helping decision makers to assess the effect of planned interventions as well
as to restructure and redesign ongoing programs in an iterative cycle. There is, however, often a considerable disjuncture between these ideals and the actual realities of
assessment processes on the ground. Overworked and under-resourced project staff
frequently perceive assessment as an additional, and somewhat unnecessary, burden
on their time that does not necessarily contribute to their own perceived effectiveness in delivering results. For many organizations, embedding assessment into their
operations continues to pose challenges (see Box 6.10).
At the same time, there is considerable methodological innovation in the processes by which decision makers are seeking to measure and understand change.
New techniques have evolved, including refinements of existing quantitative methods and experimentation with novel qualitative ones. Organizations are seeking to
include a wider range of stakeholders in assessment processes, and there has been
250 | A Manual for Assessment Practitioners
Box 6.10. Response assessment in learning organizations
Assessment is seen as an important learning opportunity by a number of organizations involved in the planning and implementation of policy interventions. It is considered of utmost
importance that learning or information generated through evaluations is fed back for better
program management. For instance, the World Bank explicitly states, under its strategic objectives of conducting evaluation, the need for “building learning opportunities into evaluations.”
Similarly, other organizations stress the need to take the feedback of stakeholders on findings,
disseminate the results of evaluation through publications, and use findings to improve the
performance of the project.
The Joint Inspection Unit (JIU) report on Managing for Results in the United Nations System recognizes the importance of learning, but also the limited extent to which this is actually
happening. It observes that “in general, there is a lack of coherence in the planning, programming, budgeting and evaluation cycle in many organizations, in particular in the final stage
of evaluation, which is not systematically used on the eve of the next budgetary cycle.” This
suggests that, at present, evaluation findings and recommendations are not being used to
their full potential to improve performance. The JIU report suggests that for a results-based
management system to be effective, “evaluation findings (must) be effectively used.” However,
the report does not provide further recommendations on how to improve the effective use of
evaluation findings in policy learning.
Organizations such as NORAD, the United Nations Development Programme, the German
development agency, the Global Environment Facility, and the Canadian International Development Agency specifically assign responsibilities and detail the processes for follow-up on
evaluation findings and their integration into a results-based management system. Guidelines
also specify the need to disseminate the findings of the evaluation among interested parties
as well as making them available for the public. Within the nongovernmental sector, “critical
stories of change”—a method used by ActionAid to document and analyze progress of an
intervention—itself serves as a learning document, as the process involves continuous dialogue with the stakeholders at various levels. The document is also available for general public
access, to allow wider stakeholder engagement with the findings.
Sources: JIU 2004, ActionAid 2006, IEG 2007.
some progress in developing participatory approaches that empower beneficiaries
to determine what matters to their lives and how change should be measured and
documented. Apart from beneficiaries, assessment feeds into organizational interaction with a wider range of stakeholders, including clients, implementation partners, donors, and benefactors. Measuring the impact of programs, and documenting
change, is a useful way to improve communication with these external stakeholders.
It also helps provide greater accountability, allowing organizational performance to
be assessed against the expectations of these other actors.
Ultimately, the context within which strategies are being implemented shapes
the extent to which assessment of interventions feeds into the decision-making cycle.
The implementing agency might be open to a process of reflexive, self-critical learning, in which case the assessment would be seen as part of an evolving cycle of
introspection and adaptive management. On the other hand, the assessment might
also be conducted by a more defensive and closed organization, which is seeking
Assessing Intervention Strategies | 251
evidence to justify its intervention but does not actually learn from experience and
integrate it into better future planning. There are probably a range of organizational
types along the spectrum that these stylized (and polarized) examples represent.
What is important is to recognize that the value of any assessment is partly shaped
by the attitude of the agency that is conducting the exercise and by its willingness to
learn from such a process.
6.5 Conclusion
Section’s take-home messages
• In any assessment, it is fundamentally important to identify the nature of the problem that
is being addressed, the range of strategies that are available, and the key actors who are
able to exert an influence on the issue.
• The value of an assessment process lies is in its ability to progressively improve the quality
of intervention strategies.
This chapter has outlined a series of steps that are useful in conducting an assessment of strategic interventions designed to positively influence ecosystems and human well-being. Such an exercise can be conducted to evaluate the performance of
previous interventions or to plan future options. In both cases, it is fundamentally
important to identify the nature of the problem that is being addressed, the range of
strategies that are available, and the key actors who are able to exert an influence
on the issue. When looking at past interventions, it is useful to consider not just the
options that have been implemented but also other options that might have been
feasible but were rejected. It is important not just to assess whether the intervention
worked—in the sense of delivering desired outcomes—but also whether it was more
(or less) effective than alternatives that might have been pursued.
Similarly, in forward-looking scenarios, a range of possible alternatives need
to be considered, and choices need to be made that address decision makers’ key
strategic objectives. In some situations this is likely to involve accepting trade-offs,
in the sense that certain objectives might be achieved through the intervention while
others might have to be sacrificed. This is the inevitability of policy choices, and
decision makers need to recognize that it is often not possible to achieve progress
on all fronts simultaneously. The criteria for choosing among alternatives need to
reflect social and political values and realities. Different decision makers, operating
in different decision contexts, might have very different rankings among competing
The value of an assessment process lies in its ability to progressively improve
the quality of intervention strategies. In a decision context that values policy learning, an assessment process is likely to yield useful insights into the sorts of strategies
that have worked well and those that have been less effective. This can be used to
communicate outcomes to stakeholders and to plan future interventions in a cycle
of adaptive learning and management. To benefit from this process, decision makers
must attempt to create a culture of reflexive learning in order to maximize the influence of the sorts of assessment procedures that have been outlined in this chapter.
252 | A Manual for Assessment Practitioners
ActionAid. 2006. Accountability, Learning and Planning System (ALPS). Johannesburg: ActionAid
Adams, W., D. Brockington, J. Dyson, and B. Vira. 2003. Managing tragedies: Understanding conflict
over common pool resources. Science 302:1915–16.
Agrawal, A., and C. C. Gibson. 1999. Enchantment and disenchantment: The role of community
in natural resource conservation. World Development 27 (4): 629–49.
ASB (Alternatives to Slash-and-Burn Programme). 2003. Balancing rainforest conservation and
poverty reduction. Policy Brief 5. Nairobi: ASB Programme.
Baland, J.-M., and J.-P. Platteau. 1996. Halting degradation of natural resources: Is there a role for
rural communities? Oxford: Oxford University Press.
BBOP (Business and Biodiversity Offset Programme Secretariat). 2008. A draft consultation paper
for discussion. Available at
Bohensky, Erin, and Timothy Lynam. 2005. Evaluating responses in complex adaptive systems:
Insights on water management from the Southern African Millennium Ecosystem Assessment
(SAfMA). Ecology and Society 10 (1): 11.
Brandon, Katrina. 2002. Perils to parks: The social context of threats. In Parks in peril: People,
politics and protected areas, ed. Katrina Brandon, Kent Redford, and Steven E. Sanderson.
Washington, DC: Island Press.
Chomitz, Kenneth M., Timothy S. Thomas, and Antonio Salazar Brandao. 2004. Creating markets
for habitat conservation when habitats are heterogeneous. Policy Research Working Paper 3249.
Washington, DC: World Bank.
EEA (European Environment Agency). 2005. Effectiveness of urban wastewater treatment policies
in selected countries: an EEA pilot study. Report No 2/2005. Copenhagen: EEA.
IEG (Independent Evaluation Group). 2007. Evaluation methodology. Washington, DC: World
IUCN (International Union for Conservation of Nature), U.N. Environment Programme (UNEP),
and World Wide Fund for Nature (WWF). 1991. World Conservation Strategy. Gland, Switzerland: IUCN, UNEP, and WWF.
JIU (Joint Inspection Unit). 2004. Implementation of results based management in the United Nations organisations: Managing for results in the United Nations system. New York: United
Kapoor, Aditi. 2003. Women, water, and work: The success of the Self-Employed Women’s Association. In World Resources 2002–2004, 198–207. Washington, DC: World Resources Institute.
Leach, M., R. Mearns, and I. Scoones. 1999. Environmental entitlements: Dynamics and institutions in community-based natural resource management. World Development 27 (2): 225–47.
Lucas, Nicolás J., Iokiñe Rodriguez, and Hernán Darío Correa. 2008. To change global change:
Ecosystem transformation and conflict in the 21st century. In Policies for sustainable governance
of global ecosystem services, ed. Janet Ranganathan, Mohan Munasinghe, and Frances Irwin.
Cheltenham, U.K.: Edward Elgar.
MA (Millennium Ecosystem Assessment). 2005a. Ecosystems and human well-being: Policy responses. Washington, DC: Island Press.
MA. 2005b. Ecosystems and human well-being: Synthesis. Washington, DC: Island Press.
Olson, Mancur. 1965. The logic of collective action: Public goods and the theory of groups. Cambridge, MA: Harvard University Press.
Ostrom, E. 1990. Governing the commons: The evolution of institutions for collective action. Cambridge, U.K.: Cambridge University Press.
Panda, Smita Mishra. 2007. Women’s collective action and sustainable water management: Case of
SEWA’s water campaign in Gujarat, India. CGIAR Systemwide Program on Collective Action
Assessing Intervention Strategies | 253
and Property Rights. Washington, DC: Consultative Group on International Agricultural
SAfMA (Southern African Millennium Ecosystem Assessment). 2004. Nature supporting people:
The Southern African Millennium Ecosystem Assessment integrated report. Pretoria: Council
for Scientific and Industrial Research.
Somanathan, E. 1991. Deforestation, property rights and incentives in Central Himalayas. Economic and Political Weekly 26 (4): 37–46.
Action cosmetics, 210
Action overshoot, 210
Action paralysis, 210
Actor-centric exercises, 155
Adaptive Cycle Model, 85, 86
Adaptive management, 97
Advisory groups, 49–51, 56
Aggregation, 118, 236, 237
Alternative metrics, 239
Alternatives to Slash and Burn Program
(ASB), 102–104, 235
Approach, defining of, 53–54
Aral Sea syndrome, 89, 91
Archiving, 55
ASB. See Partnership for the Tropical Forest
Assessment teams, 46–48, 73
Assessments, 7, 26
Assumptions, scenario exercises and,
Audiences, 64, 66–67, 68. See also
Authorizing environments, 170–173
Awareness, 42, 223, 224, 232
Ayllu, 106, 107
Ayni, 106
Bajo Chirripó, Costa Rica, 58–59, 107–108
Balance, 47
Barter economies, 142
Baselines, 80
Basket of services, 131
Bayesian framework, 146
BBOP. See Business Biodiversity Offset
Bias. See Legitimacy
Binding constraints, 232–233
Biodiversity, 123–127, 244
Bioenergy, 6
Bohensky and Lynam paradigm, 223–224,
Boundaries, 10, 43–44, 101–102
Brazil, 244, 245
Brundtland Report, 100
Budgets, 48, 49
Business Biodiversity Offset Program
(BBOP), 244
Business strategy scenarios, 160
Business-as-usual scenarios, 177, 191
Cabecar conical shaped house, 108
California Agroecosystem Assessment,
92, 93
Canadian International Development
Agency, 250
Canaima National Park, 231
Capacity, 47–48
Capacity building, 15, 58–59
Capacity to deliver given service, 131
Cap-and-trade systems, 243, 244
Capital theory approach, 91
Captions, 147–148
Caribbean Seas Assessment, 103, 104–105
Cause-and-effect chains, 22
CBA. See Cost-benefit analysis
CEA. See Cost-effectiveness analysis
Certification programs, 247
Chakana, 107
Champions, 67
Change in community composition, 131
Chile, 41, 49
China, 42
Choice modeling, 141
Clarity, 78
Clean Development Mechanism, 227–228
Climate change, 4
Clusters, pathways, or trajectories concept,
89, 90
Coherent scenarios, 189, 190
Collation of data, 144–145
Command-and-control regulations,
241, 243
Commodity prices, 237
Common currency, 244
256 | Index
of condition and trend assessment results,
formats for, 63–66
goals for, knowing audience and, 61–62
overview of, 18, 28–29
relevance, credibility, legitimacy and,
report content and style and, 62–63
of scenario exercise results, 201–212
strategies for leveraging, 67–68
of uncertainty, 63, 146
Community composition, change in, 131
Community management, 241–242
Comparable scenarios, 189, 190
Complementary scenarios, 189, 190
Complexity, 76, 96–100, 147–148
Composite indices, 126, 134
Compound indexes, 117
Conceptual frameworks
complexity, uncertainty and, 96–98
defining of, 53–54
development of, 74–78
drivers of ecosystem change and, 88–91
examples of use of, 101–110
flows of ecosystem services, stocks or
resources and, 91–93
overview of, 12, 72–74
political complexity and, 98–100
scales in space and time and, 93–96
structure and main elements for MA,
use of, 78–79
Condition and trend assessments
biodiversity and, 123–124
communicating, 142–148
conceptual frameworks and, 87
dealing with multiple scales in, 124–127
ecosystem services and, 127–132
information sources and, 119–123
link of ecosystem services to human wellbeing and, 134–142
metrics, indicators, indices and, 117–119
overview of, 16–17
selection of what to measure in, 116–117
setting scope of, 115–116
Conflict management, 57
Conical houses, 108
Conjoint analysis, 141, 238
Consecutive scenario development
processes, 189
Conservation Reserve Program (CRP), 246
Content, 24–25
Context, 204, 207–208, 211–212, 250–251
Contingent valuation, 141, 238
Control variables, 87
Convening assessment participants, 52–53
Convention on Biological Diversity (CBD),
119, 227
Conversion rates, 133
Cosmetics, 210
Cost of illness valuation, 141
Costa Rica, 58–59, 107–108, 246
Cost-based valuation, 141
Cost-benefit analysis (CBA), 236–239
Cost-effectiveness analysis (CEA), 236,
certification and, 247
communication of results and, 66–67
ecosystem assessments and, 6, 24
market-based incentives and, 244
members of assessment team and, 47
overview of, 33–34, 35
participation and, 37–39
peer review and, 17, 57–58
process rules and, 54
stated preference methods and, 238
synergies and tradeoffs with relevance,
legitimacy and, 37
Credible points of view, 24
CRP. See Conservation Reserve Program
Crusoe example, 75–76
Cultural indicators, 129
Culture, 88
Currencies, 244
Customary landowners, 108
Dahlem Desertification Paradigm (DDP), 84
Data, 55, 96–98, 144–145
Databases, 120–122
DDP. See Dahlem Desertification Paradigm
Decision context, 211
Decision making
improving, 2–8
intervention strategies and, 248–251
linkage of to assessment of responses,
linkage of to ecosystem assessments,
scenario exercises for, 159–160, 166,
summaries for, 64–65
targeting report style to, 63
Index | 257
Decision units, 211
Deductive approaches to scenario
development, 177, 178, 179
Deforestation, 95, 223, 245
Degree of fragmentation, 130
Department for International Development
(DfID), 91–93
Desertification, 84, 95
Desertification Synthesis, 131
Design stage, overview of, 10–16, 28
Development goals, 4–5
Development partnerships, 136–137
DfID. See Department for International
Diagrams, conceptual frameworks and,
75–76, 79, 98, 99
Direct drivers, 87, 88, 89
Disaggregation, 118
Discontinuities, uncertainty and, 97
Discount rates, 140, 237–238
Diseases, 135–136
Dong-Zhi Tableland, China, 42
Double-counting, 239
Drivers, 80, 87–91, 164, 182–185, 194
Drivers Pressures States Impacts Responses
(DPSIR) frameworks, 80, 81–83, 92–94
Drylands, 8, 84, 89
Early product release, 68
Ecological footprint, 131
Economic incentives, 19, 20
Economic optimism scenarios, 157
binding constraints and, 233
common errors in analyses of, 239
cost-benefit analysis, 236–239
cost-effectiveness analysis and, 239–240
as driver of ecosystem change, 88
estimating necessary funding and, 48–49
human well-being indicators and, 133–134
incentives and, 19, 20
multi-criteria analysis and, 236
stocks and flows and, 130
valuation of ecosystem services and,
Ecosystem change, 88–91
Ecosystem integrity, 130–131
Ecosystem services
Alternatives to Slash and Burn Program
and, 235
assessing link to human well-being,
balance of supply and use of, 131–132
conceptual frameworks and, 91–93
development goals and, 4–5
gathering and assessing information on,
human well-being and, 87
indicators for, 128–129
payments for (PES), 241, 245–246
stocks and flows and, 129–130
valuation of, 138–142
Ecosystem Services Review, 48
Ecosystem Wellbeing Index, 134
Ecosystem-level indicators, 126
Ecosystems, scenario exercises and, 194–197
intervention strategies and, 249, 250
Millennium Development Goals and, 135
scenario exercises for, 159, 166, 205–208
EEA. See European Environment Agency
Effectiveness, defining, 240–241
Electronic communications, 65
Emblematic events, 184
Emissions, 6, 244
Emissions Trading Scheme, 244
Empirical relationships, uncertainty and, 97
Enabling responses, 225, 226, 227–228
Energy security, 5, 6
Engagement, review process as, 56–57
Environmental conservation, 4
Environmental Impact Assessments, 248
Ethical discount rate, 140
Ethnic balance, 47
European Environment Agency (EEA), 176
EU-Ruralis project, 196, 204, 205
Evaluation, 15–16, 248
Ex ante assessments, 248
Explicit knowledge, 14–15
Explorative scenarios, 155, 206
Exploratory stage, 9–10, 28, 39–46
Extent of the ecosystem, 130
Fairness. See Legitimacy
Farm forestry, 229
Fee-based incentives, 244
Feedbacks, 225
Fellows programs, 15
Figures, results communication and, 63
Financial turnover, 140
Fire management, 231
Fit of scale, 204, 208, 212
Flexibility, 75
Flows, 129–130
258 | Index
Focal issues, 180–182, 185–186
Food production, 5
Food security, 6
Foreloop, 85
Forests, 78, 242, 244, 245. See also
Formal knowledge, 14–15
Formality, 52
Foundation responses, 225, 226
Fragmentation, degree of, 130
Frameworks. See Conceptual frameworks
Frequentist framework, 145–146
Freshwater, 4
Fuelwood, 132, 229
Funding, estimating necessary, 48–49
Gender balance, 47
Gender equality, 135, 229
Gene-level indicators, 126
GEO. See Global Environment Outlook
Geographic boundaries, 43
Geographical scale, 170, 171
German development agency, 250
Gini coefficient, 133
Global Biodiversity Assessment, 34
Global Environment Facility, 250
Global Environment Outlook (UNEP), 7,
81–82, 83, 194
Global Orchestration scenario, 198
Global sustainable development scenarios,
Global value, 139
Goals, 53, 61–62
during assessments, 54–57
designing structure of, 49–51
incorporation of assessment results and,
19, 20
interaction between technical team
and, 56
overview of, 11–12
Governments as initiating actors, 226–228
Grains, 6
Graphics, 147–148
Green Seal, 247
Greenhouse gases, 6, 244
Groups-in-conflict scenarios, 159
GTAP, 196
Gujarat, India, 229
Hananpacha, 106
Health, ecosystem services and, 4
Hedonic pricing, 141, 238
Heuristic frameworks, 156
Hierarchical systems, 80
Houses, conical, 108
Human capital valuation, 141
Human Development Index (HDI), 134
Human Ecosystem Model, 84–85
Human Poverty Index, 134
Human well-being. See Well-being, human
Hydropower, 5
IAASTD, 7, 73
Ignorance, 152–153
IMAGE, 196
Impact scope, 223, 224, 232
Impact variables, 87
Implementation, overview of, 16–18, 28–29
economic, 19, 20
fuelwood and, 229
incorporation of assessment results and,
19, 20
intervention strategies and, 223
market-based, 241, 243–245
payments for ecosystem services, 241,
Income equality, 133
Incremental approaches to scenario
development, 177
Independence, peer review and, 17
Independent scenario development
processes, 189
India, 229, 242
Indicative indicators, 200
for biodiversity, 126
condition and trend assessments and,
for ecosystem integrity, 130–131
for human well-being, 132–134
for main ecosystem services, 128–129
Indices, 117–119
Indigenous knowledge, 14, 46, 55, 108–110,
Indirect drivers, 88, 89, 95
Inductive approaches to scenario
development, 177
Inflation indexes, 133
Information. See also Knowledge
in condition and trend assessments,
119–123, 127–134
credibility of, 35
Index | 259
identifying sources of, 57
improving through participation, 38–39
legitimacy and, 36
scenario exercises and, 159, 166, 191–
194, 205–208
slogans and, 78
types of, 14–15
using nonpublished/nonpeer-reviewed
sources of, 30–31
Initiating actors, 225–227
Institutions, binding constraints and, 233
Instrumental responses, 225, 226, 227–228
Integrated multiscale assessments, 13, 171
Integrated natural resource management
(iRNM) Approach, 102–104
Integration, 51, 54
Intergovernmental Panel on Climate Change
(IPCC), 7, 34, 60, 81–82
International Assessment of Agricultural
Science and Technology for
Development (IAASTD), 7, 73
International assessments. See Assessments
International Payments for Ecosystem
Services, 246
Internet, 65, 123
Intervention strategies
appropriateness of, 224–232
binding constraints and, 232–233
confirming selection of correct, 240–247
determination of responding parties and,
effectiveness of, 240–241
linking assessment of responses to
decision making and, 248–251
resolving trade-offs and, 233–240
response options and, 241–247
selection of, 221–233
Invasive (alien) species, 234
IPCC. See Intergovernmental Panel on
Climate Change
iRNM Approach. See Integrated natural
resource management approach
Iterative scenario development processes,
Ixacavaa, Costa Rica, 107
Japan sub-global assessment, 105–106
Joint Inspection Unit (JIU), 250
Joint scenario development processes, 188
Kaypacha, 106
Kebun lindung, 78
Key users, 45
Knowledge. See also Information
boundaries between different forms of,
condition and trend assessments and, 123
credibility of, 35
intervention strategies and, 227
slogans and, 78
types of, 14–15
Kumaon, India, 242
Kyoto Protocol, 227–228
LADA. See Land Degradation Assessment of
Laguna Lake Basin, Philippines, 41
Land Degradation Assessment of Drylands
(LADA), 7
Land use changes, 90, 200
Land Use/Cover Change (LUCC)
Programme, 89, 90, 91, 92–93
Language, 25
Leadership, 52
Leakage, 244–245
communication of results and, 66–67
ecosystem assessments and, 8
members of assessment team and, 47
overview of, 33–34, 35–36
participation and, 35–36
process rules and, 54
scenario exercises and, 174
synergies and tradeoffs with credibility,
relevance and, 37
Life cycle, policy, 21–22, 167, 248
of assessment to decision-making process,
of drivers across scales, 95
drivers of ecosystem change and, 88–89
of ecosystem services and human wellbeing, 134–142
of formal and informal knowledge,
between models, 196
of scale in ecosystem assessments, 12–13
of scenarios across geographic scales,
Livelihoods approach, 91
Llankay, 106
Local and traditional knowledge, defined,
Local knowledge, 14, 55
260 | Index
Logics, scenario exercises and, 182–185
LUCC Programme, 89
MA. See Millennium Ecosystem Assessment
Magna Carta, 78
Management of assessment process, 52–59
Managing for Results in the United Nations
System (JIU), 250
Maps, 148
Marginal value, 139, 140, 239
Market value, 141
Market-based incentives (MBI), 241, 243–
MBI. See Payments for ecosystem services
MCA. See Multi-criteria analysis
Media, 63, 66
Meetings, results communication and, 65
Meta-analysis, 146
Metrics, 117–119, 239
Millennium Development Goals, 134,
Millennium Ecosystem Assessment (MA)
comparing scenarios and, 199
conceptual framework of, 80, 86, 87
coupling models and, 196
ecosystem services indicators in, 128–129
effectiveness and, 240–241
human well-being indicators in, 133
intervention strategies and, 224–225
legitimacy and, 36
nonpublished/nonpeer-reviewed sources
in, 30–31
overview of, 2, 3, 7, 11
scenario exercises and, 153, 156, 158–
159, 162, 187–188, 193, 199
structure and main elements of, 80–88
user groups in, 46
workplan schematic, 55
Modeling, 121–123, 191, 193–196
Monitoring, 247, 248
Mont-Fleur scenarios, 206
Montreal Protocol, 227
Morality, 39
Mortality, 135
Multi-criteria analysis (MCA), 236
Multiscale assessments, 13, 171
Munay, 106
National and subnational policies, 19, 20
Natural hazards, 4
Neutrality, 77
New conversation scenarios, 159
Nonintervention scenario, 196
Nonlinearity, 97
Nonmarginal changes, 237
Nonmarket values, 238
Nonnative species, 234
Nonpeer-reviewed information sources,
Non-use value, 139
NORAD, 250
Normative scenarios, 155, 206
Northern Range Assessment, 103, 104–105
Offsetting trends, 200
Oil palm plantations, 223
Omega>Alpha backloop, 85
Organization of assessment process, 46–52
Outreach kits, 67
Overshoot, 210
Pachakuti, 106, 107
Pachamama, 107
Pani samitis, 229
Papua New Guinea (PNG), 103, 108–110
Parallel scenario development processes, 188
Paralysis, 210
Participation, 35–39, 52–53, 173–176
Participatory Rapid Appraisals, 55
Partner institutions, 67
Partnership for the Tropical Forest Margins
(ASB), 102–104, 235
Pathways, 89, 90
Payments for ecosystem services (PES), 241,
Peer review process
condition and trend information and,
120, 123
credibility and, 17, 57–58, 59
as engagement, 56–57
overview of, 17–18, 28–29
Perceptions, uncertainty and, 97
Peru, 55, 66, 99, 106–107
PES. See Payments for ecosystem services
Philippines, 41
Photographs, 148
Planning focus scenarios, 209
PNG. See Papua New Guinea
Policy, intervention strategies and, 249
Policy cycle, 21–22, 167, 248
Political boundaries, 43–44
Political institutions, 233
Pollution, 244
Population change, 88
Index | 261
Portugal sub-global assessment, 68
Potsdam Institute for Climate Change, 89
Poverty, 5, 100, 135
Power, 223, 224, 232
PPP. See Purchasing power parity
Predictive scenarios, defined, 155
PRELUDE project, 176, 187–188, 204, 207
Prescott-Allen, Robert, 134
Prescriptive regulations, 23–24
Presentation, key questions and, 24
Pressure State Response (PSR) framework, 80
Pressures and threats indicators, 126
Probabilistic information, 197–198
Problem-focused scenarios, defined, 155
Process outcomes, 248–249
Process variables, 87
Production function valuation, 141, 238
Project teams, scenario exercises and,
Provisioning services, 126–127, 128, 130
Proxies, 128–129
Proximate causes, 89
PSR framework. See Pressure State Response
Public entities as initiating actors, 226–227
Public goods, 245–246
Public interest scenarios, 160
Purchasing power parity (PPP), 133
Qing-Yang, China, 42
Qualitative scenarios, 153–154, 157, 186,
Quality assurance, 55
Quantitative scenarios, 153–154, 157, 186,
Quechua people, 106–107
Questions, addressing relevant, 23–24
Quotas, 244
r>K (foreloop), 85
Random utility, 141
Recommendations, 23–24
Reference conditions, 80
Reference scenarios, 155, 177, 191
Referencing of information sources,
Reflexive interventionist scenario processes,
Reformed market scenarios, 157
Regional competition scenarios, 157
Regional sustainable development scenarios,
Regression analyses, 145
Regulating services, 128
Regulatory Impact Assessments, 248
communication of results and, 66–67
condition and trend indicators and, 118
ecosystem assessments and, 8
importance of in policy responses, 22–23
members of assessment team and, 47
overview of, 33–37
participation and, 37–39
peer review and, 57–58
synergies and tradeoffs with credibility,
legitimacy and, 37
uncertainty and, 25
Reliability, 244
Reorganization, 85
Replacement costs, 239
Reporting, 26, 62–66
Representation, 52
Reserved words, 96
Resilience, 107
Resilience Alliance, 85
Resource stocks, 91–93
Response indicators, 126
Response options, 195–197
Response variables, 87
Responses component, overview of, 16,
Revealed preference methods, 140, 238
Review process, 56–57. See also Peer review
Risk assessments, 208
Robinson Crusoe example, 75–76
Robustness analysis, 200
Rules, defining of, 54
Ruralis project, 196, 204, 205
SAfMA. See Southern African Millennium
Ecosystem Assessment
SAfMA-Gariep assessment, 177
SAfMA-livelihoods assessment, 177
SAfMA-regional assessment, 177
Sahel syndrome, 89, 91
Salar de Atacama, Chile, 41, 49
Salience. See Relevance
San Pedro de Atacama Subglobal
Assessment, 41, 49
São Paulo sub-global assessment, 123
Satellite-derived images, 148
Satoumi, 105–106
Satoyama, 105–106
262 | Index
biodiversity measures and, 127
conceptual frameworks and, 76
in condition and trend assessments,
credibility, relevance, legitimacy and, 37
defining, 42–43
drivers of ecosystem change and, 88
impact scope and, 223, 224
intervention strategies and, 222–223
linkage of in ecosystem assessments,
12–13, 95
multiple in space and time, 93–96
scenario exercises and, 170, 171, 188–
190, 202–204, 208, 212
Vilcanota framework and, 106–107
Scenario exercises
analysis of, 190–201, 215
approaches to developing, 178–180
assumptions, storylines and, 185–190
comparing, 198–201
for decision support and strategic
planning, 159–160, 166, 208–212
defining scope of, 168–170
development of, 177–190, 214
drivers, logics and, 182–185
ecosystems and human well-being and,
for education and information, 159, 166,
establishing authorizing environment and
project team for, 170–173
examples of types of, 155
focal issue identification and, 180–182
history of use of, 157
managing participation throughout,
in Millennium Ecosystem Assessment
Framework, 158–159
modeling in, 193, 196
outcomes in, 163–164
overview of, 16, 17, 152–155
PRELUDE project and, 176, 187–188
purpose in, 164–168
reasons for use of, 155–156, 159–160
role of, 162
SAfMA and, 177
scale and, 171, 188–190
for scientific exploration and research,
156, 159, 165–166, 203–205
setting up, 161–177, 214
uncertainty and, 154, 197–198
use and communication of, 201–212, 215
use of out of context, 212
Scientific research
condition and trend indicators and, 118
intervention strategies and, 227
scenario exercises for, 156, 159, 165–166,
Scientists, targeting report style to, 63
of condition and trend assessments,
of environmental assessments, 10, 42–44
impact, 223, 224
of scenario exercises, 168–170
Sector policies, 19, 20
Self-Employed Women’s Association
(SEWA), 229
Sensitivity assessments, 118, 208
Service-based integrity indicators, 131
Services. See Ecosystem services
SEWA. See Self-Employed Women’s
Shadow pricing, 141
Slogans, 78
Social conditions, 230–231, 233
Social discount rate, 140
Social Impact Assessments, 248
Social learning, 97
Social preference function, 142
Sociopolitics, 88
Southern African Millennium Ecosystem
Assessment (SAfMA), 13, 50, 54, 102,
132, 177, 234
Spatial data, 122
Species richness, 127
Species-level indicators, 126
SRES scenarios, 184, 185
advisory groups and, 49–50
communication of results and, 61–62
consultations with, 56
identifying and engaging relevant and
diverse, 44–46
SAfMA and, 50
scenario exercises and, 166, 169, 173–176
State variables, 87
Stated preference methods, 238
Statistical databases, 120–121
Statistics, 145–146
STEEP drivers, 164
Steering committees, 11–12
Stocks, 129–130
Index | 263
Story-and-simulation scenarios, 192, 193–194
Storylines, 185–190
Strategic conversation scenarios, 160
Strategic Environmental Assessments, 248
Strategic planning, 157, 159–160, 166,
Strategy evaluations, 208–209
Students, targeting reports to, 63
Stylized facts, 194–195
Substitution, 239
Substitution pricing, 141
Sulfur dioxide trading program, 244
Summaries, 64–65, 144–145
Summaries for decision makers (SDM),
Summaries for policy makers (SPM), 64–65
Supporting indicators, 129
Surprise, 153
Sustainability, 91, 136, 157
Sustainable use, 241, 247
Syndromes concept, 89–91
Synergy, 88–89
Tables, 147–148
Tailored reports, 64
TARGETS project, 188
Tax-based incentives, 244
Technogarden scenario, 198
Technology as driver, 88
Terms of reference (ToR), 51
Thresholds, 97, 98
Time, 35, 116
Time horizon, 169–170
Time scales, 93–96
Tipping points, uncertainty and, 97
ToR. See Terms of reference
Total Economic Value, 139
Total value, 139, 239
Traceability, 120–121
Tradable development rights, 244
Tradable forest rights, 245
Tradable permits, 244
Trade-off matrixes, 143
assessing, 142, 143
cost-benefit analysis and, 236–239
cost-effectiveness analysis and, 239–240
intervention strategies and, 233–240
multi-criteria analysis and, 236
payment for ecosystem services programs
and, 246
scenarios and, 199
Traditional ecological knowledge, 14
Traditional knowledge, 14, 123
Trajectories, 89, 90
Transparency, 35, 39, 52, 56–57, 120
Travel-cost methods, 141, 238
Trends. See also Condition and trend
Trends, defined, 115–116
Tropical Forest Margins Assessment, 102–104
Ukupacha, 106
U.N. Convention to Combat Desertification
(UNCCD), 84
communication of, 63, 146
conceptual frameworks and, 96–98
ecosystem assessments and, 6
relevance and, 25
scenario exercises and, 154, 182–183,
state and trend assessment and, 145–147
types of, 152–153
Underwriters Laboratories, 247
UNDP. See United Nations Development
United Nations Development Programme
(UNDP), 250
Use, sustainable, 241, 247
Use indicators, 126
Use value, 139
User groups, 46
User needs assessments, 35, 44–45, 50–51
Valuation, 77, 97, 100, 138–142, 239
Value-added processing, 239
Van panchayats, 242
Variables, conceptual frameworks and, 87
Venezuela, 231
Viewpoints, conceptual frameworks and, 77
Vilcanota, Peru, 66, 106–107
Vocabulary, 25
Volition, 153
Water, Work, and Women program, 229
Water resources
bioenergy and, 6
ecosystem services and, 4
gender equality and, 229
Salar de Atacama, Chile and, 41
scale and, 96
“Working for Water” program and, 234
Web sites, 65, 123
264 | Index
Well-being, human, 87, 100, 132–142,
Wellbeing of Nations, 134
Wikipedia, 123
Willingness to pay, 140
“Working for Water” program, 234
Workplans, need for clarity in, 54–55
Workshops, 65, 169
World Bank, 250
World Conservation Strategy: Living
Resource Conservation for Sustainable
Development, 246
World Resources Institute, 48
Yachay, 106