The importance of deliberation in valuing ecosystem services in developing countries--Evidence from the Solomon Islands

Abstract

Monetary valuation of ecosystem services enables more accurate accounting of the environmental costs and benefits of policies, but this has rarely been applied in developing countries. In such contexts, there are particular methodological and epistemological challenges that require novel valuation methodologies. This paper introduces a new participatory, deliberative choice experiment approach conducted in the Solomon Islands. The research aimed to determine the value people placed on ecosystem services and whether participatory interventions to elicit deeper held values influenced the preferences expressed. Results found that the initial willingness to pay for a number of tropical forest ecosystem services amounted to 30% of household income. Following deliberative intervention exercises, key ecosystem services effectively became priceless as participants were unwilling to trade them off in the choice experiment scenarios, regardless of financial cost. The group based deliberative approach, combined with participatory interventions, also resulted in significant learning for participants. This included a more sophisticated view of ecological-cultural linkages, greater recognition of deeper held values, and greater awareness of the consequences of human actions for the environment. The use of a group-based participatory approach instead of a conventional individual survey helped to overcome many of the practical difficulties associated with valuation in developing countries. Given the impact of learning on valuation outcomes, participation and deliberation should be integrated into valuation of any complex good, both in developing and developed economies. However, such a methodology raises questions about how valuation can deal with unwillingness to trade-off key ecosystem services, which results in the breakdown of monetary valuation methods. Evaluation of the appropriateness of valuation processes and methodologies for assessing deeper held values and use of mixed-method approaches will be essential to ensure policies take into account the extent to which human life is dependent on ecosystem services.

Full text

The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands
Jasper O. Kenter
a,1
, Tony Hyde
a
, Michael Christie
a
, Ioan Fazey
b,
*
a
Institute of Biological, Environmental and Rural Sciences, Aberystwyth University, Llanbadarn Campus, Aberystwyth, Ceredigion, SY23 3AL, Wales,UK
b
School of Geography and Geosciences, University of St Andrews, Irvine Building, North St, St Andrews, Fife, KY16 9AL, Scotland, UK
1. Introduction
There are increasing calls to estimate the value of ecosystem
services in monetary terms (Carpenter et al., 2006, 2009; Suther-
land et al., 2009; TEEB, 2010). However, research on the valuation
of environmental goods and services is limited in developing
countries, where much of the worlds biodiversity is located (Abaza
and Rietbergen-McCracken, 1998; Christie et al., 2008; Fazey et al.,
2005; Georgiou et al., 2006). This paper introduces a novel,
participatory and deliberative approach to ecosystem service
valuation in a developing country context, and discusses how
deliberation may impact on the way environmental values are
expressed.
Ecosystem services are the ecological processes and mechan-
isms that result in the conditions that fulfil and sustain human life
(Daily, 1997). They can be categorised as provisioning services (e.g.
food, fuel and fibre); regulating services (e.g. water purification,
climate regulation); supporting services (e.g. photosynthesis); and
cultural services (which provide spiritual, aesthetic, educational
and recreational benefits) (Millennium Ecosystem Assessment,
2005). The contribution of ecosystem services to human wellbeing
is enormous (Costanza et al., 1997; TEEB, 2010), but many
ecosystem services are rarely traded directly or taken into
consideration by economic markets. This has led to a lack of
appreciation in policy-making of the critical role of ecosystem
services in maintaining livelihoods and wellbeing.
Global Environmental Change xxx (2011) xxx–xxx
ARTICLE INFO
Article history:
Received 4 May 2010
Received in revised form 28 December 2010
Accepted 5 January 2011
Keywords:
Ecosystem services
Participatory approaches
Environmental values
Developing countries
Solomon Islands
Social choice experiments
Deliberative valuation
ABSTRACT
Monetary valuation of ecosystem services enables more accurate accounting of the environmental costs
and benefits of policies, but this has rarely been applied in developing countries. In such contexts, there
are particular methodological and epistemological challenges that require novel valuation methodolo-
gies. This paper introduces a new participatory, deliberative choice experiment approach conducted in
the Solomon Islands. The research aimed to determine the value people placed on ecosystem services
and whether participatory interventions to elicit deeper held values influenced the preferences
expressed. Results found that the initial willingness to pay for a number of tropical forest ecosystem
services amounted to 30% of household income. Following deliberative intervention exercises, key
ecosystem services effectively became priceless as participants were unwilling to trade them off in the
choice experiment scenarios, regardless of financial cost. The group based deliberative approach,
combined with participatory interventions, also resulted in significant learning for participants. This
included a more sophisticated view of ecological-cultural linkages, greater recognition of deeper held
values, and greater awareness of the consequences of human actions for the environment. The use of a
group-based participatory approach instead of a conventional individual survey helped to overcome
many of the practical difficulties associated with valuation in developing countries. Given the impact of
learning on valuation outcomes, participation and deliberation should be integrated into valuation of
any complex good, both in developing and developed economies. However, such a methodology raises
questions about how valuation can deal with unwillingness to trade-off key ecosystem services, which
results in the breakdown of monetary valuation methods. Eva luation of the appropriateness of valuation
processes and methodologies for assessing deeper held values and use of mixed-method approaches will
be essential to ensure policies take into account the extent to which human life is dependent on
ecosystem services.
ß2011 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +44 0 1334 463937; fax: +44 0 1334 453949.
E-mail addresses: jasper.kenter@abdn.ac.uk (J.O. Kenter), ath@aber.ac.uk
(T. Hyde), mec@aber.ac.uk (M. Christie), ioan.fazey@st-andrews.ac.uk (I. Fazey).
1
Current address: Oceanlab, Institute of Biological and Environmental Sciences,
University of Aberdeen, Newburgh, Aberdeenshire, AB41 6AA, Scotland, UK.
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001
Contents lists available at ScienceDirect
Global Environmental Change
journal homepage: www.elsevier.com/locate/gloenvcha
0959-3780/$ – see front matter ß2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.gloenvcha.2011.01.001

Over the last 10 years, there has been increasing interest in
methods that can estimate the monetary value of ecosystem
services so that the environmental costs and benefits of policies
and land use change can be accounted for (Carpenter et al., 2006,
2009; Costanza et al., 1997; Sutherland et al., 2009; TEEB, 2010). In
developing countries, the number of Payment for Ecosystem
Services schemes is increasing, for example where water users pay
to protect upstream water resources or farmers are paid to prevent
erosion through afforestation. Such projects, if implemented as
community-based schemes, have the potential to both conserve
nature and improve the welfare of rural people (Tallis et al., 2008).
However, relatively few environmental valuations have been
conducted in countries with the least developed economies (Abaza
and Rietbergen-McCracken, 1998; Christie et al., 2008; Fazey et al.,
2005; Georgiou et al., 2006), despite such countries harbouring the
majority of the world’s biodiversity and the high dependence of
people on those services for their survival and livelihoods (Christie
et al., 2008).
There are many different tools for monetary valuation of
ecosystem services. These include market-based, revealed prefer-
ence and stated preference methods. An example of a market-
based approach is replacement-cost analysis, where the cost of
replacing certain ecosystem services by technology is calculated,
e.g. replacing flood protection by mangroves with coastal defence
works (Gunawardena and Rowan, 2005; Winpenny, 1991).
However, the reliability of such methods is frequently undermined
by limited or incomplete knowledge of ecological systems
(Gunawardena and Rowan, 2005; McCauley and Mendes, 2006).
Also, they fail to capture the total value of such environmental
goods because many services, such as nutrient cycling (supporting
service) or aesthetic values of ecosystems (cultural service) are not
easily replaced by technology.
Another approach is to use revealed preference methods, where
some kind of marketed good that has a monetary value, such as
house prices or travel cost, is used as a proxy to reveal the value of a
non-marketed good. It is nonetheless difficult to establish mean-
ingful relationships betweenthe price of a marketed good and all but
a few ecosystemservices. This is particularlythe case where markets
such as property and transport are underdeveloped,as in many rural
areas of developing countries (Christie et al., 2008).
The third approach is to use stated preference methods, which
do not rely on existing markets. Instead, respondents are asked for
their willingness to pay for environmental goods in a number of
hypothetical scenarios. For example, participants may be asked for
their willingness to pay for conservation programmes which
improve biodiversity or a number of ecosystem services such as
erosion protection, water availability or even ‘ecosystem health’
(Barkmann et al., 2007). Because a hypothetical market is
simulated, there is the advantage that practically any good can
be valued, including more subtle benefits of the environment such
as those provided by cultural ecosystem services. Stated prefer-
ence techniques are therefore likely to be the most suitable
approach for monetary valuation of many ecosystem services in
developing countries.
Nonetheless, a number of theoretical, methodological and
epistemological challenges remain. These include low literacy
rates and language barriers, especially as the techniques often rely
on questionnaires (Christie et al., 2008; Whittington, 1998);
difficulties in explaining hypothetical scenarios (Whittington,
1998); lack of local research capacity for implementing complex
techniques (Alam, 2006; Christie et al., 2008; Whittington, 1998,
2002); and assumptions by researchers that participants have
similar ways of thinking as they do (Alam, 2006; Christie et al.,
2008; Lu et al., 1996).
Another challenge are the utilitarian assumptions associated
with welfare-economic theory that form the basis of monetary
valuation (Hanemann, 1984). These state that individuals seek to
maximise their benefit and minimise their cost, that preferences
are stable and transitive, and that utility curves are comparable
between individuals (Kahneman, 1986; Urama and Hodge, 2006).
Further, values may be lexicographic (meaning that they will not
be traded-off), or individuals may express multiple values (Spash,
1998; Urama and Hodge, 2006). These issues can be challenging in
any context. In many developing countries, however, people can
have limited experience of market mechanisms if they rely on
subsistence livelihoods. It is also not clear whether the assump-
tions underlying monetary valuation are upheld in these
circumstances.
A key issue specific to stated preference methods is that they
require respondents to take income constraints into account when
stating preferences (Arrow et al., 1993). But when incomes are low
and when people heavily rely on biodiversity for their livelihoods,
the values expressed may not properly reflect the true value of an
environmental good or service to their wellbeing (Abaza and
Rietbergen-McCracken, 1998; Hearne, 1996). For example, for
those dependent on subsistence farming, livelihoods will rely
heavily on nutrient cycling services. However, their monetary
income could be very low relative to what they believe is the actual
value of nutrient cycling for maintaining wellbeing. If participants
take their income restraints into account when asked what they
would be willing to pay for this service, the full value of the services
to them is not reflected. If they don’t take income constraints into
account, the assumption that people make choices as if they would
actually have to pay the amount asked for in a hypothetical
scenario, is violated.
A further, more general issue with stated preference techniques
is the assumption that preferences are pre-formed. This has
encouraged methods that use individual preference surveys.
However, values are not pre-formed but ‘constructed’ through
deliberation, and conventional methodological individualism fails
both to capture collective values and to make use of dialogue in
order to encourage reflection of what a persons’ values truly are
(Spash, 2008b).
Communal discussion for decision-making is particularly
important for many indigenous societies, such as those with
customary tenure systems, where land management and decision-
making on environmental goods is decided upon at the clan or
extended family level. Dialogue and deliberation that promote
reflection are also a key component of participatory and action
research methods, which are receiving increasing attention as
research funders recognise the need to find new mechanisms for
knowledge exchange and the co-production of knowledge
(Armitage et al., 2008; Fazey et al., 2010; Folke et al., 2005;
Pahl-Wostl, 2009). Such processes are important for enhancing
learning at a range of scales, promoting adaptive capacity for
responding to complex social and ecological issues, and for
promoting more equitable decision-making (Armitage et al.,
2008; Fazey et al., 2010; Folke et al., 2005; Pahl-Wostl, 2009).
Not surprisingly, introducing communal discussion and pro-
viding time for participants to think in valuation has been shown to
improve the quality of decisions (Urama and Hodge, 2006;
Whittington et al., 1992). This is particularly the case when
research is conducted with those who have had poor access to
education (Urama and Hodge, 2006). It has also been suggested
that group dynamics draw out greater attention to less obvious
values of the environment (Kaplowitz, 2001; Kaplowitz and
Hoehn, 2001), which is important for indigenous societies that
may have strong but subtle ties to the natural capital upon which
they depend.
While there is certainly potential for integrating group
discussion with stated preference techniques, it is not clear
whether these methods and their underlying assumptions can be
J.O. Kenter et al./ Global Environmental Change xxx (2011) xxx–xxx
2
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

applied for regions where monetary economies are weak, where
there is strong dependence on biodiversity, and where collective
decision-making is common. It is also not clear whether group
deliberation changes the values expressed by participants and
what this means for understanding the value people place on the
services they rely on. There is a paucity of economic valuation
studies from countries with developing economies where strong
reliance on subsistence-based economies remains and there have
been no previously published studies of social, group based
decision-making in environmental choice experiments.
This paper therefore presents a case study from the Solomon
Islands that develops and implements a participatory, group-based
methodology to determine the value of ecosystem services. It uses
stated preference techniques in a context where there is a strong
emphasis on subsistence economies (Bourke et al., 2006) and
where precedents have been set for participatory research
methods that aim to promote capacity building and enhance local
learning about complex social and ecological issues (Fazey et al.,
2010). The expectation is that the participatory and deliberative
methodology elicits deeper held values and more sophisticated
points of view on environmental goods. Specifically, the research
addresses three main questions: (1) What value do local people
place on ecosystem services? (2) Does deliberation change the
values expressed? And (3) What type of learning was encouraged
by deliberation? After a presentation of methods and results, we
discuss the implications of this research for valuing the environ-
ment in rural areas of developing countries.
2. Materials and methods
2.1. Study area
Kahua (16280
0
E to 162815
0
E, 10825
0
Sto10840
0
S), on the island of
Makira, consists of Wards 12 and 13 of the Makira-Ulawa province
of the Solomon Islands (SI) (Fig. 1). Average rainfall in Makira is
3600–4000 mm, with little annual variation (Allen et al., 2006),
while rainfall may be significantly higher in the more mountainous
Kahua region (Garonna et al., 2009). Topography consists of a
narrow coastal area, steeply rising into rainforest-covered hills.
Kahua has two large rivers, the Pihuru in the east and Warihito that
marks the border of Kahua in the west. Leaf-hut villages are dotted
along the coast and along these rivers. People largely subsist on
banana and sweet potato, with copra and cocoa being produced for
the market on a small scale (Allen et al., 2006; Fazey et al., 2007).
There are no roads or electricity. Rough seas also limit transport.
The SI archipelago has extremely high species endemism and
diversity (Lamoreux et al., 2006) and is one of a limited number of
places globally with large remaining tracts of coastal tropical
rainforest (Bayliss-Smith et al., 2003). The SI harbour an
exceptional cultural diversity, and are part of the Melanesian
biocultural hotspot (Loh and Harmon, 2005; Maffi, 2005).
Rural communities in the SI live a life that is directly dependent
on local ecosystem services, relying on subsistence agriculture,
fishing, hunting, wild food and non-food harvesting (Bourke et al.,
2006; Jansen et al., 2006). But they face a number of pressures on
resources, such as those from logging and mining companies,
conservation organisations (Hviding, 2003), increasing impacts of
climate change (Donner et al., 2005; Ebi et al., 2006) and
endogenous processes such as rapid population growth (Bourke
et al., 2006; Fazey et al., 2007). Although it is one of the few
unlogged areas of the SI (URS Sustainable Development, 2006), in
other aspects the Kahua region is no exception. Recent remote
sensing research shows that primary productivity has decreased in
recent years, with more marked decreases near villages (Garonna
et al., 2009). Locals state that the provision level of many
ecosystem services on which they depend, including provision
of building materials, water purification, crop disease regulation
and soil fertility building has decreased. The stress that this causes
in communities leads people to look for monetary solutions
(Bourke et al., 2006; Fazey et al., 2007), but this appears to
reinforce many of the problems by intensifying pressure on
resources. While social life revolves around traditional customs
and values, these have also come under pressure from the drive for
monetary wealth and population growth (Fazey et al., 2007).
Kahua consists of western, central and eastern zones, and
between these, there are clear differences in levels of income and
amount of outside influence and cultural change, related to the
amount of cash crops grown and transport opportunities for
accessing markets. The largest central area has little access to flat,
fertile land and no sheltered anchorages, making shipping
connections unreliable and extremely limited. The eastern and
especially western areas have access to wider river flood plains
enabling the planting of more cash crops than in the central region.
Further, west Kahua has a reliable regular shipping connection to
the SI capital Honiara and is close to a small road that leads to the
provincial capital Kirakira. This means that the west has both the
greatest area of highly fertile land for planting cash crop and has
greater access to markets. West Kahuans have on average over four
times the number of cocoa trees as east Kahuans, and over 12 times
more than the central region (Fig. 2). Nonetheless, even in the west,
monetary income is extremely limited.
In 2000, a grassroots non-governmental organisation, the
Kahua Association (KA), was founded by local people without
external assistance to improve ability for communities to respond
to mounting pressures. The KA is committed to sustainability and
equality, has a focus on learning and building capacity, and has
strong support within Kahuan society (Fazey et al., 2010). Both past
research and that presented in this paper placed significant
emphasis on co-management and co-design of research pro-
grammes. This approach has been widely supported by the KA and
local communities, and has received high rates of participation,
interest, and trust in external and trained local researchers (Fazey
et al., 2010).
In summary, Kahua has high biodiversity, strong dependence
on the local environment, low levels of income and education and
high levels of illiteracy, customary land-ownership, strong
traditional culture, and precedents and desire from local people
to engage in participatory and deliberative research approaches.
Kahua therefore provides a useful context for investigating the
influence of deliberative valuation approaches.
2.2. Research approach
The research approach combined participatory methods with
stated preference valuation techniques to develop new delibera-
tive and group based valuation methods. We adopted a set of key
participatory principles: local people were considered to be
capable of analysing their own realities and would be empowered
to do so; the main purpose of outside researchers was to catalyse,
convene, and facilitate the process; and learning of the participants
should be experiential rather than determined by transmissive
modes of information delivery managed by outsiders (Kumar,
2002). These principles were implemented by conceptualising a
process of learning occurring within three-tiers where external
researchers (tier one) trained and worked closely with local
research assistants (tier two), who facilitated the implementation
of focus groups and participation of community members (tier
three) (see Fazey et al., 2010). Group facilitation by local research
assistants, where input from external researchers was kept limited,
not only helped bridge language barriers, but also reduced the
chances of participants conforming to what they might believe
outside researchers would want to hear.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
3
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Fig. 1. The study area and its location (from Fazey et al., 2010).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
4
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Choice experiments (CEs) provided the basis to the valuation
methodology. CEs are a stated preference technique where
respondents are presented with a series of choices between more
or less desirable alternatives (Hanley et al., 1998). These consist of
a number of attributes, with each attribute being available at
different levels. For example, a simple CE could include water
purification; plants used for building materials and cost as
attributes. One alternative would be four months of clean water
per year and 15 min walk to ample building materials for $20.
Another could be 12 months of clean water and 3 h walk for
building materials for $30. An example of a choice task used in this
study is depicted in Fig. 3. The choice outcomes are used to
construct a probability model that is used to calculate the relative
importance of each attribute (Louviere et al., 2000). Usually, one of
these attributes provides a price tag (monetary, or a proxy measure
of value, such as travel time or distance), and this allows the
analyst to calculate a marginal willingness to pay for each attribute.
By interacting social, economic or demographic characteristics of
participants, such as gender, income or level of comprehension of
the choice tasks, with the choice attributes, differences in
willingness to pay between groups can be evaluated.
The general participatory approach provided the basis for
social, group based rather than individual based choice experi-
ments. While there is a disadvantage in that sample sizes are
reduced by using focus groups, the group format encouraged
greater involvement of people and reduced the resources and effort
that would have been required to conduct individually based
interviews. A further disadvantage of a group-based approach may
be that participants conform to social norms rather than form their
own opinion. This was ameliorated to some degree by facilitators
proactively prompting participants for their personal opinion.
Further, the social approach was more appropriate for a cultural
context where resource management decisions tend to be made
collectively through a process of discussion (Allen et al., 2006;
Bourke et al., 2006; Fazey et al., 2010).
To increase comprehension, deliberation, and reflection, the
choice experiment exercises were complemented by additional
participatory exercises and interventions (Section 2.3.4).
CEs for valuing ecosystem services of tropical forests in
developing countries have only been used in two previously
published studies, with both conducted in Indonesia (Barkmann
Fig. 2. Cash crops in Kahua per region, per household (adapted from Fazey et al.,
2007). Striped bars indicate coconut trees, shaded bars cocoa trees. Error bars
represent standard errors.
Fig. 3. Stages in the research design, execution and analysis, and staged structure of the focus groups. KA stands for Kahua Association (see Section 2.1).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
5
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

et al., 2007, 2008; Glenk et al., 2006, 2008; and see TEEB, 2010 for
an overview of studies). As far as we are aware, there are no studies
that have used social choice experiments for valuing ecosystem
services.
2.3. Research design and data collection
2.3.1. Overall experimental design and data collection
Key stages of the valuation are outlined in Fig. 4. In the initial
design phase research objectives were developed with the KA, local
facilitators were trained, and logistical and methodological aspects
designed and piloted. Data was then collected using a total of 46
focus groups held across Kahua, in 18 communities. 8 groups (17%)
were held in the eastern area, 25 (54%) in the largest central region,
13 (28%) in west Kahua. Groups were conducted separately for
women and men. In total 447 people participated from 30 villages
(10% of the total population) (Fazey et al., 2007). All village
inhabitants were invited to participate. Average size of groups was
9.7 (
2.89), average mean age 36 (8.7). With 55% of the Kahua
population under 20 (Fazey et al., 2007), young people were under-
represented, as most of the research period fell within the school
term.
Focus groups lasted between 3 and 4 h, and comprised several
staged exercises (Fig. 4). The central element was a choice
experiment, where focus groups expressed their preferences for
different scenarios describing change in a number of ecosystem
services. The workshop started with a warm-up exercise, which
asked participants to find 15 items representing important plants
from the forest, discuss their use, and whether their abundance
was increasing or decreasing, and why. The aim of the exercise was
to get participants to think about their forest environment and
discuss change within it, and to deliberate on why this change was
taking place. It also had the purpose of building participants’
confidence in their own knowledge. After this exercise, focus
groups undertook one set of choice tasks, then two intervention
exercises, followed by a second set of choice tasks. The purpose of
these interventions was to stimulate deliberation on issues related
to the ecosystem services that were being valued. Following the
second set of choice tasks, participants were asked to state what
they had learned most from the workshop. Further details on
choice experiment design, attribute selection and the intervention
exercises are outlined below.
2.3.2. Choice experiment design
The design of choice experiments includes decisions about
alternatives, attributes and their levels, a potential base or
reference scenario, the nature of the cost attribute, and the
framing of choices in putting them to participants. A dichotomous,
fold over choice task design with four attributes was used (Table 1),
resulting in eight possible choice tasks. Four sets of choice task
orders were made which were randomly allocated to groups. The
simple design, which omitted gradients in attribute levels, was
chosen to allow participants to focus on discussion rather than
having to re-interpret attributes, and to aid illiterate people by only
having two different attribute levels that could quickly be
committed to memory.
The baseline alternative at no extra cost consisted of all
attributes at base level. This approach was chosen instead of a
status quo scenario, because of variation between villages and the
complexity of implementing a variable status quo scenario in
combination with a dichotomous design without intermediate
levels. It also maximised time spent on deliberating scenarios
without the need to establish a status quo with participants.
The meaning of the alternatives, attributes and their levels was
explained in great detail before commencing with the choice
experiment, and then repeated more summarily before each choice
task in order to remind participants of the meaning of attributes
and their levels (Table 1). The importance of taking income
constraints into account was underlined. Alternatives were framed
as scenarios similar to development programmes, which partici-
pants were familiar with. Participants were therefore making
choices between programmes based upon their preference and
their ability to pay the associated cost for delivering such a
programme. In terms of decisions, participants reached consensus,
Fig. 4. A sample choice task, where participants would choose between three
hypothetical scenarios. shows number of stones placed, representing attribute
levels. Scenarios were described as the potential future outcome of a development
program. Costs attached to the programmes were the costs per household, which
would go into a community fund to finance the program. Program C is the default
scenario at no extra cost. This program remains identical across tasks.
Table 1
Choice attributes, their scenario levels and current situation in Kahua. Participants perform a number of tasks where they choose between three scenarios that consist of a
combination of ecosystem service attributes at either a base or an alternative level. One scenario always stays the same (the default scenario), has all attributes at the base
level, and has nil cost. The other scenarios have a cost attached to them, but provide alternatives that are generally seen as improvements. By modelling repeated choice tasks,
the relative importance of attributes and their monetary values can be estimated.
Attribute ES types Base level Alternative level Current situation
Gue (rattan/vine)
(proximity)
Cultural, provisioning 4 h walk needed to harvest
enough to build a house
15 min walk needed to harvest
enough to build a house
2–5 h walk needed in most places.
Water quality Regulating Water unfit to drink for
8–10 months/year
Water of drinking quality
all year round
Mostly good year round, occasional
issues in eastern and western areas.
Gardens (type) Provisioning, cultural,
supporting
Three cocoa gardens,
one food garden
Three food gardens, one
cocoa garden
Cash crops very limited in central area,
increasing in east and especially west.
Subsistence mainstay everywhere.
Cost (SBD) No cost SBD $300 (US $20) or $500
(US $33) into a community
development program
ES: ecosystem services; SBD: Solomon Island dollar, SBD $15 equalled circa US $1 at time of data collection.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
6
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

or, when differences of opinion went unresolved, voted. Facil-
itators recorded discussed motivations for making particular
choices as well as the choices themselves.
Various participatory tools were used to enhance comprehen-
sion and inclusion. For example, ecosystem service attribute levels
were represented by stones, and when moving on to the next
choice task, facilitators would move stones around, explicitly
illustrating how scenarios (and therefore attribute levels) changed
from one scenario to the other. In the debates around which
choices to make, participants actively interacted with the scenarios
through pointing and temporarily altering represented scenarios
by moving stones around themselves to illustrate their position
and facilitate debate (Figs. 5 and 6). Participation of illiterate and
otherwise potentially less empowered members of the focus
groups members was enhanced by representing scenarios on
outside ground rather than using a school class setting, repeated
explanations, active inclusion by facilitators of all participants and
other techniques common to participatory action research
(Chambers, 2002; Kumar, 2002).
2.3.3. Attribute selection
Attribute selection aimed to identify goods that were compre-
hensive in terms of representing different ecosystem services as
well as being both relevant and understandable to participants.
The four attributes used in the choice experiments were gue, water
quality, food/cocoa gardens, and cost (Table 1). ‘Gue’ (Calamus spp.)
is the local name for a rattan that is harvested for use as rope in
construction of dwellings. There has been a severe decrease in its
local distribution due to overharvesting, and as such gue is
representative of overexploited provisioning forest ecosystem
services. It also provides cultural services, as it is a key material for
maintenance of traditional huts and feasting houses.
The second attribute was water quality. Hydrological regulato-
ry services have not been degraded in most areas of Kahua, but are
known by locals to have diminished in neighbouring regions after
logging operations, and the relationship between deforestation,
erosion and water quality is generally understood. As a compre-
hensive good, water quality relates to both turbidity and sanitation
issues. In most villages, either water from standpipes connected to
uphill sources or from rainwater collection tanks is used for
drinking, but rivers and streams are used for washing and bathing.
A combination of subsistence and cocoa gardens formed the
third attribute. One of the most important supporting service
outcomes of forest functions is to generate soil, prevent its erosion
and replenish nutrients for shifting cultivation food gardens. As
they are tended by extended families under traditional tenure,
where rights, benefits and labour are shared between relatives,
food gardens perform an important cultural as well as provisioning
function. Cash crop planting leads to an increase in monetary
income, but also to erosion of cultural elements associated with
subsistence agriculture (see Section 4). Food gardens have been
partially displaced by cocoa and coconut, though Kahuans still
have a far greater reliance on subsistence than cash crops. Gardens
are ca. 0.5–1 ha, and in 2009 an average sized cocoa garden
generated a profit of around SBD $2700 (US $180) year
1
.
Households typically manage three to four shifting cultivation
gardens; though in west and east Kahua some families also have
access to more extensive floodplain plantations. Cocoa and food
gardens form one rather than two attributes, because households
have a limit to how much gardens they can tend, and displacement
of one by the other is more realistic than additive scenarios.
Scenarios did not include complete displacement either way.
For the cost attribute, a standard monetary version was used, as
previous research in Kahua showed that almost all households
interact with the monetary economy through growing cocoa or
copra (Fazey et al., 2007), and most families pay school fees. In
terms of the payment vehicle, the attribute was presented as a
contribution to a community fund. Attributes and their levels were
established in a separate, exploratory focus group with Kahua
Association representatives and cost levels established using a
bidding game. The research design was piloted with the same
group, and a second pilot was held with three additional focus
groups, which led to upward adjustment of the cost attribute
levels. Levels of the attributes are detailed in Table 1. Data from the
pilots were not used in the analysis of the results presented in this
paper.
2.3.4. Intervention exercises
Two intervention exercises were conducted between the sets of
choice tasks to determine whether choices changed as the result of
deliberation (Fig. 4). These interventions provided specific focus on
the role of some of the ecosystem services participants were
discussing. The exercises took around 30 min each. No outside
information from, or perspectives of the researchers or local
facilitators was delivered. A detailed review of the results of the
intervention exercises is beyond the scope of this paper, but they
are summarised briefly below.
In the first intervention exercise, participants were asked to
quantify their use per household for a range of subsistence and
wild goods, and then price them according to the local market and
total them. The purpose of this was to make the use value of goods
associated with the forest environment more explicit in monetary
Fig. 5. A focus group considering a choice task.
Fig. 6. The outside setting, where choice attributes are physically represented by
stones, makes it easy to interact with the choice scenarios. This aids deliberation
and participation.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
7
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

terms. In each group there were at least one or two participants
who had the basic arithmetic skills necessary to help the group
complete the exercise. All of the focus groups stated surprise at the
total value, which was on average SBD $13,149 (SE = 632) per
annum, equivalent to US $877.
The second intervention exercise asked participants to name
and rank five negative impacts of cocoa and other cash crops, and
then name and rank five negative impacts on kastom specifically.
This Melanesian Pijin word broadly relates to traditional culture,
including religion and economics, practices and values. Kahuans,
like other Melanesians, generally identify strongly and positively
with kastom, both under the influence of intense cultural diversity
of Melanesia and the pressures of modernisation (Akin, 2004;
Lindstrom and White, 1994). The discussions that ensued were
mostly related to the five themes outlined in Table 2: deforestation
and decrease of food security; violation of cultural principles;
privatisation of land; loss of community and social cohesion; and
more inequality, jealousy and division. Participants clearly
established a number of social-ecological linkages and exposed
vulnerabilities associated with an increase of cash crop growing,
stating explicitly the relationship between environmental degra-
dation and social-economic changes.
2.4. Data analysis
Choice results were used to construct a multinomial logit model
using NLOGIT 3 software. The model is specified in the appendix,
for more detail also see Kenter (2010). For a full derivation, see
McFadden (1973) or Louviere et al. (2000). For further details on
using choice modelling to value the environment, see Boxall et al.
(1996) and Hanley et al. (1998).
Separate models were used for the first and second set of
choices (rather than coding the intervention as a dummy variable
and interacting it with the attributes). The rationale for this was
the excellent model fit of these models (see Section 3.1), and that it
facilitated evaluating interactions between the intervention and
social-economic variables. A single alternative specific constant
(ASC) was set to 0 for the baseline scenario, and 1 for the other
scenarios. Age, gender, comprehension rating and area were effects
coded and interacted with attributes rather than the ASC, as the
number of choices for the baseline scenario was limited (Section
3.1). Associations between these data were tested with Pearson’s
chi-square, using SPSS 17.
Learning outcomes per participant were coded into a multiple
response set of 16 categories, with 2 out of 286 (0.17%) answers not
being assigned at least one code. Codes were aggregated per focus
group, as individuals tended to respond to, add to or confirm each
other. Using SPSS, we calculated Kendall tau-B coefficients to
evaluate correspondence of learning outcomes between different
areas, level of comprehension and gender. The coefficient
t
(tau)
expresses the proportion of pairs of values, in this case the number
of groups stating a certain learning outcome, that is concordant at
the ordinal level.
3. Results
3.1. The value of ecosystem services
Model fit was excellent for both for the first (
r
2
= 0.45,
p<0.0001) and second set of choices (
r
2
= 0.44, p<0.0001). In
both choice tasks participants overwhelmingly chose for scenarios
with environmental improvements (Table 3). Overall, the baseline
scenario with no additional cost was chosen in a mere 0.72% of the
tasks. Water quality (1st set:
b
= 2.120, p<0.0001; 2nd set:
b
= 1.934, p<0.0001) and subsistence gardens (1st set:
b
= 1.893,
p<0.0001; 2nd set:
b
= 2.157, p<0.0001) were deemed to be
more important than the vine gue (1st set:
b
= 0.678, p<0.01; 2nd
set:
b
= 0.644, p<0.0001). The lower value and significance of the
gue attribute suggests it is of second-order influence compared to
water quality and gardens in choice making.
Willingness to pay for an attribute can easily be calculated by
dividing its coefficient by the negative of the cost coefficient. In the
first set of choice tasks, before the intervention exercises,
participants were willing to pay SBD $495 (US $33) per household
per year for an increase in water quality, SBD $442 (US $29) for an
increase in food over cash crop gardens and SBD $159 (US $11) for
an improvement in gue abundance. The value attributed to gue
from direct market prices obtained during the intervention
exercise was similar to that obtained in the first set of choices
(SBD $141, SE = 18.07). While the latter is non-marginal, given the
wide margin maintained between our dichotomous attribute
levels, the comparison suggests that implicit prices from this
choice model reflect reality to a high degree. Certainly, these prices
are very high, given that annual monetary income for almost all
households in Kahua is US $150–500 per year (estimate based on
survey results from Fazey et al., 2007).
Table 2
Themes surfacing in discussions in the second intervention exercise.
Deforestation and decrease
of food security
Cash crop plots are cleared in the forest, or more commonly they displace food gardens away from the coastal and river
plains, where the most fertile land is, into newly cleared areas in the hills, where soil erodes quickly and yields are lower.
This is partly because men tend to take care of cash crops and prefer to grow them nearby, forcing women to grow food further
away. There is also less time to grow food, while the income from cash crops is irregular, not always sufficient to purchase
imported rice, and is often spent on other things such as school fees or alcohol.
Violation of Kahua
principles
The four Kahua principles consist of the related principles of discussion, asking permission, sharing and care. These are
violated when an area of forest is cleared and cash crops are planted without proper discussion/permission, and when land,
labour and proceeds are not shared with others who have rights to the same land.
Privatisation of land This leads to land disputes, and this is rarely resolved by uprooting cash crops, but usually land is divided up, such that the
dispute-plot is privatised to the original user and other plots or uncleared forest are allocated to relatives who shared rights to
the original plot. This privatisation has major cultural impacts in that it erodes tradition of shared land, labour and risk, and
leads to further forest clearance.
Loss of community and
social cohesion
People are increasingly concerned with private instead of community business, mostly due to increased work pressure.
Increasingly money comes before kastom, and people with a lot of money are reported to be more disrespectful of others, as
they can afford to pay compensation to the aggrieved party, the traditional method of doing justice. The ‘cocoa life’ is very
different from the ‘kastom life’, where families grow food and work together; they share and take care of each other. Because of
the increasing importance of money and private business, family members now sometimes ask to be paid for their labour/help,
people sell food to each other instead of sharing and do not respect kastom prices set by chiefs but demand market prices.
More inequality, jealousy
and division
While some people have started to grow a lot of cocoa, others either do not have the land available or do not want to grow a lot
of cash crops. This leads to inequality, hatred and jealousy in communities between the cocoa growers and others. Because of this
it is becoming more difficult for people to work together in communities.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
8
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Eastern and central areas showed lower value outcomes (1st
set:
b
cost
=0.638, p<0.001) than overall (1st set:
b
cost
=0.428,
p<0.001). Young groups (mean age <30) expressed a stronger
preference for lower-cost scenarios (1st set:
b
cost
=0.708,
p<0.01), while middle aged-groups had an average cost coeffi-
cient (1st set:
b
cost
=0.464, p<0.01). Thus value placed on the
environmental services increased with age.
3.2. Did deliberation influence choices?
The intervention strongly influenced choices made (Table 3).
The cost attribute became insignificant, indicating that overall this
attribute stopped influencing choice making in the second set of
choices. In other words, after the intervention, participants solely
based their decisions on the level of environmental improvements
and began to ignore the monetary costs. While it is clear that
willingness to pay for the ecosystem services was much greater in
the second than the first set of choices, because respondents
overall did not trade off cost against other attributes, it was not
possible to estimate implicit prices for the post-intervention
choices.
There were also differences in the choices made in different
regions, in gender, and in relation to the extent to which
participants comprehended the exercise and discussions. For
regional differences, when the western area was excluded from
analysis, the remaining areas showed a cost-coefficient that still
bears significance, though much reduced in strength, as is the co-
efficient itself (before intervention:
b
=0.638, p<0.0001; after:
b
=0.273, p= 0.02). Therefore, those from the west placed higher
value on the environment from the beginning and were affected
more by the intervention than those in east and central regions.
With regards to gender, there was no major difference between
women and men as regards total willingness to pay, but there were
gender-specific effects of the intervention exercises on the
perception of subsistence versus cocoa gardens. Whereas men
cared less for food gardens than women before the intervention
(overall:
b
= 1.893; males:
b
= 1.518), in the second set of choices
men chose more strongly in favour of subsistence (overall:
b
= 2.157; males:
b
= 2.467).
Finally the quality of deliberation in the choice experiments
also influenced outcomes. Discussion quality was measured in that
it formed a component of the comprehension rating (CR) set by
facilitators per group after completion of the choice experiments.
None of the focus groups had a very poor or poor rating, 9% was
deemed to be moderate, 37% good and 54% very good. A moderate
Fig. 7. The different aspects learnt by participants through engaging in the participatory valuation process.
Table 3
Choice models (multinomial logit).
b
expresses the weight of an attribute in choices
made. Positive values indicate that preference for a scenario increases with an
increase in the level of the attribute. Indented variables represent the weight of
attributes when interacted with certain characteristics of the participants. ASC
represents variation that cannot be ascribed to other variables. Overall probability
of the model expresses the likelihood that it is significantly different from a
constants only model.
r
2
(Pseudo R
2
) is an expression of model goodness-of-fit.
Weight parameter (
b
) Choices before
intervention
Choices after
intervention
Gue 0.678 ** 0.644 **
Water 2.120 **** 1.934 ****
Gardens (subsistence over cocoa) 1.893 **** 2.157 ****
Male 1.518 *** 2.467 ****
Cost 0.428 *** (0.131) NS
East + central areas 0.638 **** 0.273 *
Mean age <30 0.708 ** 0.431 T
Mean age 30–40 0.464 ** (0.184) NS
CR: good 0.641 ** (0.150) NS
CR: very good 0.280 T (0.055) NS
ASC (non-default) 2.647 *** 2.406 *
Observations 184 230
Log-likelihood 54.7 91.6
p(Chi
2
); DF <0.0001; 4 <0.0001; 4
r
2
(Pseudo R
2
) 0.45 0.44
****Significant at p<0.0001; ***p<0.001; **p<0.01; *p<0.05; T: tendency at
p<0.01; ASC: alternative specific constant, equal to 1 for non-default scenarios; CR:
comprehension rating; NS: not significant; DF: degrees of freedom; cost-
coefficients at SBD $100 (Solomon Island dollars);
r
2
in reference to a constants
only model –
r
2
>0.4 is comparable to an R
2
of over 0.8 in a linear regression model
(Domencich and McFadden, 1975; Hensher et al., 2005).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
9
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

CR meant that participants generally understood the tasks but had
not fully grasped the meaning of all attributes. ‘Good’ meant that
participants had a complete understanding of all aspects of the
choice tasks. The difference between ‘good’ and ‘very good’
comprehension was defined as the latter having a higher degree of
discussion and deliberation before making choices, with choice
making being more considered.
CR clearly influenced results. Groups with ‘merely’ a good CR
showed a much stronger preference for less costly scenarios in pre-
intervention choices (
b
cost
=0.641, p<0.01) than those with the
highest rating (
b
cost
=0.280, p= 0.06). The low significance of the
cost attribute for the highest CR groups shows that they only
tended to look at cost in choosing scenarios, even before the
intervention, with decisions largely being based on the environ-
mental attributes.
3.3. What did groups learn from the participatory, deliberative
approach?
Participants stated a wide range of learning outcomes. These
relate to 16 topics (Fig. 7), ranging from improved knowledge, such
as uses of certain plants, and skills (both raised by 9% of groups) to
general learning outcomes about the environment, such as
decreasing forest resources (73% of groups), that the forest or
land needs better management (71%) or that they felt they now had
a better understanding of the value of the environment (71%). The
majority of groups also stated that they had learnt about the
negative impacts of cash crops (64% of groups).
Learning outcomes were not homogenous across groups (Table
4,Fig. 8). There were differences across different regions of Kahua
(
t
= 0.56, p<0.01; Fig. 8a), comprehension rating (
t
= 0.65,
p<0.01; Fig. 8b) and gender (
t
= 0.33, tendency at p<0.1;
Fig. 8c). In relation to regions, groups in the eastern and central
areas of Kahua less often raised cultural values or cash crops than
western groups, and more often made more general statements,
such as that they had learnt to be more concerned about the
environment, or that they had acquired a better understanding of
its importance. The same differences can be seen between groups
with the highest comprehension rating, and groups with a
moderate to good CR. This similarity in the ‘shape’ of learning
outcomes (Fig. 8) is confirmed by tau coefficients: outcomes for the
eastern/central area are considerably more akin to those of groups
with a moderate to good comprehension rating (
t
= 0.84,
p<0.001) than to the western area (
t
= 0.56, p<0.01). Chi square
tests show no significant associations between regions and
comprehension ratings, which implies that there is some other
reason why western groups and groups with the highest CR, and
thus the highest quality of discussion, have similar learning
outcomes that are distinct from other groups. These differences in
learning outcomes appear to reflect differences in valuation
outcomes, where both high-CR groups and western groups showed
much stronger than average preferences for environmental
improvements.
Care for future generations, impacts of population growth,
displacement of subsistence resources (such as food gardens),
concerns about cultural values, and rising food prices due to
increased demand and reduced supply were more often raised as a
learning outcome by women than men (Fig. 8c). Men said they
learnt most regarding the need for better land management and
the negative sides of cash crops.
4. Discussion
4.1. What values do people of Kahua place on ecosystem services?
In the second choice tasks the value placed on ecosystem
services was so high as to be incalculable; as cost of the scenarios
stopped influencing choices, no tradeoffs are made, so it becomes
impossible to anchor preferences to a monetary value. But even in
the first tasks, participants were willing to pay markedly high
amounts for improvement in environmental goods given that most
households make less than US $1 per day. The total willingness to
pay was US $73, approximately 30% of mean annual monetary
income per household. Most notable is the strong preference for
subsistence over cocoa gardens. This outcome may in part be a case
of status quo bias, the effect where attachment to the present
situation and aversion to loss limits a person’s enthusiasm for
change (Kahneman et al., 1991; Samuelson and Zeckhauser, 1988).
However, important explanations given for these choices were that
people preferred home grown food for its traditional cultural
significance (including its use in feasting) and greater nutrition
than substitute imported food. Thus, though the amount of cash
crops grown has been increasing in Kahua (Fazey et al., 2007),
when having to make a direct trade-off, subsistence is clearly
preferred.
This has important implications for managing vulnerability.
Planting of cash crops may decrease food security, as they are
grown in high fertile areas previously used for food crops. At the
same time there is considerable desire to earn money, driven as
much by personal interests in accumulating wealth as by the need
to meet important needs such as paying for health care, transport,
and school fees. The continued recognition of the importance of
food gardens highlighted in this study means that the desire for
money has not yet superseded considerations for maintenance of
essential ecosystem services, and may provide an important social
hook to help promote more sustainable activities.
As regards water quality, participants strongly associated this
with a clean and healthy environment and intact forests. The
attachment to living in such a place and maintaining it for future
generations seemed to be of greater importance than the simple
use value of clean water, which, participants realised, could also be
achieved by rainwater catchment tanks or similar substitutes. As
such, the attribute of water quality effectively represented
regulating ecosystem services.
Gue, a vine that is mainly used as rope for traditional houses,
was seen to be of less fundamental importance than the other two
Table 4
Kendall tau-B coefficients of similarity between learning outcomes of different areas, comprehension ratings and genders. Tau (
t
) coefficients express the proportion of rank
concordance.
East + central West CR: M-G CR: VG Male
East + central 1.00
West 0.56
**
1.00
CR: M-G 0.84
***
0.65
**
1.00
CR: VG 0.71
***
0.80
***
0.65
**
1.00
Male 0.62
**
0.53
**
0.60
**
0.56
**
1.00
Female 0.68
***
0.76
***
0.72
***
0.77
***
0.33 T
***Significant at p<0.001 (2-tailed); **p<0.01; T: tendency at p<0.1; CR: comprehension rating; M: moderate; G: good; VG: very good (highest rating).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
10
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Fig. 8. Comparison of the proportion of groups that identified different learning outcomes, for the nine most stated outcomes. Comparisons are between (a) different regions
of Kahua, (b) comprehension ratings by facilitators and (c) gender. The axes indicates the percentage of groups stating a certain outcome.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
11
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

attributes. Though it was associated with other plants and trees
used in building that were also in decline, having food security and
clean water were of greater importance than improved proximity
of plants that were only required once in a few years.
4.2. Did deliberation influence valuation outcomes?
The results clearly demonstrated that deliberation influenced
valuation outcomes. Three outcomes pointed to this: overall
intervention impact, specific impact of the intervention on male
choices, and the influence of discussion quality during choice
experiments on choice outcomes.
The main impact of the intervention was that the price of
scenarios lost significance to the focus groups. In other words,
participants started making decisions based almost solely on the
levels of the environmental attributes rather than the monetary
cost. This was clearly a result of new insights gained from the
deliberative exercises, as there was a congruent shift in discussions
and decision making in the focus groups. Participants, when asked,
consistently stated that they would pay their whole income
towards safeguarding ecosystem services; they would simply have
to ‘make sacrifices’ because of this imperative, which did not
surface as such in the first round of choices. Simply put,
participants stated that because ecosystem services are so
valuable, their conservation should take absolute priority over
short-term economic gains. This is rarely fully or explicitly
expressed by participants of valuation studies in countries with
more developed economies, suggesting that, when given time to
think and deliberate, those relying on subsistence more readily
appreciate the value of ecosystem services than those depending
on a monetary income. This raises important questions about
whether willingness to pay approaches are able to fully capture the
benefits of ecosystem services, unless those surveyed are in their
daily lives directly confronted with their dependence on these
services.
The second indicator of the importance of deliberation is the
specific impact of the interventions on the extent to which men
appreciate subsistence gardens. At first, while men still valued
subsistence over cocoa gardens, they did so to a lesser extent than
women. This is not surprising: women spend more time in
subsistence gardens, especially in areas where cash crops are more
abundant. Conversely, tending plantations (and spending profits
gained from them) is men’s work. After the intervention
discussions, men radically strengthened their preference for
subsistence, which was confirmed by more frequently stating
the negative impacts of cash crops as a learning outcome than
women (Fig. 8c).
Thirdly, the impact of deliberation within choice experiments is
clearly shown by the differences in valuation outcomes between
groups with the highest comprehension rating and groups with
lower ratings (Table 3), as the difference between the highest
rating and the level below was solely defined in terms of quality of
discussion and participation. Those who lacked lively discussion
tended to choose cheaper scenarios. Notably, quality of delibera-
tion appears to be much more important than level of formal
education in terms of valuation; while education levels of
participants were not known, we do know that females and
older-aged tend to be less educated (Fazey et al., 2007), but they
neither had lower comprehension of the scenarios nor bid lower
(Table 3).
4.3. What type of learning was encouraged by deliberation?
4.3.1. Range of learning outcomes
Results clearly demonstrated a range of learning outcomes for
participants demonstrating the potential value of participatory
approaches to valuation research (Fig. 7). First, participants better
understood the use value of subsistence goods. They became aware
that prices of these goods, if one were to buy them, would rise if
more cash crops were grown at the expense of traditional crops.
Second, many individuals had not realised before that new
plantations displaced traditional resources, and participants
worried about food security if they could depend less on
subsistence. Third, there was increased awareness of the cultural
value of the attributes and participants expressed worry that the
decrease in ecosystem services would lead to an erosion of cultural
principles, loss of community cohesion, and increased greed,
disputes, and social problems (c.f. Sillitoe et al., 1998). Fourth,
participants stated that they only realised after the focus groups
that decreased abundance of wild resources, such as gue, was a
result of human impact (overharvesting and increased land
clearing). This related to increased awareness of the interdepen-
dence between humans and their environment. Fifth, participants
expressed a greater bequest value of the environment. That is, they
expressed a desire for their children to be able to maintain current
ways of living, and a desire to conserve the environment for future
generations. These findings suggest that the interventions elicited
deeper held beliefs and better appreciation of socio-cultural,
economic and environmental sustainability, and realisation that
sustainability was threatened by erosion of natural and cultural
capital.
4.3.2. Learning and understanding of environment-culture
relationships
The results from the participatory approach that encouraged
learning have important implications for the way people from
Kahua relate to their environment, and how recognition of this
relationship can be used to promote more sustainable activities in
the region. Ecological knowledge, environmental management,
values, culture and identity are intimately intertwined in many
indigenous societies (Houde, 2007) and this may especially be the
case in places such as the Solomon Islands, where subsistence
dominates livelihoods. Key principles of Kahua society are sharing,
particularly of land and labour; discussion with and asking
permission of everyone who may be affected by a decision; and
mutual care (Fazey et al., 2010). Reproduction of these principles is
strongly linked to traditional interactions with the environment,
such as shared felling, tillage and crop care in shifting cultivation
systems under customary tenure, where land and labour is shared
within the clan or extended family. Thus natural capital performs
socio-cultural functions as well as ecological and economic ones
(Berkes and Folke, 1994; Chiesura and de Groot, 2003).
A mixture of environmental and cultural concerns is reflected in
stated learning outcomes (Fig. 7). The two most frequently
expressed outcomes relate to environmental themes. These are
recognition of the need for better forest and land management and
improved understanding of the importance of natural resources.
These themes were stated by a majority of groups in all regions, for
both genders and for all levels of comprehension. The third and
fourth most stated themes, the negative impacts of cash crops and
concern for cultural values, were specifically most stated by those
groups who also made the highest bids in the choice tasks: people
from West Kahua and groups with the highest comprehension
rating scores (Table 3,Fig. 8). In west Kahua, access to transport
and markets is greatest and cash crops have been developed most
extensively (Fazey et al., 2007). Remote sensing suggests that areas
in West Kahua have also suffered some of the greatest
environmental change (Garonna et al., 2009). There are also
indications that it is in the western area where the most rapid
recent social change has occurred, such as in consumption of
alcohol and erosion of social cohesion, which may be mediated by
greater circulation of money generated through cash crops. Thus,
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
12
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

people from West Kahua have probably had the greatest
experience of environmental and social change, and thus a greater
awareness of its consequences. Similarities in the shape of learning
outcomes between western groups and groups with very good
comprehension suggest that the latter have achieved a similar
awareness, but through deliberation instead of experience, as
increased learning led to high value outcomes for both groups.
4.3.3. Relating learning to theories of environmental values
The findings can be understood in relation to theories of human–
environment relationships. In environmental psychology, the
values-beliefs-norms (VBN) theory of environmental behaviour
(Stern, 2000; Stern et al., 1999) has become increasingly established
(Dietz et al., 2005; Hansla et al., 2008; Snelgar, 2006), though it has
not been tested in a developing country context. The theory
describes the process of translation of deeperheld values, consisting
of a three-factorbase of self-interest, altruisticand biocentric values,
into behaviour regarding the environment, through a number of
linkages (Fig. 9). Deeper values shape an ecological worldview, and
this is linked to awarenessof consequences of actions and perceived
ability to reducethreats to the environment, which eventually direct
norms and behaviours (Hansla et al., 2008; Stern, 2000; Stern and
Dietz, 1994; Stern et al., 1999). The VBN theory predicts the type of
learning that wouldneed to take place to generate changes in choice
outcomes: the surfacing of deeper held environmental values,
increased awareness of human responsibility for environmental
degradation, anda realisation that this outcome is not inevitable, i.e.
the result of an implicit choice.
Each of these types of learning can be seen in the results of this
study. In terms of deeper held values, utilitarian reasoning (best
goods for the best price), made way for more altruistic concerns
such as the Kahua principles and consideration for future
generations, and for biocentric values in the sense of a realisation
of the interdependence with nature. In terms of norms, partici-
pants expressed a need for better management, citing examples of
what they now understood to be wrong, such as felling large trees
for planking and leaving half of it to waste due to lack of planning,
and then doing the same thing over again the next year. And
indeed, focus groups that may have had a greater awareness of the
consequences of land use change also showed greater appreciation
for environmental goods.
4.3.4. Synthesis: the effects of learning on perceptions of the
environment
The results demonstrated five main consequences as a result of
learning. These were: (1) Increased and more sophisticated
understanding of social–ecological linkages (e.g. linking the
environmental and cultural consequences of land privatisation);
(2) Explicit recognition of deeper held values, such as the wish to
leave a healthy environment to one’s children, or the sense that one
is interdependent with nature; (3) Clarification of the use value of
non-marketed goods, such as subsistence resources; (4) Increased
appreciation of less obvious values of environmental goods (e.g.
goods may not justbe important for their direct usebut also for their
contribution to cultural identity); (5) Increased awareness of the
consequences of actions, and of one’s ability to intervene or change
behaviour (e.g. if resources are decreasing, deliberation may help
one realise thatthis may be due to humanactions rather than natural
processes). This type of learning also includes better appreciation of
the ‘you can’t eat your cake and have it too’ principle.
4.4. Implications for environmental valuation
Our case study was conducted in a remote area where the
mainstay of livelihoods consists of subsistence gardening, access to
markets was extremely limited, land is not privately owned and
local prices are set through traditional mechanisms rather than the
market. Nonetheless, the excellent model fit and strong consis-
tency in cost appreciation in the first set of choices clearly
demonstrate that participants were able to engage in utilitarian
thinking. Moreover, utilitarian strategies, in particular maximising
benefits while minimising costs, dominated the discussions during
the first round of choices. The high values expressed, which
represented 30% of participants’ estimated modal monetary
income, demonstrated that participants were able to translate
their reliance on these ecosystem services into rational choices.
This study therefore confirms that, at least in terms of participants’
capacity, choice experiments can be applied to measure non-
market environmental values in rural areas of developing
countries, even where subsistence makes up the mainstay of
livelihoods (cf. Barkmann et al., 2007, 2008).
The application of the methodology was, however, made easier
by existing precedents for participatory research approaches, and
because participants already relied on deliberative, group-based
decision-making structures and tenure systems. The use of a
group-based participatory approach instead of a more conven-
tional individual style survey helped to overcome many of the
practical difficulties associated with valuation in developing
countries (e.g. illiteracy), though this is dependent on facilitators
being given appropriate training. It also addressed the problem of
establishing equal knowledge among participants, which is
heterogeneous to start with (Urama and Hodge, 2006).
Fig. 9. Representation of the values-beliefs-norms theory of environmental behaviour (adapted from Dietz et al., 2005; Stern, 2000).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
13
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

However, after the post-intervention choices, the choice experi-
ment broke down, as the cost attribute became insignificant.
Participants were unwilling to trade-off ecosystem services as a
result of social learning. The importance of information and
discussion to valuation has been expressed many times (Arrow
et al., 1993; Hoehn and Randall, 2002; Macmillan et al., 2006;
Niemeyer and Spash, 2001; Schiffman and Kanuk, 1991; Spash,
2002, 2007, 2008a; Urama and Hodge, 2006; Whilehead and
Blomquist, 1991; Whittington, 1998; Whittington et al., 1992), but
in many cases this is only related to the direct consequences of a
change in management or policy. Rarely is a fully participatory
process proposed where information originates with participants
rather than experts, and where external researchers chiefly act as
facilitators (cf. Kumar, 2002; Pretty et al., 1995). High quality
discussion, particularly in combination with appropriate interven-
tions that stimulate analysis and linking social-economic, ecological
and cultural processes and change, increase the degree to which
deeper held value-bases and more sophisticated perspectives are
realised. This is particularly important for capturing the value of the
more subtle aspects of environmental goods, such as the cultural
ecosystem service of identity formation. Also, political theory
suggests that deliberation will prompt participants to look beyond
their immediate self-interest (Niemeyer, 2004).
If, as this research suggests, such learning is a requirement for
full realisation of the importance of nature, then the short surveys
that form the mainstay of valuation practice are inappropriate for
the elicitation of preferences for environmental goods, irrespective
of whether they are conducted in developing or developed
countries. At the least, preferences are not pre-formed but are
constructed through the process of eliciting them (Hoehn and
Randall, 2002; Spash, 2002; Urama and Hodge, 2006). This means
that environmental economists need to engage with other
disciplines that have developed theories of value (e.g. ethics,
social psychology) or give insights into social processes (e.g.
political science, anthropology), in order to design more sophisti-
cated processes for eliciting preferences. Issues that need to be
considered include confidence and trust in process and structure,
models of decision making, power relations, legitimacy and
inclusiveness, and the need to understand attitudes and motiva-
tion (Spash, 2007). Evaluation of the appropriateness of processes
and methodologies for eliciting and assessing deeper held values
will be essential.
Incorporating deliberation into the valuation process has,
however, raised an important methodological issue. Those
surveyed in this study came to the conclusion that key ecosystem
services and money were not of the same order of importance and
that the services they depended upon for livelihoods were
priceless. This poses a major methodological challenge, as tradeoffs
between services are required to generate estimates of monetary
value. We can see five methodological approaches that can help
reduce the chances of eliciting lexicographic preferences. These are
(1) using a so-called willingness to accept instead of willingness to
pay approach; (2) valuing less drastic, more marginal changes in
ecosystem service provision; (3) only valuing less important
ecosystem services; (4) scenarios without a cost attribute; (5)
scenario’s using a proxy good as a cost attribute (e.g. bags of rice,
livestock). However, these options have significant disadvantages
(Table 5), such as the risk of reducing quality of deliberation and
social learning.
The alternative is to accept cost-insignificance as a valid
outcome in its own right and to use non-monetary methods to
qualify values in addition to monetary valuation techniques. If the
primary aim of the valuation process is learning for both
participants and researchers regarding why nature is valuable,
rather than just a figure outcome, the conclusion that key
ecosystem services are priceless is quite valid.
Table 5
Methods that can help to avoid lexicographic preferences with their advantages and disadvantages.
Description Advantages Disadvantages
Willingness to accept
(WTA) approach
Asks participants for their
willingness-to-accept compensation
for loss of ecosystem services,
rather than their willingness to pay.
Avoids income constraints and
can make it possible to price
attributes of extremely
high value.
Does not reflect real market behaviour
in the absence of budgetary considerations.
Can lead to vastly inflated values.
Discouraged by accepted valuation
protocols (Arrow et al., 1993).
Attributes with marginal
changes only
Avoidance of drastic changes in
ecosystem service provision levels
in choice scenarios. Participants
are not asked to make major
sacrifices one way or the other.
Tradeoffs more tolerable/
plausible.
Difficult to establish in advance what
constitutes a non-marginal change.
More subtle values of environment
may not be elicited. Potentially less
meaningful discussion and reduced
social learning.
Attributes of lesser
importance only
Survey restricted to less important
ecosystem services (e.g. amount of
shading in plantations or specific
non-timber forest products)
Tradeoffs more tolerable/
plausible.
Important services not valued. Potentially
less meaningful discussion and reduced
social learning.
Scenarios without
cost attribute
Choice experiments where participants
make tradeoffs between ecosystem
services only.
Key ecosystem services can
be effectively traded off against
each other (e.g. food crops
versus forest produce)
providing insight into their
relative value.
No monetary estimates. Attributes need
to be carefully selected. Risk of reducing
quality of discussion and reduced social
learning.
Proxy cost attribute Use of some good which can replace money
as a cost-indicator, e.g. Shyamsundar and
Kramer (1996) used bags of rice, or one
might ask for willingness to invest time
to achieve a scenario outcome.
May improve comprehension
of choice tasks.
Using culturally highly significant goods
(e.g. pigs or shell money in the Solomon
Islands) may lead to unexpected implications
in terms of how tradeoffs are made. Relatively
neutral non-local goods (e.g. bags of rice) or
time may still face issues of budget constraints
and still lead to lexicographic preferences
regarding important ecosystem services
(unless a WTA approach is used in concordance).
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
14
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Acknowledgements
This research was made possible by funding from Wild Earth
and XminY (both Netherlands) and Aberystwyth University (UK).
We would like to thank Jacqueline Ketara, Angela Mataku, Aniceto
Giro Junior, Tony Piringi, James Ta’ata and others of the Kahua
Association for their guidance and support.
Appendix A. Specification of the multinomial logit model
In order to analyse choice experiments, a model is constructed
based on choice outcomes. This allows us to understand the
influence that each attribute of the experiment had on choices
made. Here, a mult inomial logit model was utilised, which is the most
common choice model used. Choice models are rooted in Lancastrian
consumer theory, and random utility theory. The first states that
when an alternative is chosen, it is not preferred per se,butforits
component attributes (Lancaster, 1966). In valuing ecosystem
services, these alternatives consist of different environmental
scenarios , composed of a number of ecosys tem services as att ributes.
Random utility theory states that utility U, or net benefit of an
alternative ihas a systematic compon ent V
i
that can be observed and
related to k nown variables, and a stochastic, unobs erved component
e
i
, which reflects individual idiosyncrasies (Manski, 1977; McFad-
den, 1973):
U
i
¼V
i
þe
i
(1)
The randomness referred to here is only from the point of view of
the researcher; respondents are fully aware of their utility function.
Utility is the satisfaction one gains from, or desirability of, a good.
When we assume that choice experiment participants aim to
maximise their utility, the probability Pthat an alternative iis
chosen over jfrom choice set Ccan be expressed as:
PðijCÞ¼PðU
i
>U
j
Þ¼P½ðV
i
þe
i
Þ>ðV
j
þe
j
Þ ¼ P½ðV
i
V
j
Þ>ðe
j
e
i
Þ
8i;j2C;i6¼ j
(2)
In order to estimate the probabilities in Eq. (2), one needs to make
a number of assumptions on the nature of the random component.
Multinomial logit models, also known as conditional logit models,
assume that the random term is independently but identically
distributed (IID) across alternatives. This means that, while values of
the random term are different between alternatives, we maintain the
assumption that their variance will be identical, while there is no
cross-correlation between them. As a consequence of this, alter-
natives have to be independent from irrelevant alternatives (IIA). This
means that the probability ratio of choosing iover jis unaffected by
the presence of other alternatives (This can be tested using a
Hausman test). If the IID or IIA assumptions are violated, a mixed logit
(random parameters) or nested logit model can be used instead of
multinomial logit. The distribution that is typically chosen for the
random component of utility is a type 1 extreme values (Gumbel)
distribution, which is appropriate for large numbers of IID random
values (Hensher et al., 2005). Armed with these assumptions, we can
now estimate the probability of an alternative ibeing chosen from
choice set C(McFadden, 1973):
PðijCÞ¼ exp mV
i
P
C
j¼1
exp mV
j
8i;j2C;j¼1; :::; i; :::C;i6¼ j(3)
where
m
is a scale factor, which may be normalised to one.
Then, we maintain that V
i
is composed of individual attributes,
accompanied by weights that determine their relative contribution to
utility:
V
i
¼b
0
ASC þb
1
x
1i
þb
2
x
2i
þb
3
x
3i
þ  þb
m
x
mi
(4)
where x
ni
is the nth attribute of I,
b
m
is the weight coefficient of
attribute x
mi
,
b
0
is the weight coefficient associated with an
alternative specific constant (ASC) and represents the average of
unobserved utility, i.e. utility that cannot be ascribed to other
parameters (Tardiff, 1978). An ASC equals 1 for a single alternative
and 0 for other alternatives, or the inverse. If there are j
alternatives, a maximum of j1 ASCs may be included. Social-
economic variables can either be interacted with attributes or an
ASC.
As we only know the relative utility obtained from choosing one
alternative over another, the weight of different attributes cannot be
interpreted in absolute terms, though we can estimate marginal rates
of substitution between attributes. If one of the attributes is a
reflection of cost (b
$
), then we can calculate tradeoffs as implicit
prices:
Implicit price ¼b
m
b
$
(5)
Ceteris paribus, the implicit price reflects marginal willingness to
pay (WTP) for improvement of an attribute, or in this case, an
ecosystem service.
In order to estimate significance of a model, the log likelihood ratio
is tested over a base model, assuming a chi-square distribution. This
base or constants-only model consists of alternative specific
constants only, ignoring attributes or socio-demographic variables.
The test shows whether the likelihood of the model, and thus
predictive ability of the model, has significantly improved by
introducing variables, and can also be used to evaluate whether
two models are significantly different. The test value Dcan be
calculated as follows:
D¼2ðLL
base model
LL
estimated model
ÞX
2
(6)
Similarly, one can calculate goodness of fit (expressed as
McFadden’s pseudo-R
2
,orr
2
):
r
2
¼LL
base model
LL
estimated model
LL
base model
(7)
For a full derivation of the multinomial logit model, see McFadden
(1973) or Louviere et al. (2000). For further details on using choice
experiments to value the environment see Boxall et al. (1996) and
Hanley et al. (1998).
References
Abaza, J., Rietbergen-McCracken, J., 1998. Environmental Valuation: A Worldwide
Compendium of Case Studies. United Nations Environment Program, New York.
Akin, D., 2004. Ancestral vigilance and the corrective conscience: kastom as culture
in a Melanesian society. Anthropological Theory 4, 299–324.
Alam, K., 2006. Valuing the environment in developing countries: problems and
potentials. Asia Pacific Journal on Environment and Development 13, 27–44.
Allen, M.G., Bourke, R.M., Evans, B.R., Iramu, E., Maemouri, R.K., Mullen, B.F., Pollard,
A.A., Wairiu, M., Watoto, C., Zotalis, S., 2006. Solomon Islands Smallholder
Agriculture Study. vol. 4, Provincial Reports. AusAID, Canberra, Australia.
Armitage, D., Marschke, M., Plummer, R., 2008. Adaptive co-management and the
paradox of learning. Global Environmental Change 18, 86–98.
Arrow, K., Solow, R., Portney, P., Leamer, E., Radner, R., Schuman, H., 1993. Report of
the NOAA panel on contingent valuation. Federal Register 58, 4602–4614.
Barkmann, J., Glenk, K., Handi, H., Sundawati, L., Witte, J.P., Marggraf, R., 2007.
Assessing economic preferences for biological diversity and ecosystem services
at the central sulawesi rainforest margin – a choice experiment approach. In:
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
15
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Tscharntke, T., Leuschner, C., Zel ler, M., Guhardja, E., Bidin, A. (Eds.), Stability
of Tropical Rainforest Margins: Linking Ecological, Economic and Social Con-
straints of Land Use and Conservation. Springer Verlag, Berlin, pp. 181–208.
Barkmann, J., Glenk, K., Keil, A., Leemhuis, C., Dietrich, N., Gerold, G., Marggraf, R.,
2008. Confronting unfamiliarity with ecosystem functions: the case for an
ecosystem service approach to environmental valuation with stated preference
methods. Ecological Economics 65, 48–62.
Bayliss-Smith, T., Hviding, E., Whitmore, T., 2003. Rainforest composition and
histories of human disturbance in Solomon Islands. Ambio 32, 346–352.
Berkes, F., Folke, C., 1994. Investing in cultural capital for sustainable use of natural
capital. In: Jansson, A.M., Hammer, M., Folfe, C., Costanza, R. (Eds.), Investing in
Natural Capital: The Ecological Economics Approach to Sustainability. Island
Press, Washington, DC, pp. 22–37.
Bourke, R.M., McGregor, A., Allen, M.G., Evans, B.R., Mullen, B.F., Pollard, A.A.,
Wairiu, M., Zotalis, S., 2006. Solomon Islands Smallholder Agriculture Study.
vol. 1, Main Findings and Recommendations. AusAID, Canberra.
Boxall, P.C., Adamowicz, W.L., Swait, J., Williams, M., Louviere, J., 1996. A compari-
son of stated preference methods for environmental valuation. Ecological
Economics 18, 243–253.
Carpenter, S.R., DeFries, R., Dietz, T., Mooney, H.A., Polasky, S., Reid, W.V., Scholes,
R.J., 2006. Millennium ecosystem assessment: research needs. Science 314,
257–258.
Carpenter, S.R., Mooney, H.A., Agard, J., Capistrano, D., DeFries, R.S., Diaz, S., Dietz, T.,
Duraiappah, A.K., Oteng-Yeboah, A., Pereira, H.M., Perrings, C., Reid, W.V.,
Sarukhan, J., Scholes, R.J., Whyte, A., 2009. Science for managing ecosystem
services: beyond the millennium ecosystem assessment. Proceedings of the
National Academy of Sciences of the United States of America 106, 1305–1312.
Chambers, R., 2002. Participatory Workshops. Earthscan, London.
Chiesura, A., de Groot, R., 2003. Critical natural capital: a socio-cultural perspective.
Ecological Economics 44, 219–231.
Christie, M., Fazey, I., Cooper, R., Hyde, T., Deri, A., Hughes, L., Bush, G., Brander, L.,
Nahman, A., De Lange, W., Reyers, B., 2008. An Evaluation of Economic and Non-
Economic Techniques for Assessing the Importance of Biodiversity to People in
Developing Countries. DEFRA, London.
Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K.,
Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., van den Belt, M., 1997.
The value of the world’s ecosystem services and natural capital. Nature 387,
253–260.
Daily, G.C., 1997. Introduction: what are ecosystem services. In: Daily, G.C. (Ed.),
Nature’s Services: Societal Dependence on Natural Ecosystems. Island Press,
Washington, DC, pp. 1–10.
Dietz, T., Fitzgerald, A., Shwom, R., 2005. Environmental values. Annual Review of
Environment and Resources 30, 335–372.
Domencich, T.A., McFadden, D., 1975. Urban Travel Demand: A Behavioral Analysis.
Elsevier, New York.
Donner, S.D., Skirving, W.J., Little, C.M., Oppenheimer, M., Hoegh-Guldberg, O.,
2005. Global assessment of coral bleaching and required rates of adaptation
under climate change. Global Change Biology 11, 2251–2265.
Ebi, K.L., Lewis, N.D., Corvalan, C., 2006. Climate variability and change and their
potential health effects in small island states: information for adaptation
planning in the health sector. Environmental Health Perspectives 114, 1957–
1963.
Fazey, I., Fischer, J., Lindenmayer, D.B., 2005. Who does all the research in conser-
vation biology? Biodiversity and Conservation 14, 917–934.
Fazey, I., Kesby, M., Evely, A., Latham, I., Wagatora, D., Hagasua, J.-E., Reed, M.S.,
Christie, M., 2010. A Three-Tiered Approach to Participatory Vulnerability
Assessment in the Solomon Islands. Global Environmental Change Human
and Policy Dimensions 20, 713–728.
Fazey, I., Latham, I., Hagasua, J.E., Wagatora, D., 2007. Livelihoods and Change in
Kahua, Solomon Islands. Aberystwyth University, Aberystwyth.
Folke, C., Hahn, T., Olsson, P., Norberg, J., 2005. Adaptive governance of social-
ecological systems. Annual Review of Environment and Resources 30, 441–473.
Garonna, I., Fazey, I., Brown, M.E., Pettorelli, N., 2009. Rapid primary productivity
changes in one of the last coastal rainforests: the case of Kahua, Solomon
Islands. Environmental Conservation 1–8.
Georgiou, S., Whittington, D., Pearce, D., Moran, D., 2006. Economic Values and the
Environment in the Developing World. Edward Elgar, Cheltenham.
Glenk, K., Barkmann, J., Marggraf, R., 2006. Unveiling Regional Preferences for
Biological Diversity in Central Sulawesi: A Choice Experiment Approach.
STORMA Discussion Paper Series, Sub-program A on Social and Economic
Dynamics in Rain Forest Margins 16.
Glenk, K., Barkmann, J., Marggraf, R., 2008. The value of ecosystem services and
agrobiodiversity in Central Sulawesi. In: Kontoleon, A., Pascual, U., Smale, M.
(Eds.), Agrobiodiversity, Conservation and Economic Development. Routledge,
London.
Gunawardena, M., Rowan, J.S., 2005. Economic valuation of a mangrove ecosystem
threatened by shrimp aquaculture in Sri Lanka. Environmental Management 36,
535–550.
Hanemann, W.M., 1984. Welfare evaluations in contingent valuation experiments
with discrete choice responses. American Journal of Agricultural Economics 66,
332–341.
Hanley, N., Wright, R., Adamowicz, V., 1998. Using choice experiments to value the
environment. Environmental and Resource Economics 11, 413–428.
Hansla, A., Gamble, A., Juliusson, A., Garling, T., 2008. The relationships between
awareness of consequences, environmental concern, and value orientations.
Journal of Environmental Psychology 28, 1–9.
Hearne, R.R., 1996. Economic valuation of use and non-use values of environmental
goods and services in developing countries. Project Appraisal 11, 255–260.
Hensher, D.A., Rose, J.M., Greene, W.H., 2005. Applied Choice Analysis. Cambridge
University Press, New York.
Hoehn, J.P., Randall, A., 2002. The effect of resource quality information on percep-
tions and contingent values. Resource and Energy Economics 24, 13–31.
Houde, N., 2007. The six faces of traditional ecological knowledge: challenges and
opportunities for Canadian co-management arrangements. Ecology and Society
12, 34.
Hviding, E., 2003. Contested rainforests, NGOs, and projects of desire in Solomon
Islands. International Social Science Journal 55, 539–553.
Jansen, T., Mullen, B.F., Pollard, A.A., Maemouri, R.K., Watoto, C., Iramu, E., 2006.
Solomon Islands Smallholder Agriculture Study. vol. 2, Subsistence Production,
Livestock and Social Analysis. AusAID, Canberra, Australia.
Kahneman, D., 1986. Comments. In: Cummings, R.G., Brookshire, D.S., Schulze,
W.D. (Eds.), Valuing Environmental Goods: An Assessment of the Contingent
Valuation Method. Rowan and AUanheld Press, Totowa, NI.
Kahneman, D., Knetsch, J.L., Thaler, R.H., 1991. Anomalies: the endowment effect,
loss aversion, and status quo bias. The Journal of Economic Perspectives 5, 193–
206.
Kaplowitz, M.D., 2001. Assessing mangrove products and services at the local level:
the use of focus groups and individual interviews. Landscape and Urban
Planning 56, 53–60.
Kaplowitz, M.D., Hoehn, J.P., 2001. Do focus groups and individual interviews reveal
the same information for natural resource valuation? Ecological Economics 36,
237–247.
Kenter, J.O., 2010. Values of the Environment in Kahua, the Solomon Islands and the
Importance of Deliberation in Valuing Ecosystem Services in Developing Coun-
tries. BSc Thesis. Aberystwyth University, Aberystwyth.
Kumar, S., 2002. Methods for Community Participation. Practical Action Publishing,
Rugby, Warwickshire.
Lamoreux, J.F., Morrison, J.C., Ricketts, T.H., Olson, D.M., Dinerstein, E., McKnight,
M.W., Shugart, H.H., 2006. Global tests of biodiversity concordance and the
importance of endemism. Nature 440, 212–214.
Lancaster, K., 1966. A new approach to consumer theory. Journal of Political
Economy 74, 132–157.
Lindstrom, L., White, G.M., 1994. Culture, Kastom, Tradition: Developing Cultural
Policy in Melanesia. Institute of Pacific Studies, University of the South Pacific.
Loh, J., Harmon, D., 2005. A global index of biocultural diversity. Ecological Indi-
cators 5, 231–241.
Louviere, J., Hensher, D.A., Swait, J., 2000. Stated Choice Methods – Analysis and
Application. Carmbridge University Press, Cambridge.
Lu, A.Y., Bishop, R.C., Welsh, M.P., 1996. Measuring the benefits of air quality
improvement in Taipei: a comparison of contingent valuation elicitation tech-
niques. In: Mendelsohn, R., Shaw, D. (Eds.), The Economics of Pollution Control
in the Asia Pacific. Edward Elgar, Cheltenham, pp. 182–198.
Macmillan, D., Lienhoop, N., Hanley, N., 2006. Contingent valuation: environmental
polling or preference engine? Ecological Economics 60, 299–307.
Maffi, L., 2005. Linguistic, cultural, and biological diversity. Annual Review of
Anthropology 34, 599–617.
Manski, C.F., 1977. Structure of random utility models. Theory and Decision 8, 229–
254.
McCauley, C., Mendes, S., 2006. Socio-Economic Assessment Report: Montserat
Centre Hills Project. Montserat Centre Hills Project, Montserat.
McFadden, D., 1973. Conditional logit analysis of qualitative choice behavior. In:
Zarembka, P. (Ed.), Frontiers in Econometrics. Academic Press, New York, pp.
105–142.
Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-Being:
Synthesis. Island Press, Washington, D.C.
Niemeyer, S., 2004. Deliberation in the wilderness: displacing symbolic politics.
Environmental Politics 13, 347–372.
Niemeyer, S., Spash, C.L., 2001. Environmental Valuation Analysis, Public Delibera-
tion, and Their Pragmatic Syntheses: A Critical Appraisal. Environment and
Planning C-Government and Policy 19, 567–585.
Pahl-Wostl, C., 2009. A conceptual framework for analysing adaptive capacity and
multi-level learning processes in resource governance regimes. Global Envi-
ronmental Change 19, 354–365.
Pretty, J.N., Guijt, I., Thompson, J., Scoones, I., 1995. Participatory Learning and
Action: A Trainer’s Guide. IIED, London.
Samuelson, W., Zeckhauser, R., 1988. Status quo bias in decision making. Journal of
risk and uncertainty 1, 7–59.
Schiffman, L.G., Kanuk, L.L., 1991. Consumer Behaviour. Prentice Hall International,
London.
Shyamsundar, P., Kramer, R.A., 1996. Tropical forest protection: an empirical
analysis of the costs borne by local people. Journal of Environmental Economics
and Management 31, 129–144.
Sillitoe, P., Bentley, J.W., Brokensha, D., Cleveland, D.A., Ellen, R., Ferradas, C.,
Forsyth,T.,DarrellAddison,P.,Stirrat,R.L.,Stone,M.P.,Warren,D.M.,Zuberi,
M.I., 1998. The development of indigenous knowledge: a new applied an-
thropology [and comments and reply]. Current Anthropology 39, 223–
252.
Snelgar, R.S., 2006. Egoistic, altruistic, and biospheric environmental concerns:
measurement and structure. Journal of Environmental Psychology 26, 87–99.
Spash, C.L., 1998. Environmental Values and Wetland Ecosystems: CVM, Ethics and
Attitudes. Cambridge Research for the Environment, Department of Land
Economy, University of Cambridge, Cambridge, U.K..
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
16
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001

Spash, C.L., 2002. Informing and forming preferences in environmental valuation:
coral reef biodiversity. Journal of Economic Psychology 23, 665–687.
Spash, C.L., 2007. Deliberative monetary valuation (DMV): issues in combining
economic and political processes to value environmental change. Ecological
Economics 63, 690–699.
Spash, C.L., 2008a. Deliberative monetary valuation and the evidence for a new
value theory. Land Economics 84, 469–488.
Spash, C.L., 2008b. How much is that ecosystem in the window? The one with the
bio-diverse trail. Environmental Values 17, 259–284.
Stern, P.C., 2000. New environmental theories: toward a coherent theory of en-
vironmentally significant behavior. Journal of Social Issues 56, 407–424.
Stern, P.C., Dietz, T., 1994. The value basis of environmental concern. Journal of
Social Issues 50, 65–84.
Stern, P.C., Dietz, T., Abel, T., Guagnanon, G.A., Kalof, L., 1999. A social psychological
theory of support for social movements: the case of environmentalism. Human
Ecology Review 6, 81–97.
Sutherland, W.J., Adams, W.M., Aronson, R.B., Aveling, R., Blackburn, T.M., Broad, S.,
Ceballos, G., Cote, I.M., Cowling, R.M., Da Fonseca, G.A.B., Dinerstein, E., Ferraro,
P.J., Fleishman, E., Gascon, C., Hunter, M., Hutton, J., Kareiva, P., Kuria, A.,
MacDonald, D.W., MacKinnon, K., Madgwick, F.J., Mascia, M.B., McNeely, J.,
Milner-Gulland, E.J., Moon, S., Morley, C.G., Nelson, S., Osborn, D., Pai, M.,
Parsons, E.C.M., Peck, L.S., Possingham, H., Prior, S.V., Pullin, A.S., Rands,
M.R.W., Ranganathan, J., Redford, K.H., Rodriguez, J.P., Seymour, F., Sobel, J.,
Sodhi, N.S., Stott, A., Vance-Borland, K., Watkinson, A.R., 2009. One hundred
questions of importance to the conservation of global biological diversity.
Conservation Biology 23, 557–567.
Tallis, H., Kareiva, P., Marvier, M., Chang, A., 2008. An ecosystem services framework
to support both practical conservation and economic development. Proceedings
of the National Academy of Sciences of the United States of America 105, 9457–
9464.
Tardiff, T.J., 1978. The Use of Alternative Specific Constants in Choice Modelling.
University of California Institute of Transportation Studies, Irvine, CA.
TEEB, 2010. The Economics of Ecosystems and Biodiversity: The Ecological and
Economic Foundations. Earthscan, London.
Urama, K.C., Hodge, I., 2006. Participatory environmental education and willingness
to pay for river basin management: empirical evidence from Nigeria. Land
Economics 82, 542–561.
URS Sustainable Development, 2006. National Forest Resource Assessment Update
2006. AusAid, Canberra, Australia.
Whilehead, J.C., Blomquist, G.C., 1991. Measuring contingent values for wetlands:
effect of information about related environmental goods. Water Resources
Research 27, 2523–2531.
Whittington, D., 1998. Administering contingent valuation surveys in developing
countries. World Development 26, 21–30.
Whittington, D., 2002. Improving the performance of contingent valuation surveys
in developing countries. Environmental and Resource Economics 22, 323–367.
Whittington, D., Smith, V.K., Okorafor, A., Okore, A., Liu, J.L., Mcphail, A., 1992. Giving
respondents time to think in contingent valuation studies – a developing-
country application. Journal of Environmental Economics and Management 22,
205–225.
Winpenny, J.T., 1991. Values for the Environment: A Guide to Economic Appraisal.
HMSO, London.
J.O. Kenter et al. / Global Environmental Change xxx (2011) xxx–xxx
17
G Model
JGEC-841; No. of Pages 17
Please cite this article in press as: Kenter, J.O., et al., The importance of deliberation in valuing ecosystem services in developing
countries—Evidence from the Solomon Islands. Global Environ. Change (2011), doi:10.1016/j.gloenvcha.2011.01.001