See discussions, stats, and author profiles for this publication at: [Link]

net/publication/23644290

Economic valuation of biodiversity: A comparative study

Article  in  Ecological Economics · September 2008

DOI: 10.1016/[Link].2008.03.003 · Source: RePEc

CITATIONS READS

195 3,732

3 authors:

Peter Nijkamp Paulo A.L.D. Nunes

Tinbergen Institute Amsterdam Food and Agriculture Organization of the United Nations
1,082 PUBLICATIONS   24,190 CITATIONS    219 PUBLICATIONS   5,324 CITATIONS   

SEE PROFILE SEE PROFILE

Gabriella Vindigni
University of Catania
39 PUBLICATIONS   1,036 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The Role of Small and Medium Towns (SMTs) in the Development Processes View project

AG2020 View project

All content following this page was uploaded by Paulo A.L.D. Nunes on 28 January 2019.

The user has requested enhancement of the downloaded file.

 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / e c o l e c o n

Economic valuation of biodiversity: A comparative study

Peter Nijkampa , Gabriella Vindignib , Paulo A.L.D. Nunesc,⁎

a
Free University, Department of Economics, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands
b
University of Catania, DISEAE, Via Santa Sofia, 98, 95123 Catania, Italy
c
University of Venice, Department of Economics and Foundazione Eni Enrico Mattei, Palazzo Querini Stampalia, 30122 Venice, Italy

AR TIC LE I N FO ABS TR ACT

Article history: In recent years, an intensive debate on the economic valuation of biodiversity has entered
Received 23 June 2007 the environmental-economics literature. The present paper seeks to offer first a critical
Received in revised form review of key concepts that are essential for a proper understanding of such evaluation
27 February 2008 issues. Particular attention is given here to various monetary valuation approaches and to
Accepted 5 March 2008 comparative (i.e., meta-analytical) methods from the perspective of conservation and
Available online 4 July 2008 sustainable use of biodiversity. Several illustrative examples are presented in order to
highlight the usefulness of the various approaches discussed. Next, an attempt is made to
Keywords: infer general findings and lessons from past applied research by means of meta-analysis. In
Meta-analysis this context, a multi-dimensional technique originating from the field of artificial
Biodiversity values intelligence is deployed. It allows us to identify the most important variables responsible
Policy formulation for changes in economic estimates of biodiversity.
Non-market valuation methods © 2008 Elsevier B.V. All rights reserved.
Market valuation methods
Benefit transfer

1. Introduction as they address similar issues. In this context, a multi-

dimensional technique originating from the field of artificial
Biodiversity requires our attention for two reasons. First, it intelligence is deployed. Estimation results allow us to identify
provides a wide range of indirect benefits to humans. Second, the most important variables responsible for changes in
human activities have been contributing to unprecedented economic estimates of biodiversity.
rates of biodiversity loss, which threaten the stability of The organization of the article is as follows. Section 2
ecosystems in terms of their provision of goods and services to discusses the challenge that comparative research is able to
humans. Consequently, in recent years many studies of put forward in the field of economic valuation of environ-
biodiversity and its loss have appeared. This article critically mental quality, in general, and biodiversity in particular.
evaluates the application of economic valuation methods for Section 3 offers a classification of biodiversity value, char-
the assessment of monetary values for biodiversity benefits. acterizing the approach adopted in the evaluation here
Particular attention is given to comparative (i.e. meta-analy- offered, i.e. economic approach. Section 4 critically evaluates
tical) methods as an alternative valuation approach to the well the use of the economic approach to the valuation of
known, and often costly, non-market methods. Finally, an biodiversity and its wide range of revealed and stated
attempt is made to infer general findings and lessons from preference methods. Section 5 shows an empirical attempt
available valuation studies by means of meta-analysis, as far to infer general findings and lessons from past applied

⁎ Corresponding author.
E-mail addresses: pnijkamp@[Link] (P. Nijkamp), vindigni@[Link] (G. Vindigni), pnunes@[Link],
[Link]@[Link] (P.A.L.D. Nunes).

0921-8009/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/[Link].2008.03.003
218 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

research by means of meta-analysis, discusses the range of rigorous valuation tools in order to cope with complicated
empirical findings and evaluates their basis against the earlier trade-offs in environmental policy analysis in the context of
presented framework. Section 6 concludes. sustainable development initiatives and emerging policies
which take explicitly account of the variety in the earth's
ecosystem. The current biodiversity conservation programmes
2. Biodiversity as a comparative in various countries require for their implementation con-
research challenge siderable financial expenditures, which have to be traded-off
against alternative uses. Although world-wide much progress
In recent years, the awareness has grown that biological has been made in identifying and prioritising such pro-
diversity is of critical importance for the stability of the earth's grammes, innovative valuation strategies are still needed to
ecosystem, as it forms the base for sustainable functions of generate additional information in order to support the actions
natural systems. In addition, it also offers a great potential for advocated in Agenda 21 of the 1992's Earth Summit United
human use (such as recreation or scientific research) (ten Kate Nations Conference in Rio, Brazil. Biodiversity conservation
and Laird, 2004). Biodiversity may reflect a great variety of programme funds have, in general, a rather poor underpinning
appearances depending on specific geophysical and climato- and are not based on solid and explicit economic choice
logical conditions. For example, European ecosystems encom- mechanisms. The reasons for this are manifold, but in general
pass more than 2500 habitat types and 215,000 species they are due to insufficient information on a given biodiversity
(Stanners and Bourdeau, 1995). Biodiversity has both quanti- issue as well as on undefined property rights, high transaction
tative and qualitative characteristics. It is generally accepted costs, divergence between private and social costs, inappropri-
that biodiversity cannot exclusively be expressed in numbers, ate economic instruments and bureaucratic inertia of relevant
as it also depends on the ecological structure of a whole area. It political institutions. Public authority choices concerning
is nowadays broadly recognized that human activities are biodiversity preservation programmes should ideally be
adversely affecting the earth's biological diversity, as a result based on sound economic principles and information, such
of prevailing production and consumption patterns and of as fair market prices, benefits of specific biodiversity policies
land use changes (cf. van Kooten et al., 2000). Consequently, and cost-opportunities of alternative decisions.
biodiversity tends to become a scarce economic good, for There is a growing awareness that biodiversity conserva-
which however a proper pricing system does not exist. In the tion programmes may generate many social benefits but
past years, research on the economic valuation of living sometimes at high costs, in particular in terms of management
natural resources and also of biodiversity has shown a and information gathering. Against this background, many
significant progress, but there is certainly not yet an estab- efficiency problems and fair public funds allocation issues
lished framework for valuing biological variety. Apart from the have arisen. Although general information about biodiversity
lack of a solid economic valuation mechanism for biological programmes is available through traditional policy channels, it
diversity, there is also a serious lack of reliable and up-to-date is challenging to allocate and manage biodiversity funds
information and monitoring systems with a sufficient geo- adequately from the perspective of the non-market value of
graphical detail on biodiversity. Clearly, studies on biodiver- environmental resources. In order to obtain a balanced trade-
sity require a pluridisciplinary approach (see also Cattizone, off between programme costs and benefits, it is necessary to
1999). Any economic approach to biodiversity is therefore, by optimise an efficient use of the information available (van den
definition, limited and partial in nature (see e.g. Pearce and Bergh et al., 1999). Fortunately, the number of studies
Moran, 1994; Barbier et al., 1995). Although various approaches concerning monetary biodiversity evaluation is quickly grow-
deploy contingent valuation methods, it ought to be recog- ing. Consequently, there is the need to deploy and develop
nized that this class of methods may certainly be helpful in adjusted methodologies and analysis instruments that can
assessing the use value of biodiversity, but has serious improve our understanding of economic biodiversity values
shortcomings in case of non-use values such as bequest and and, concurrently, that would allow for a more accurate
existence values (see also Desaigues and Ami, 2001). forecast of biodiversity values. Comparative analysis of many
The economic valuation of natural resources, in general, case studies is a key for enhancing an understanding.
and biodiversity, in particular, is among the most pressing and The ecological economics of biodiversity centres around the
challenging issues confronting today's environmental econo- crossroads of natural and human values of ecosystems. In
mists. Economists value biodiversity because valuation allows addition to an analysis of methodological complexities, there is
for a direct comparison with economic values of alternative also a need to draw policy lessons and general findings from
options, a corner stone for any cost-benefit analysis exercise. past applied research. The large number of applied economic
In addition, the monetary valuation of biodiversity allows valuation studies currently available has induced the search for
economists to perform environmental accounting, natural commonalities and contrasts in different empirical investiga-
resource damage assessment, and to carry out benefit assess- tions and has also induced the current popularity of meta-
ment. Valuation is also essential in the research of individual analysis and value transfer. In particular, in recent years we
consumer behaviour. It indicates the opinion of individual have seen a rising number of publications on the economic
consumers about certain biodiversity management objectives aspects of biodiversity, both theoretical and empirical. This
and identifies individual consumer motivations with respect prompts the intriguing question of whether more general
to biodiversity conservation. valuation conclusions might be inferred from a set of specific
Despite some flaws in economic valuation approaches to empirical investigations on closely related research themes or
biodiversity, there is a clear need to continue with developing issues in the area of biodiversity. Meta-analysis has originally
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 219

been developed in the context of the natural sciences (in methods difficult. Therefore, the use of specific procedures and
particular, in medical sciences) as a statistical tool for develop- of new instruments of analysis becomes increasingly necessary.
ing comparative studies and for creating synthetic knowledge The goal of this paper is to propose an operational methodolo-
from controlled experimentation studies (see Florax et al., 2002 gical framework to support decision-makers in synthesizing
for a detailed review). More recently, meta-analysis has been information concerning the economic value of biodiversity con-
diffused to the social sciences, including psychology, sociology servation studies. This study reviews economic and other
and economics. This has generated a new stream of quantita- values ascribed to biodiversity as well as related evaluation
tive research seeking for a synthesis of scientific results based methods. The focus is in particular on the issues associated with
on statistical applied analysis. The application of conventional a multi-dimensional evaluation of biodiversity, i.e. connections
statistical methods in meta-analysis is particularly appropriate, between economic and non-economic valuation. However, a
when the study focuses on identifying common trends and clear economic foundation in assessing the loss of biodiversity
relationships between the variables under investigation. The is still missing, but is certainly a potentially promising tool for
methods are also appropriate for the construction of synthetic coping with this loss. The case study approach developed here is
indicators and the determination of parameters or other constructed in order to illustrate some recent issues in meta-
common elements that can be described in quantitative terms analysis. It deals in particular with rough set theory based
(more details can inter alia be found in Hedges and Olkin, 1985; approaches to decision rules induction for examining different
Hunter and Schmidt, 1990; Wolf, 1986). price attributes to biodiversity conservation programmes. A
In recent years, meta-analysis has proven to be a research discussion on different approaches of decision rules induction
instrument of great potential for research synthesis of (i.e., conditional predictions) will be presented and extended to a
previously undertaken case studies. The main objective of small illustrative example of a set of studies characterized by the
modern meta-analysis is to synthesize in quantitative terms use of contingent valuation methods. The main aim is here to
the empirical results obtained from different studies on a pre- find rules that determine whether an object (a study) belongs to
determined field or topic of research. Although of great a particular subset called a decision class (i.e., a numerical range
interest and potentiality, this analysis instrument is of of monetary values of a given biodiversity conservation plan).
considerable methodological complexity. In particular, it is These technicalities will be further discussed in Section 2.
characterized by a high degree of transversality, both hor-
izontal and vertical. Transversality refers to both the hetero-
geneous nature of the studies and the heterogeneous nature 3. Meta-analytical methods for comparative
of the different empirical processes and/or the diverse biodiversity valuation
evaluation approaches of policies in the selected studies.
Therefore, this analysis instrument requires the identification 3.1. Prefatory remarks
and the selection of comparable empirical case studies as well
as the statistical application of comparative analyses by Meta-analysis refers to the use of quantitative methods – mainly
means of appropriate technical research methods (horizontal statistical – that can be deployed for the comparison or
transversality). Moreover, it also needs the adoption of a synthesis of outcomes from a set of empirical investigations
‘vertical’ perspective in terms of the description of the prob- on a common, or largely similar, issue (Stanley and Jarrell, 1989;
lem studied, the objective of the study, and the utilization of Cooper and Hedges, 1994). In contrast, value transfer aims to
the results obtained (Matarazzo and Nijkamp, 1997). develop a quantitative framework for the transferability of value
The intrinsic methodological complexity of meta-analysis (or benefit) estimates for policy decisions. Recently, many
becomes particularly obvious when transferring applications efforts are undertaken to carry out comparative studies using
from the field of natural sciences to the field of social sciences, a framework offered by meta-analysis and value transfer (for a
in particular in the area of environmental economics. In this survey, see van den Bergh et al., 1997). In this context, applied
field, the cause–effect relations are often characterized by environmental-economic research has in recent years proposed
uncertainty, since social phenomena are influenced by indi- to develop a test on value transfer by conducting two parallel
vidual behaviour (limited rationality, subjective preferences, case studies with the aim of deriving non-market environ-
rigidity in behaviour etc.) and by the adoption of specific mental values and comparing them with the obtained results
multi-faceted policies (Munda et al., 1995). (see Bergland et al., 1995; Kirchhoff et al., 1997; Bateman et al.,
Variations in comparable case studies can thus be attrib- 1995). Besides the case study approach used to obtain the
uted to the various dimensions of the studies, time and space required values for value transfer purposes, statistical methods
factors, the presence of a number of determining factors (i.e., based on meta-analysis can be used to obtain quasi-estimations
moderator variables), and finally by variation in time in regard of non-market values (see van den Bergh and Button, 1999).
to the policy effects to be expected. This is closely linked to the Thus, comparative studies seem to cover new ground in
fact that in the social sciences it is only possible to speak of economics, in general, and in environmental economics, in
“quasi-experiments” and “quasi-scientifically methods” (But- particular.
ton and Jongma, 1995). Meta-analysis clearly has many advantages in applied
For this reason, it is not always possible or methodologically quantitative research. It avoids the need to develop a costly
correct to apply standard statistical methods in meta-analysis: and new methodological basis for new case studies, as far as
the absence of general laws and the difficulty to generate they address similar issues related to past studies. It allows
random, independent and equal–dimensional samples render for the statistical identification of major driving factors in
the use of descriptive statistics and classical Bayesian inference causal relationships and can also act as a robustness check on
220 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

existing studies. In the context of value transfer, such research also may also concern numerical methods for dealing with
synthesis experiments also offer operational frameworks for categorical data. The format of study findings and the
making conditional forecasts. In the present section we will objective of synthesis in many areas of the social sciences,
offer a slightly more extensive review of meta-analytical including economics, are not exclusively based on an experi-
methods for research comparison. mental investigation of empirical phenomena (van den Bergh
et al., 1997). As a result, over the past decades several new
3.2. Statistical techniques non-statistical methods that synthesize study findings from
many areas of the social sciences have been developed.
Meta-analytical methods have gained much popularity, not Complementary non-statistical or non-parametric statistical
only in the context of research synthesis of previous case study techniques used for synthesis can be classified into main
surveys, but also as a tool for comparative case study research categories, in particular: (1) rough set analysis; (2) fuzzy set
and benefit transfer in case of uncertain outcomes in different analysis; and (3) content analysis. These tools appear to be
situations. In recent years, meta-analysis has seen many promising for drawing quantitative inferences, even on the
applications in environmental economics, particularly in basis of a collection of qualitative study findings (Hogenraad,
studies that have performed monetary valuations of environ- 1989). We will offer here a concise introduction and discussion
mental goods — see Table 1 for some examples. of these complementary meta-analytical methods.
A variety of statistical methods has been applied in order to
compare outcomes from different methods on a given issue 3.3.2. Rough set analysis
(van den Bergh et al., 1997; Florax et al., 2002). In general, Rough set analysis – developed by Pawlak (1982) – is the most
different outcomes from various studies can be explained by suitable technique for synthesis when the studies are to be
differences in the formulation of the research, the size and grouped and classified according to numerical characteristics
type of data analyzed, the statistical methods applied, the that are imprecisely measured. This analysis originates from
publication bias, and temporal and geographical characteris- artificial intelligence and aims to pinpoint data regularities
tics of the studies under consideration. The use of (quantitative that are not immediately evident; it searches for the possible
and qualitative) multi-criteria techniques is of great impor- existence of the principle of causality among data sets and
tance, when the objective of the synthesis is the qualitative attempts to eliminate irrelevant information. An important
comparison of studies and when the result of studies can be feature of this synthesis technique is that it does not
interpreted in terms of compliance with desirable ‘criteria’. necessarily require numerical information about the data
The application of these methods is especially important when being used, provided it is classified in distinct groups. In this
the studies do not provide one simple value per indicator or way it is able to synthesize a mixture of classified qualitative
when the analysis focuses on a number of indicators. In and quantitative data as well as to combine study findings that
principle, meta-analytical methods are able to encapsulate are subject to inconsistencies and inaccuracy (Pawlak, 1982;
these multi-dimensional components. Slowinski and Stefanowski, 1989). In fact, rough set theory also
has the advantage that it is able to create a ranking of actions in
3.3. Alternative techniques multi-attribute decision support processes (van den Bergh
et al., 1997). Given its features, it is clear that rough set analysis
3.3.1. Introduction is a readily applicable synthesis technique for decision-making
Meta-analysis comprises not only the class of quantitative- processes that deal with a large input of similar study findings.
statistical techniques for synthesizing research outcomes, but Several software packages are at present available.

Table 1 – Applications of meta-analysis

Subject area Studies

Urban pollution valuation Smith (1989), Smith and Huang (1993), Smith and Huang
(1995), Schwartz (1994), van den Bergh et al. (1997)
Recreation benefits Smith and Kaoru (1990a), Walsh et al. (1989a,b)
Recreational fishing Sturtevant et al. (1995)
Water quality Magnussen (1993)
Valuation of life estimates van den Bergh et al. (1997)
Contingent valuation versus revealed preference Carson et al. (1996)
Wetlands valuation Brouwer et al. (1999), Woodward and Wui (2001)
Noise nuisance Nelson (1980), Button (1995), Schipper (1996)
Travel congestion Waters (1993)
Visibility improvement Smith and Osborne (1996)
Transport issues Button (1995), Button and Kerr (1996)
Multiplier effects of tourism Nijkamp and Baaijens (2001)
Price elasticity of demand and travel cost Smith and Kaoru (1990b)
Price elasticity of gasoline demand Espey (1996)
Valuing morbidity Johnson et al., (1996)
Forest ecosystem services Markandya et al. (2008)

Source: Nunes et al. (2004), adapted.

 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 221

3.3.3. Fuzzy set analysis valuation studies (‘study site’) and to transfer their esti-
Fuzzy set analysis is applicable for the analysis of the same mates' values to the site where the new value estimate is
category of issues as rough set analysis. However, it is the most needed (‘policy site’) — see Brouwer et al. (1999); and Navrud
suitable technique for synthesis, whenever study findings con- and Bergland (2001). Value transfer brings up an important
tain a clear component of linguistic uncertainty in terms of im- research question: which lessons can be drawn from a
precise measurement (see, for example, Kacprzyk, 1978; Munda comparative analysis of monetary estimates derived from
et al., 1993). As Dubois and Prade (1992) have demonstrated the earlier empirical studies for an additional similar case not
distinction between fuzzy set analysis and rough set analysis is included in the meta-sample? The solution is to perform
more subtle than is commonly expected. For example, instead of essentially a meta-analysis, and to use the estimation
the discrete classes used in rough set analysis, fuzzy set analysis results for a prediction of the new or as yet unknown
employs a continuous classification scale. Fuzzy sets refer to empirical case. In economics, value transfer can be applied
linguistically defined variables that do not have an unambiguous across different sites – spatial value transfer – or, for one
measurement scale. Fuzzy variables can be categorized into specific site or valuation object over time — temporal
classes for which the boundaries are weakly demarcated, so that valuation transfer.
variables can belong to these classes with some degree of A major advantage of value transfer for policy guidance is
membership. A good example of the use of fuzzy set theory in that it ensures more comparability and consistency across
environmental quality valuation can be found in Munda (1995), different evaluation studies. It may also be helpful in an initial
who also developed the NAIADE software package. screening of a large number of public projects that cannot be
investigated in full detail. Furthermore, the outcome of a
3.3.4. Content analysis previous study may be used as a benchmark against which
Content analysis is a method for making inferences by results of studies can be evaluated. But the most important
identifying characteristics of text messages in a systematic advantage of benefit transfer is that it is a cost-effective way to
way in order to convert the text message into distinct classes make quantitative statements about phenomena that have
that can be studied with the use of quantitative methods. not been subject to previous analysis (see Johnson and Button,
Therefore, this synthesis technique is able to consider 1997).
simultaneously the dynamic context of cultural, social, In empirical research, two value transfer approaches are
economic and political factors. Content analysis is able to available: unit value transfer and function value transfer. Both
synthesize all kinds of verbal messages and texts by means of can be based on the principles of meta-analysis. The former
quantitative methods, concluding a mapping of specific types transfers mean monetary value estimates, for example, mean
of words in a text into fewer content categories. Besides willingness-to-pay, directly from the study site to the policy
coding, sophisticated computer software can be used, such as site, with possible income adjustments. The latter, instead of
the LISREL econometric package (see Weber, 1983). It is clear transferring individual willingness-to-pay estimates, explores
that a collection of published study findings from previously the use of more information, such as characteristics of the
undertaken case studies rather than a single text may serve as object of valuation and the subject who performs the
the basis for content analysis. Many interesting applications valuation exercise, and it is able to generate a benefit function
can be thought of such as newspapers, Internet sites, and that allows ‘prediction’ for the policy site. Examples of these
scientific magazines. Even when we limit ourselves to the approaches can be found in income elasticity studies, savings
social sciences, a wide range of application possibilities can be rate studies, consumer's surplus studies, accessibility studies,
observed. According to Weber (1983, p. 128), in general, “social value of time studies and evaluation studies on environmen-
scientists will be able to use computer-aided content analysis tal decay. The implicit assumption is then, that the degree of
with greater confidence to address a wide variety of theore- variation in estimated parameters is sufficiently small to be
tical problems involving the relationships among cultural, able to deploy valuations from the original data in a given case
social, economic and political change”. As Weber (1983) points study to assess corresponding parameters from other similar
out, content analysis is not only capable of considering mutual cases, usually in different contextual settings. Clearly, the
socio-economic relationships but also their dynamic aspects, more uniform the set of previous studies, the more likely the
which are highly useful in economic research. In particular, validity of the above implicit assumption (see Smith and
content analysis can be used to generate a flow of benefits Kaoru, 1990a; Smith, 1992).
over time and to compute net present values, which can serve In conclusion, next to quantitative-statistical techniques
as a value relating to a particular scenario of ecosystem such as standard meta-regression analysis, there is a variety
change or management. Studies that consider these features of complementary methods which also offer a great potential
of content analysis are, for example, Namenwirth (1969) and in research synthesis in the economic valuation of biodiver-
Rosengren (1981). These show the existence of a strong sity. The application of these research synthesis techniques is,
relationship over time between political statements and the however, anchored to a set of primary valuation studies
state of the economy. It is clear that content analysis is also a which, in turn, are exposed to some discussion. The central
powerful tool to extract quantitative or coded information problem refers to assessing the validity of value measures
from qualitative data input. obtained from any primary economic valuation method since
we observe an absence of an unambiguous clear criterion
3.3.5. Value transfer against which to compare those measures. In fact, many
In the context of ecosystem and biodiversity valuation, the environmental goods and services, including biodiversity, are
general idea is to explore the use of previous and original not directly observable and consequently different factors
222 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

may cause differences in biodiversity estimation. In Section 3 perceive things. Something has a value if it contributes to the
we will address this discussion in detail. welfare of someone. Therefore, goods do not have a value per
se, but their value is related to people's perceptions.

4. Caveats in the use of the economic 4.2. A classification of biodiversity benefits

approach for the valuation of biodiversity
Independently of the valuation approach engaged, value
4.1. Market and non-market approaches estimation is concerned with a change in people's welfare,
which stems from a change in the provision or enjoyment of
Economic, monetary valuation of biodiversity can proceed in the good. This change may relate to the quantity of the good or
different ways: using market price information and elicit its quality, or to whichever characteristic of it (or even of a
consumer's preferences using a wide range of non-market related good). These changes may be marginal or discrete.
valuation methods. Monetary indicators of biodiversity values Prices reflect marginal changes; many valuation methods are
are based on market price valuation mechanisms such as the used to account for discrete changes. Values may thus be
value of the financial revenues from tourism activities related expressed in marginal or discrete terms. Changes may of
to visits to natural areas and the value of contracts signed by course be positive or negative. When they refer to values, they
firms and governmental agencies, also known in the literature could reflect an increase or a decrease in welfare. Both can be
as bioprospecting contracts. Such contracts are characterized accounted for. Generally, it is expected that a loss of biological
by the search among the genetic codes contained in living diversity results in a decrease in welfare, and therefore in a
organisms, for the development of chemical compounds of negative value expressed in monetary terms. Similar patterns
commercial value, i.e. market priced value (Simpson et al., can also be observed for landscapes, although it is not as
1996). This is dominated by pharmaceutical research since straightforward, and it may vary more from person to person.
most prescribed drugs are derived, or patented after natural Once the physical change is identified, the value should reflect
sources. In this context, the marginal value of such an input, the welfare change related to the individuals whose welfare
often translated in terms of genetic information for medicinal has been affected by it, or the average welfare change of the
purposes, is measured by its contribution to the improvement individuals of the population (Riera, 2001).
of health care. Recent registrations and applications of It should be noted that biological diversity – and in
bioprospecting contracts and agreements between states and particular the preservation of threatened species – can affect
pharmaceutical industries represent important benchmarks of the welfare of many people, even living far away from the site
monetary indicators for these types of biodiversity values. The concerned. In other words, people may derive satisfaction out
most noted of these agreements is the pioneering venture of knowing that there is an improvement in biodiversity for
between Merck and Co., the world's largest pharmaceutical present and future generations, even if they would not benefit
firm, and “Instituto National de Biodiversidad” (INBio) in Costa from it directly. Therefore, the relevant population is often not
Rica. At the moment of the signature of the contract, in 1991, local, but global. These welfare gains are usually known as
Merck paid Costa Rica about 1 million dollars and agreed to pay passive or non-use values. The consideration of both use and
royalties whenever a new commercial product was explored. non-use values introduces the notion of total economic value. In
Since then, INBio has signed contracts on the supply of genetic the literature, there is a wide range of systematization of the
resources with Bristol-Myers Squibb, other companies and total economic value of the environment.1 They differ both in
non-profit organizations (ten Kate and Laird, 1999; INBio, 2001). terms of the services included and in the terminological
Another illustration of the market value of genetic diversity definition. The total economic value (TEV) of a species or habitat
refers to the commercial agreement signed in 1997 between is constituted by a combination of the use value and non-use
Diversa, a San Diego based biotechnology, and the US National value:
Park Service. Diversa paid $175,000 for the right to conduct
research on heat-resistant microorganisms found in hot - The use value is a value related to the present or future use
springs in Yellowstone National Park (Sonner, 1998; Macilwain, of a particular habitat by individuals. It can be subdivided
1998). More recently, a Brazilian company, Extracta, signed a into direct use values and indirect use values. Direct use values
$3.2 million agreement with Glaxo Wellcome, the world's are derived from the actual use of a resource either in a
second largest pharmaceutical company, to screen 30,000 consumptive way or a non-consumptive way (e.g., timber
samples of compounds of plant, fungus and bacterial origin in forest, recreation, fishing); indirect use values refer to the
from several regions in the country (Bonalume and Dickson, benefits derived from ecosystem functions (e.g., watershed
1999). Despite the fact that these agreements show a positive protection or carbon sequestration by forests);
economic value of genetic diversity, concern remains with - The non-use values are associated with the benefits derived
respect to the fairness of such deals. Indeed, some environ- simply from the knowledge that a natural resource – such
mental groups have been very critical, claiming that these are
unequivocally “biopiracy” actions (see RAFI, 2001). In the
1
A frequent question concerns how TEV is related to the notion
absence of market prices for biodiversity values – which is
of intrinsic value. Intrinsic value is often regarded as a value that
commonly the case – different methods have been developed
resides in the assets in question, especially in the environmental
to derive consumers' preferences — these will be discussed in assets, but that is independent from human preferences. Since by
detail in Section 4. However, value is a cultural and psycholo- definition TEV relates to preferences of individual human beings,
gical concept, and is related to the question how human beings it cannot encompass an intrinsic value (Pearce and Moran, 1994).
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 223

as species or habitat – is maintained. By definition, such a economic relevance of non-market ecological functions. Farn-
value is not associated with the use of the resource or the worth et al. (1981) identified a value category, inherent value,
tangible benefits deriving from its use. It can be subdivided which was defined as “values that support other values” in
into two parts that overlap in relation to its definition. First, ecological systems. It includes natural processes of selection
there are existence values which are not connected to the and evolution and life support functions of ecosystems in an
real or potential use of the good, but reflect a value that is all encompassing perspective.
inherent in the fact that it will continue to exist indepen- Another category of value is the contributory value which
dently from any possible present or future use of indivi- focuses on the fact that species can only survive in interactive
duals. Secondly, bequest values are associated with the relationships and therefore each species contributes to the
benefits of the individuals derived from the awareness survival of other species (Norton, 1986). Wood (1977) illustrates
that future generations may benefit from the use of the importance of the contributory value in the example of the
the resource. These can be altruistic values, when e.g. the productive use of the wild species for the preservation of the
resource in question should in principle be available to resistance of cultivated plants. Because of their limited genetic
other individuals in the current generation. diversity, cultivated plants can only perform minor adapta-
tions to changes in environmental conditions. Wild species on
A separate category is made up by option values, i.e. values the other hand possess higher adaptability to environmental
attributed by individuals from the knowledge that a resource conditions because of their higher genetic diversity. Thus, the
will be available for future use. Thus it can be considered like wild species forms the basis for preserving or improving the
an assurance that a resource will be able to supply benefits in a resistance of cultivated plants against disease or pest, by
possible future, but not for present use. The quasi-option value, cross-breeding with wild species.
which is sometimes classified as a non-use value, represents Finally, besides ecological values, psychological values are
the value derived from the preservation of the future potential gaining attention in the field of environmental psychology
use of the resource, given some expectation of increase of (Boerwinkel, 1992; Nunes 2002; Nunes and Onofri, 2004). In this
knowledge. The quasi-option value is important when the field the concept of value is interpreted as a reason why people
decisions on consumption are characterized by a high feel that certain things are important (de Boer and Hissche-
reversibility. möller, 1998). Examples of such values are justice, safety, and
However, any economic valuation of biodiversity according beauty. While ecological values are meant to determine the
to this definition is based on an instrumental perspective on well functioning of a system, the psychological values are
the value of biodiversity. This means that the value of used to determine the perceived quality or the perception of
biodiversity is regarded as the result of the integration nature. As such, the differences between ecological values and
between humans and the object of valuation, which is the the psychological values pertain to what is being valued: the
change in biodiversity. Economic valuation in biodiversity quality of the system versus how the system is perceived.
changes is based on a reductionist approach of the TEV; it is However, empirical findings show that the psychological
regarded as a result of the aggregation between various use value has a strong positive relation with the total economic
and non-use values, reflecting a variety of human motiva- value, because they both measure social preferences.
tions, as well as the aggregation of local values to attain a
global value, i.e. a bottom-up approach (Nunes and van den
Bergh, 2001). 5. Alternative economic valuation methods

4.3. A critical evaluation 5.1. Revealed preferences techniques

This classification of the TEV is not always used in straightfor- Revealed preferences techniques seek to elicit preferences
ward way and free of significant criticism. While its success is from actual, observed market-based information. Preferences
based on several legal cases (e.g., the Exxon Valdez Alaskan oil for environmental goods are usually revealed indirectly, when
spill disaster), this model has its criticism among those who an individual purchases a market good to which the environ-
raise the question about its ability to capture actual values of mental good is related in some way. They are all indirect,
natural resources. Nevertheless, an understanding of the wide because they do not rely on people's direct answers to
range of values attributed to biodiversity forms a basis for questions on how much they may be willing to pay (or accept)
keeping policy-makers informed on their choices on its for an environmental quality change. Thus the emphases of
conservation and sustainable use (Spash, 2001). According to these techniques are mainly on their contribution to valuing
Brown and Moran (1994), the economic valuation of biodiver- biological resources. The values obtained could be considered
sity is required for the purpose of placing a “common concern” sufficient for cost-benefit purposes, but they will rarely reflect
in the context of the management of natural resources. Thus, biological health. As such these techniques provide only a
the above classification is often complemented with other lower bound estimate of the value of a particular biological
categories of values which consider the ecological–functional resource. The techniques included in this group are in
importance of biodiversity in natural systems (see for a review particular the travel cost method, or hedonic price and wage
Fromm, 2000). Early thoughts on these complementary techniques on adverting behaviour. In most cases they only
relationships in the literature can be found in an analysis capture use values, leaving non-use values out of considera-
carried out by Farnworth et al. (1981) aimed at distinguishing tion. This is not the case, though, with simulated market
the manifold values of the ecosystems and to point out the methods.
224 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

Fig. 1 – Methodologies for economic valuation of biodiversity.

The common underlying feature is a relationship between method is currently one of the most often used techniques
a market good and the environmental commodity. For for the valuation of environmental goods. This is partly due to
example, when using the travel cost method, researchers CV features that constitute important advantages over
estimate the economic value of recreational sites by looking at revealed preference methods. First, the CV method is the
the generalized travel costs of visiting these sites (Bockstael only valuation technique that is capable of shedding light on
et al., 1991). Conversely, practitioners of the hedonic price the monetary valuation of the non-use values, which typically
method estimate the economic value of an environmental leave no ‘behavioural market trace’. Ignoring such values will
commodity, say, clean air, by studying the relation between lead to a systematic bias in the estimation – essentially an
house prices and air quality (Palmquist, 1991). The averting underestimation – of the total benefits of biodiversity. Second,
behaviour or production cost function methods is character- CV allows environmental changes to be valued even if they
ized by exploring the relationship of the environmental have not yet occurred (i.e., ex ante valuation). Therefore, CV
commodity through a generalized cost function (Cropper and offers a greater potential scope and flexibility than revealed
Freeman, 1991). For instance, improvement of air quality can preference methods. It allows the specification of hypothetical
be assessed on the basis of expenditures made to avert or policy scenarios or states of nature that lie outside the current
mitigate the adverse effects of air pollution. Avoided cost or past institutional arrangements or levels of provision.
damage costs, preventive expenditures, repair costs (or Third, CV allows to enrich the information base by submitting
restoration), compensation costs, replacement costs, and the process of value formation to public discussion, and hence
relocation costs are specific instances of this method. Finally, it is recognized as “an effective tool for policy decision-
the production factor method estimates the economic value of making” (Sen, 1995).
an environmental commodity through the input–output rela- Fig. 1 and Table 2 provide a summary of monetary
tionship of such a commodity in a production function. For valuation techniques and their suitability to measure different
example, the economic value of a cleaner soil is related to the components of the TEV of biodiversity. Not all valuation
value of the increased agricultural output through a dose methods are usually suitable to measure non-use values. For
response method (Nunes et al., 2004). instance, if a market would exist, it may be that it fails to
capture non-use values. Sometimes, the ability to capture
5.2. Stated preferences techniques passive use values could be a decisive criterion for choosing
for a specific method selection process.
Stated preferences (SP) techniques are based on the simula-
tion of the market, and thus on ‘prices observed’ for the good 5.3. Application of the valuation techniques to biodiversity
to be valued. Results are achieved through a questionnaire to
be filled out by the population, or a sample of it. In simulated Extending these procedures to biological biodiversity is
market conditions, the supply side is represented by the complex. Indeed, the evaluation for biodiversity is perhaps
interviewer, who typically offers to provide a given amount of the most challenging issue in the context of economic
units of the good at a given price. The respondent, who either valuation. It is worth noting that all the techniques briefly
accepts or rejects the offer, represents the demand side. One of described above have been applied to biodiversity valuation.
the most crucial issues in this kind of method is to be precise Each one of them has some advantages over the overs, but
in the description of the market, and yet simple and clear also some disadvantages. The use of one or another depends
enough for people to understand it. This is important, because mainly on the purpose of the valuation exercise and the
biological and landscape diversity are among the goods for availability of data and resources.
which it is difficult to simulate a clear, credible, precise and So far, the most popular valuation methodology has been
understandable market in a poll process. They are based on the family of CVM and SP methods. The reasons for the
collecting data by means of questionnaires. The best known momentum of SP methods are diverse. It has a format that
method is the contingent valuation methodology (Mitchell respondents tend to find comfortable, thus reducing the
and Carson, 1989). Indeed, the contingent valuation (CV) proportion of no-answers and protest-answers. SP methods
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 225

Table 2 – Review of economic methods for biodiversity conservation

Method Pros Cons

Travel cost Use of real market data Can estimate use values only
method May have substantial data requirements
Requires estimates of value of travel/leisure time
Cannot predict the changes in use values due to environmental changes
without prior information
Random Estimates recreational use value of (i) changing Can estimate use values only
utility model environmental quality of site attributes and (ii) site in May have substantial data requirements
total Requires estimates of value of travel/leisure time
Problems arise with multi-purpose trips
Cannot predict the changes in use values due to environmental changes
without prior information
Can be hard to handle participation decisions (i.e. whether to make the
visit or not)
Hedonic Use of real market data Can estimate use values only
pricing Requires extensive house market data
method Cannot predict the changes in use values due to environmental changes
without prior information
Current evidence suggests it is not suitable for use in benefits transfer
Avertive Modest data requirements Can estimate use values only problems arise when (i) individuals make
expenditure multiple averting expenditures, (ii) there are secondary benefits of an
method averting expenditure and (iii) averting behaviour is not a continuous
decision but a discrete one (e.g., double glazing is either purchased or
not)
Use of real market data Cannot predict the changes in use values due to environmental changes
without prior information
Contingent Can estimate both use and non-use values Relatively expensive
valuation Suitable for valuing environmental changes Complex and multi-dimensional scenarios may be too much of a
irrespective of whether or not they have a precedence cognitive burden for respondents
Completed surveys give full profile of target The concept of diversity may similarly be difficult to put across to the
population respondents
Choice Can estimate both use and non-use values Not yet as widely tested as CV
modelling Suitable for valuing environmental changes Some techniques are not based on economic theory
irrespective of whether or not they have a precedence
Completed surveys give full profile of target The concept of ‘diversity’ may be difficult to put across to the
population respondents

can cope with valuing different attributes of a forest, like hensively enough in order to convey detailed information on
biological and landscape diversity aspects, in an integrated changes in ecosystem life support functions and processes as
manner, therefore being more informative. SP methods tend related to biodiversity changes so that the latter are not
to cope better with the so-called embedding problem (valua- regarded by respondents as too cumbersome. Such informa-
tion being rather insensitive to the scale of the physical tion is crucial to obtain a practical, reliable and effective
change) in as far as the respondent gets a richer perspective of questionnaire.
the scale of the changes proposed. Both CVM and SP ‘design’ There tends to be an inverse relationship between famil-
exactly the market so as to value the good of interest, whereas iarity with the good and the ability of the respondents to
with other methods it is often difficult to isolate the value of answer meaningfully. The biodiversity-related goods tend to
the good from other closely related goods. On the other hand, be very unfamiliar for a market situation. Therefore, the use of
expressing biodiversity changes in simple, accurate and CVM and SP requires state-of-the-art practice to overcome this
understandable terms in a questionnaire can prove to be a and other potential related problems. In general, the more
challenging task. specific the change in biodiversity is, the more reliable are the
Table 2 shows that certain valuation methods are more values obtained by all methods. This is especially the case
appropriate than others to address certain types of biodiver- with CVM and SP methods. Other techniques tend to be more
sity value. For example, revealed preference methods can only suitable for ex post valuation, since they rely on existing
be used for a limited number of biodiversity value categories, markets, whereas CVM and SP tend to be more adequate for
as they do not allow for a monetary assessment of non-use valuing changes ex ante. They can also be used in ex post
values. On the contrary, the contingent valuation method is in valuation, but there might be a lack of incentives to answer
principle applicable to a multiplicity of biodiversity value (see for details Nunes et al., 2004).
categories. However, one needs to recognize that this method In summary, even though estimating the economic values
will fail for those biodiversity value categories that the general of changes in biological and landscape diversity of forests is
public is not informed about nor has experience with. In this not a straightforward task, the tools developed by environ-
respect a questionnaire should be designed that is compre- mental economics makes it possible and, overall, fairly
226 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

Table 3 – List of the valuation case studies

Biodiversity and habitat values

Type of good Range of the monetary value estimate Method Year of study
Country

Biodiversity
UK 8.01–15.42 once-off payment/person CV–OE 1993
UK 46.99–62.26/household/year CV–OE 1993
Norway NOK 194 annual payment/year CV–DC 1991
Germany DM 16.1/month CV 1990⁎
UK 75/household CV–OE 1994
UK 308/household CV–DC 1994
Wildlife
UK 13.83–61.74/household/year CV–PC 1994
UK 0.98–3.12/visitor/year CV–OE 1986
UK 13–65 ha/year TCM 1986
UK 25 ha/year/person CV–OE 1985
UK 0.6–1.7 visit/person TCM 1985
UK 1.02–2.3 visit/person (full travel cost) TCM 1986
UK 1.18–2.53/adult CV 1988
UK 1.99–2.60/person ZTCM 1988
UK 16.8 once-off payment/person CV 1990⁎
National parks and nature reserves
Hungary 6/visitor CV–PC 1996
UK 24/household CV 1990
UK 4.54/person CV 1985
UK 0.82/person CV 1986
Watercourses
Norway $50–$100/person/year CV 1990
UK 0.75–0.95/adult visitor CV 1991
UK 0.36/visitor CV 1989
UK 13.90–16.20/household/year CV–OE 1988
UK 13.59–5.56/person/year CV 1987
UK 546–582 once-off payment/household CV–OE 1987
UK 12.08/person/year CV 1987
UK 3–5% increase in property sale price HPM 1989
UK 9.2–11.2/person/visit ITCM 1988
UK 4.9% increase in property sale price HPM 1990⁎
UK 0.51–3/visit TCM 1989⁎
Landscape
Austria ATS 9.2/visitor/day CV–OE 1991
Sweden SEK 750/person/year CV–PC 1991
UK 49–55/household/year CV–DC 1994
Netherlands NLG 55/household/year CV–PC 1993
Netherlands NLG 80/household/year CV–OE 1994
Endangered Species
Norway NOK 1700–NOK 2750/person/year CV 1991
Norway $15/person/year CV 1990
Sweden $7/person/year CV 1990
Sweden SEK 85/person/year CV 1991
UK 2.94/year CV 1993⁎
Sweden SEK 406/person CV–DC 1993
Wetlands
UK 67/household/year CV–OE 1991
UK 75/household/year CV–IB 1991
Austria ATS 329.25/Austrian/year CV–OE 1993
UK 21.75/household/year CV–OE 1991
UK 76.74/visitor/year CV–OE 1991
UK 83.67/visitor/year CV–IB 1991
Woodlands
UK 9.94/household/year CV–OE 1991
UK 18.5–20.7/household/year CR 1995
Netherlands NLG 22.83/household/month – 1987
Sweden SEK 95/person/year – 1991
Sweden SEK 1014/household/year CV–OE 1988
Sweden $3–$4/person/year CV–OE 1988
Sweden $5–$8/person/year CV 1990
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 227

Table 3 (continued)
Biodiversity and habitat values

Type of good Range of the monetary value estimate Method Year of study
Country

Woodlands
Norway $13–$18/person/year CV 1990
UK 1.21–7.09/person/year CV–OE 1991
UK 9.73/visitor/year CV 1990
UK 0.53/visitor CV 1988
UK 0.33/visitor CV 1987
UK 1.3–3.3/visit ZTCM 1988⁎
UK 1.25/visit CV–PC 1987
UK 15.13/visitor/year TCM 1988⁎
UK 1/visitor CV 1986⁎
UK 14.6–24.5/visitor/year TCM 1986⁎
UK 7.1% increase in property sale price HPM 1990
UK 43 increase in property sale price HPM 1988
UK 0.06–0.96/visitor TCM 1988
UK 0.43–0.72/person CV 1988
UK 1.95/visitor TCM 1988
UK 0.53/visitor CV–OE 1988
UK 12.55/year CV 1994⁎
UK 3.51/year CV 1994⁎
UK 1.82–2.78/visit CV–OE 1990
UK 20.6–30.59/person CV 1990⁎
UK 0.33–0.93/visitor/visit CV–OE 1991

Source: RIVM (2000).

Note: CV = Contingent valuation; CR = contingent ranking; OE = open ended; PC = payment card; DC = dichotomous choice; TCM = travel cost
method; ZTCM = zone travel cost method; – = valuation method unspecified, and ⁎ = base year unspecified and assumed to be the year before
presentation of the paper.

reliable. Provided, of course, that the methods are applied • the country or region concerned;
according to the state-of-the-art knowledge. • the evaluation method deployed (e.g., contingent valuation,
hedonic prize);
• the monetary value of the asset (or a range), including
6. Biodiversity values in plural measurement units and scales used.

6.1. Valuation data set The monetary valuation figures have been presented as
standardized willingness-to-pay figures in the study by ten
Comparative analysis, such as meta-analysis, serves to Brink et al. (2000), and these will be used by us — see Table 4 for a
provide a framework for quantitative research synthesis. In summary. This database contains both numerical information
particular, it may also serve as a quantitative framework for (such as willingness-to-pay) and alpha-numerical, linguistic
the comparative study of previously undertaken studies that and categorical information (e.g., country, year of study, etc.).
have generated different values. As an illustration of the This makes the application of standard statistical tools rather
potential of this approach, a meta-analysis study will be here problematic. Nevertheless, even a linguistic information base
presented based on a data set originating from the Dutch may incorporate a hidden structure in terms of associations
National Institute of Public Health and the Environment between patterns, or the frequency of occurrence of a given
(RIVM) (ten Brink et al., 2000) – Table 3 presents a full list of phenomenon (or qualitative characteristic or parameter). In this
the case studies. This data set contains a set of 75 distinct context, we may resort to qualitative pattern recognition
empirical case studies on the valuation of different aspects of methods, elucidated inter alia in the artificial intelligence
biodiversity and different kinds of habitat, ranging from literature. There is a wide variety of such methods, such as
wildlife and endangered species preservation to the protection computational neural networks, genetic algorithms, fuzzy and
of national parks and nature areas (ten Brink et al., 2000) rough set methods, decision tree induction methods, etc. An
provide a full list of the case studies and details. The database interesting recently developed algorithm in this framework is
includes the following items: the a priori algorithm, which is able to identify association rules
among qualitative data (Agrawal et al., 1996). This meta-
• type of environmental asset (e.g., biodiversity, wildlife, analytical method is used by us and will be concisely described
landscape, wetlands, etc.); in Subsection 5.2. Subsection 5.3 will then present the quanti-
• study characteristics; tative results of our comparative study.
228 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

Table 4 – Mean willingness-to-pay for biodiversity and habitat services and values
Type of good study WTP/person/year (in €, 2006) Countries of the study

Biodiversity preservation 28.66 UK, Norway, Germany

Wildlife preservation 1.8 UK
National parks and nature reserves 8.7 UK, Hungary
Wetlands 35.0 UK, Austria
Watercourses 27.2 UK, Norway
Landscape 57.5 UK, Netherlands, Austria, Sweden
Endangered species protection 120.9 UK, Sweden, Norway
Woodlands 18.8 UK, Netherlands, Sweden, Norway

6.2. Assessment of values ‘by association’ variables: viz. on the basis of valuation method and the good
under valuation. The second set is characterized by the price
Nowadays, various methods for data set characterization are variable. In addition, we use the information about the
rather well developed. In particular, association rules provide number of studies that support an association rule to be
a competitive advantage by using repeated interactions identified and its relative share in terms of frequency, which
among factors within the data, the a priori algorithm, first incorporates the weight of the association rule on the
introduced by Agrawal et al. (1993), is one of the most popular database (see Table 5). This table offers also the relevant
methods. It identifies combinations of attribute values or information concerning the strength of the association rule
items that occur with a higher frequency than might be within the database used in our comparative analysis.
expected, if the values or items where considered indepen-
dent of one another. In such a context, a relation R contains n- 6.3. Empirical results
tuples over a set of Boolean attributes A1, A2,… An. Let I = {Ai1,
Ai2,… Ain} and J = {Aj1, Aj2,… Ajn} be two sets of attributes. Then The use of the a priori algorithm leads to interesting empirical
I ⇒ J is an association rule, if the following two conditions are findings. Table 5 shows that that there are 18 possible
satisfied: the support for the set I ∪ J appears in at least an s- association rules, which correspond to the most frequently
fraction of the tuples; and a confidence amongst the tuples appearing regularities in the database. One particularly
shows up if at least a c-fraction also has J appearing in them. interesting association rule that emerges from the database
To some extent, the relative popularity of this method can links (2) the good under valuation to (3) the price, which is
be attributed to the simplicity of the problem statement and interpreted in terms of WTP. For example, if the good under
its wide applicability in identifying hidden patterns in a data consideration is ‘woodland habitat’, the corresponding asso-
set. However, its success has also much to do with the ciation rule (see first row) shows that the woodland monetary
availability of an efficient algorithm (a priori). valuation falls between € 9.4 and € 20.1. This price range is
The analysis aims to identify all valid association rules supported by 30 cases, which corresponds to 38% of all case
within the database for the valuation of different biodiversity studies in the database. The empirical magnitude of this
indicators and different kinds of habitats. The first step in the association rule appears to be very high: it corresponds to a
a priori analysis is to split the database into two distinct sets of confidence level of 100% (see the last column). In other words,

Table 5 – Empirical findings for alternative association rules

Association rule Share of Number Confidence
studies (%) of studies level (%)
Good, method (LHS) if Price in € (RHS) then

(1) Good = woodlands 9.4 b price ≤ 20.1 38.0 30 100.0

(2) Good = watercourses 20.1 b price ≤ 31.1 13.9 11 91.7
(3) Good = wildlife Price ≤ 9.4 11.4 9 100.0
(4) Good = endangered species Price N 38.0 8.9 7 100.0
(5) Good = wetlands 31.1 b price ≤ 38.0 7.6 6 100.0
(6) Good = biodiversity 20.1 b price ≤ 31.1 7.6 6 100.0
(7) Good = landscape Price N 38.0 6.3 5 100.0
(8) Good = national parks and nature reserves Price ≤ 9.4 5.1 4 100.0
(9) Method = CV 9.4 b price ≤ 20.1 16.5 13 48.1
(10) Method = CV–OE 31.1 b price ≤ 38.0 5.1 4 20.0
(11) Method = CV–OE 20.1 b price ≤ 31.1 6.3 5 25.0
(12) Method = TCM 9.4 b price ≤ 20.1 5.1 4 57.1
(13) Good = woodlands and method = CV 9.4 b price ≤ 20.1 16.5 13 100.0
(14) Good = woodlands and method = TCM 9.4 b price ≤ 20.1 5.1 4 100.0
(15) Good = woodlands and method = CV–OE 9.4 b price ≤ 20.1 8.9 7 100.0
(16) Good = wetlands and method = CV–OE 31.1 b price ≤ 38.0 5.1 4 100.0
(17) Good = watercourses and method = CV 20.1 b price ≤ 31.1 6.3 5 83.3
(18) Good = endangered and method = CV Price N 38.0 5.1 4 100.0
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 229

all the ‘woodland habitat’ valuation cases support this assessment of an environmental system, but rather system
association rule. If the good under consideration is ‘biodiver- changes. The goal is then to assess the human welfare
sity’, this price-good association rule (see row 6) shows us that significance of biodiversity change under consideration,
the respective price ranges between € 20.1 and € 31.1, which is through the determination of the changes in provision of
true for all biodiversity case studies presented in the database. biodiversity-related goods and services and consequent
However, the empirical significance of this rule is rather low, impacts on the well-being of humans who enjoy both use or
since it is only valid for 7.6% of the case studies in the non-use benefits from such a provision. Different instruments
database, i.e. only 7.6% of the sample of case studies was are available to assess the economic value of biodiversity. The
concerned with biodiversity. choice is not always evident. Survey valuation studies have
Another association rule that emerges from the database often been used, because the use of revealed preference
links (a) the method of valuation to (c) the price. This rule methods leaves out important biodiversity value types,
shows (see row 9) that if the valuation method under notably non-use values. Alternatively, researchers can com-
consideration is contingent valuation the respective monetary bine valuation techniques. Special attention however, should
estimates range between € 9.4 and € 20.1. This is true for 13 of then be given to value aggregation across the resulting values
all the 27 case studies that used CV as the selected valuation so as to avoid double counting. Meta-analysis appears to be a
method. Therefore, the empirical magnitude of this associa- fruitful instrument in this context.
tion rule corresponds to a confidence level with a strength of There is a clear need to obtain information about the cause,
48.1%. In other words, about half of the CV valuation cases type, and persistence of stress on biodiversity and the
support this price association rule, which corresponds to estimation of the respective impacts on human welfare. The
16.5% of the database. A similar price association rule can be combination or integration of the ecological and economic
found for the TCM, i.e., the respective price estimates range characteristics to assess and value biodiversity leads to an
between € 9.4 and € 20.1 (see row 12). However, despite the integrated framework. Interdisciplinary work is thus required,
fact that this price rule appears to have stronger empirical involving both economists and ecologists transferring ele-
support because it corresponds to a higher confidence level ments or even theories and models from one discipline to
with a strength of 57%, this association rule is valid for only another and transforming them for their specific, mutually
5.1% of the case studies registered in the database for our consistent purpose. In other words, the underlying objective is
analysis. the development of a common way of thinking about
Clearly, the present database is somewhat too small for the modelling and valuing biodiversity.
identification of robust compound rules. It seems that the
relatively aggregate and standardized valuation figures are too
REFERENCES
imprecise to support firm conclusions and are thus rather
poor to be used for policy guidance. Nevertheless, the
technical method deployed there indicates clearly that there Agrawal, R., Imielinski, T., Swami, A.N., 1993. Database mining: a
is no more reliable quantitative information to be extracted performance perspective. IEEE Transactions on Knowledge and
from the present database. In general, the above described Data Engineering 5 (6), 914–925.
Agrawal, R., Mannila, H., Srikant, R., Toivonen, H., Vercamo, I.,
multi-dimensional classification methods are in principle able
1996. Fast discovery of association rules. In: Fayyad, U.,
to identify relevant causal linkages and structures in a
Piatetsky-Shapiro, G., Smyth, P., Uthurusamy, R. (Eds.),
qualitative database. Advances in Knowledge Discovery and Data Mining. MIT press,
Cambridge MA, USA.
Barbier, E.B., Burgess, J.C., Bishop, J.T., Aylward, B.A., 1995. The
7. Concluding remarks Economics of the Tropical Timber Trade. Earthscan, London.
Bateman, I.J., Brainard, J.S., Lovett, A.A., 1995. Modelling woodland
recreation demand using geographical information systems: a
A comprehensive assessment of ecosystem biodiversity char-
benefit transfer study. CSERGE Global Environmental Change
acteristics, structure and functioning requires the analyst to
Working Paper 95-06, Centre for Social and Economic Research
undertake various important actions. First, the causes of on the Global Environment, University of East Anglia and
biodiversity loss should be determined, in order to improve University College, London.
understanding of socio-economic impacts on biodiversity Bergland, O., Magnussen, K., Navrud, S., 1995. Benefit transfer:
processes and attributes. Second, the range and degree of testing for accuracy and reliability. Discussion paper #D-03/
biodiversity functioning should be assessed, especially in 1995. Agricultural University of Norway, Oslo, Norway.
Bockstael, N.E., McConnell, K.E., Strand, I., 1991. In: Braden, J.B.,
terms of ecosystem–functional relationships. Third, the sus-
Kolstad, C.D., Miltz, D. (Eds.), Recreation, Measuring the
tainability of biodiversity uses should be assessed and the Demand for Environment Quality. Elsevier, Amsterdam.
consequences of biodiversity loss should be determined. Boerwinkel, H.W.J., 1992. Omgevingspsychologie —
Finally, alternative biodiversity management strategies should Landschappelijke Kwaliteitsbeleving. Landbouwuniversiteit,
be assessed and spatial and temporal systems analysis of Wageningen.
alternative biodiversity conservation scenarios should be Bonalume, N., Dickson, D., 1999. $3m deal launches major hunt for
drug Leads in Brazil. Nature 400 (6742), 302.
carried out.
Brouwer, R., Langford, I.H., Bateman, I.J., Turner, R.K., 1999. A
The concept of economic value has its foundations in
meta-analysis of wetland contingent valuation studies.
welfare economics. Therefore, valuation in an economic sense Regional Environmental Change 1, 47–57.
is always the result of an interaction between the subject and Brown, K., Moran, D., 1994. Valuing biodiversity: the scope and
an object. Moreover, economists do not pursue total value limitations of economic analysis. Biodiplomacy: 2Genetic
230 EC O LO GIC A L E CO N O M ICS 6 7 ( 2 00 8 ) 2 1 7 –2 31

Resources and International Relations, African Centre for Macilwain, C., 1998. Bid to block yellowstone enzymes deal. Nature
Technology Studies. ACTS Press, Nairobi. 392 (6672), 177.
Button, K.J., 1995. Evaluation of transport externalities: what can Magnussen, K., 1993. Mini meta-analysis of Norwegian water
we learn using meta- analysis? Regional Studies 29, 507–517. quality improvements valuation studies. Unpublished
Button, K.J., Jongma, S., 1995. Meta-analysis methodologies and manuscript, Agricultural University of Norway, Oslo, Norway.
microeconomics. Trace Discussion Paper TI 95-143. Tinbergen Markandya, A., Chiabai, A., Ding, H., Travisi, C., Nunes, P.A.L.D.,
Institute, Amsterdam. (2008) “Economic Valuation of Forest Ecosystem Services:
Button, K.J., Kerr, J., 1996. Effectiveness of traffic restraint policies: Methodology and Monetary Estimates", in final report The Cost
a simple meta-regression analysis. International Journal of of Policy Inaction (COPI): the case of not meeting the 2010 biodiversity
Transport Economics 23, 213–225. target, European Commission call on ENV.G.1/ETU/2007/0044,
Carson, R.T., Flores, N.E., Martin, K.M., Wright, J.L., 1996. Brussels, Belgium.
Contingent valuation and revealed preference methodologies: Matarazzo, B., Nijkamp, P., 1997. Methodological complexity in
comparing the estimates for quasi-public goods. Land the use of meta-analysis for empirical environmental case
Economics 72, 80–99. studies. International Journal of Social Economics 34 (719),
Cattizone, M., 1999. From ecosystem research to sustainable 799–811.
development. Towards a new paradigm for ecosystem research. Mitchell, R.C., Carson, R.T., 1989. Using Surveys to Value Public
Ecosystem Research Reports N° 26. European Commission, Goods: The Contingent Valuation Methods. Resources for the
Luxembourg. Future, Washington DC, US.
Cooper, H., Hedges, L.V., 1994. The Handbook of Research Munda, G., 1995. Multicriteria Evaluation in a Fuzzy Environment:
Synthesis. Russell Sage Foundation, New York. Theory and Application in Ecological Economics. Physica-Verlag,
Cropper, M.L., Freeman III, A.M., 1991. Environmental health Heidelberg, Germany.
effects. In: Braden, J.B., Kolstad, C.D., Miltz, D. (Eds.), Measuring Munda, G., Nijkamp, P., Rietveld, P., 1993. Information precision
the Demand for Environmental Quality. Elsevier, Amsterdam. and multicriteria evaluation models. In: Giardina, E., Williams,
de Boer, J., Hisschemöller, M., 1998. Omgevingswaarden en de A. (Eds.), The Theory and Practice of Cost-Benefit Analysis.
Kust. VU-IVM Rapport R98/12. Vrije Universiteit, Amsterdam. Edward Elgar, Aldershot, pp. 43–64.
Desaigues, B., Ami, D., 2001. An estimation of the social benefits of Munda, G., Nijkamp, P., Rietveld, P., 1995. Fuzzy multigroup conflict
preserving biodiversity. International Journal of Global resolution for environmental management. In: Weiss, J. (Ed.),
Environmental Issues 1 (1), 73–86. The Economics of Project Appraisal and the Environment.
Dubois, D., Prade, H., 1992. Combination of fuzzy information in Edward Elgar, Aldershot, UK, pp. 161–183.
the framework of possibility theory. In: Abidi, M., Gonzalez, R. Namenwirth, J.Z., 1969. Marks of distinction: a content analysis of
(Eds.), Data Fusion in Robotics and Machine Intelligence. British mass and prestige newspaper editorials. American
Academic Press, San Diego, California, pp. 481–505. Journal of Sociology 74, 343–360.
Farnworth, E.G., Tidrick, T.H., Jordan, C.F., Smathers, W.M., 1981. Navrud, S., Bergland, O., 2001. Value transfer and environmental
The value of economic systems: an economic and ecological policy. In: Spash, C.L., Carter, C. (Eds.), Policy Research Briefs —
framework. Environmental Conservation 8, 275–282. Environmental Valuation in Europe. Cambridge, UK.
Espey, M., 1996. Explaining the variation in elasticity estimates of Nelson, J.P., 1980. Airports and property values: a survey of recent
gasoline demand in the United States: a meta-analysis. The evidence. Journal of Transport Economics and Policy 19, 37–52.
Energy Journal 17, 49–60. Nelson, P., Baaijens, S., 2001. A comparative analysis of multiplier
Florax, R.J.G.M., Nijkamp, P., Willis, K.G. (Eds.), 2002. Comparative effects in tourist regions. Journal of Policy Modelling 22 (7),
Environmental Economic Assessment. Edward Elgar, 821–858.
Cheltenham, UK. Norton, B.G., 1986. On the inherent danger of undervaluing
Fromm, O., 2000. Ecological structure and function of biodiversity species. In: Norton, B.G. (Ed.), The Preservation of Species.
as elements of its total economic value. Environmental and Nunes, P.A.L.D., 2002. Using factor analysis to identify consumer
Resource Economics 16, 303–328. preferences for the protection of a natural area: evidence from
Hedges, L.V., Olkin, I., 1985. Statistical Methods for Meta-Analysis. a valuation survey in Portugal. European Journal of Operational
Academic Press, New York, US. Research 140, 499–516.
Hogenraad, R., 1989. The Rhetoric Texture of an Advertisement in Nunes, P.A.L.D., Onofri, L., 2004. The Profile of a Warm-Glower: a
Consumer Behavior and Strategic Marketing: Anything New? Note on Consumer’s Behavior and Public Policy Implications.
ISIDA, Palermo, Italy. Fondazioni Eni Enrico Mattei - Note di Lavoro, 13.2004,
Hunter, J.E., Schmidt, F.L., 1990. Methods of Meta-Analysis: Fondazioni Eni Enrico Mattei, Milan, Italy.
Correcting Error and Bias In Research Findings. Sage, California. Nunes, P.A.L.D., van den Bergh, J.C.J.M., 2001. Economic valuation
INBio – Instituto Nacional de Biodiversidad, 2001. Biodiversity of biodiversity. Ecological Economics 39, 203–222.
Prospecting. Available at: [Link] Nunes, P.A.L.D., van den Bergh, J.C.J.M., Nijkamp, P., 2004. The
[Link]. Economics of Biodiversity. Edward Elgar, Cheltenham, UK.
Johnson, K.H., Button, K., 1997. Benefit transfers: are they a Palmquist, R.B., 1991. Hedonic methods. In: Braden, J.B., Kolstad,
satisfactory input to benefit cost analysis. Working paper no. 1, C.D., Miltz, D. (Eds.), Measuring the Demand for Environment
Aviation Policy Program, Institute for Public Policy George Quality. Elsevier, Amsterdam, The Netherlands.
Mason University, Fairfax, Virginia. Pawlak, B.C., 1982. Rough sets. International Journal of Information
Johnson, F.R., Fries, E.E., Banzhaf, H.S., 1996. Valuing Morbidity: An and Computer Sciences 11 (5), 341–356.
Integration of the Willingness-To-Pay and Health-Status Index Pearce, D., Moran, D., 1994. The Economic Value of Biodiversity.
Literatures. TER Technical Working Paper No. T-9601, Triangle Earthscan, London, UK.
Economic Research, Durham, NC, US. RAFI – Rural Advancement Foundation International, 2001.
Kacprzyk, J., 1978. Decision making in fuzzy environment with Bioprostecting/Biopiracy and Indigenous People. [Link]
fuzzy termination of time. Fuzzy Sets and Systems 1 (3), [Link]/[Link].
169–181. Riera, P., 2001. Assessment of methodologies for valuing biological
Kirchhoff, S., Colby, B.G., LaFrance, J.Y., 1997. Evaluating the diversity of forests. Internal Report 4. European Forest Institute,
performance of benefit transfer: an empirical inquiry. Joensuu, Finland.
Journal of Environmental Economics and Management 33, Rosengren, K.E., 1981. Advances in Content Analysis. Sage, Beverly
75–93. Hills, US.
 EC O L O G IC A L E C O N O M IC S 6 7 ( 2 0 08 ) 21 7 –2 31 231

Schipper, Y.J.J., 1996. On the valuation of aircraft noise: a meta- Sturtevant, L.A., Johnson, F.R., Desvousges, W.H., 1995. A Meta
analysis. Tinbergen Institute - Ph.D. Research Bulletin 9 (2), 1–18. Analysis of Recreational Fishing. Triangle Economic Research,
Schwartz, J., 1994. Air pollution and daily mortality: a review and a Durham, North Carolina, US.
meta analysis. Environmental Economics 64, 36–52. ten Kate, K., Laird, S., 1999. The Commercial Use of Biodiversity:
Sen, A., 1995. Inequality Re-examined. Oxford University Press, Access to Genetic Resources and Benefit-Sharing. Earthscan,
Oxford. London UK.
Simpson, R., Sedjo, R.A., Reid, J.W., 1996. Valuing biodiversity for use in ten Brink, B.J.E., van Strien, A., van Hinsberg, A., Reijnen, M.J.S.M.,
Pharmaceutical Research. Journal of Political Economy 101, Wiertz, J., Alkemade, J.R.M., van Dobben, H.F., Higler, L.W.G.,
163–185. Koolstra, B.J.H., Ligtvoet, W., van der Peijl, M., Semmekrot, S.,
Slowinski, R., Stefanowski, J., 1989. Rough classification in incomplete 2000. Natuurgraadmeters voor de behoudsoptiek. RIVM
system. Mathematical Computer Modelling 12 (10–11), 1347–1357. rapport 408657005. RIVM, Bilthoven.
Smith, V.K., 1989. Can we measure the economic value of ten Kate, M.M., Laird, S.A., 2004. The Commercial Use of
environmental amenities? Southern Economic Journal 56, Biodiversity. Earthscan, London.
865–878. van den Bergh, J.C.J.M., Button, K., 1999. Meta-analysis, economic
Smith, V.K., 1992. On separating the defensible benefit transfers valuation and environmental economics. In: van den Bergh,
with travel cost models. Water Resources Research 28, 685–694. J.C.J.M. (Ed.), Handbook of Environmental and Resource
Smith, V.K., Kaoru, Y., 1990a. Signals or noise: explaining the Economics. Edward Elgar, Cheltenham, UK.
variation in recreation benefit estimates. American Journal of van den Bergh, J.C.J.M., Button, K., Nijkamp, P., Pepping, G., 1997.
Agricultural Economics 72, 419–433. Meta-Analysis in Environmental Economics. Kluwer,
Smith, V.K., Kaoru, Y., 1990b. What have we learned since Dordrecht.
hotelling's letter? A meta analysis. Economics Letters 32, van den Bergh, J.C.J.M., Button, K., Nijkamp, P., Pepping, G., 1999. Meta-
267–272. Analysis in Environmental Policy Analysis. Kluwer, Dordrecht.
Smith, V.K., Huang, J.-C., 1993. Hedonic models and air pollution: van Kooten, G.C., Bulte, E.H., Sinclair, A.E.R. (Eds.), 2000.
twenty-five years and counting. Environmental and Resource Conserving Nature's Diversity. Ashgate, Aldershot, UK.
Economics 3, 381–394. Walsh, R.G., Johnson, D.M., McKean, J.R., 1989a. Market values
Smith, V.K., Huang, J.-C., 1995. Can markets value air quality? A from two decades of research on recreational demand. In: Link,
meta-analysis of hedonic property values models. Journal of A.N., Smith, V.K. (Eds.), Advances in Applied Economics, vol. 5.
Political Economy 103, 209–227. JAI Press, Greenwich.
Smith, V.K., Osborne, L., 1996. Do contingent valuation estimates Walsh, R.G., Johnson, D.M., McKean, J.R., 1989b. Issues in
pass a scope test? A meta analysis. Journal of Environmental nonmarket valuation and policy application: a retrospective
Economics and Management 31, 287–301. glance. Western Journal of Agricultural Economics 14, 178–188.
Sonner, S., 1998. Suit Tries to Block ‘Bioprospecting’ in Waters, W.G., 1993. Variations in the value of travel time savings:
Yellowstone. Idaho News, 6 March 1998, online at [Link] empirical studies and the values for road project evaluation.
[Link]/030698/A_SECTIO/[Link]. Unpublished manuscript.
Spash, C., 2001. Broadening democracy in environmental policy Weber, R.P., 1983. Measurement models for content analysis.
processes: editorial. Environment and Planning. C, Government Quality and Quantity 17, 127–142.
& Policy 19 (4), 475–481. Wolf, F., 1986. Meta-Analysis: Quantitative Methods for Research
Stanley, T.D., Jarrell, S.B., 1989. Meta-regression analysis: a Synthesis. Sage, Beverly Hills, US.
quantitative method of literature surveys. Journal of Economic Wood, H.B., 1977. Hydrologic differences between selected
Surveys 2, 161–170. forested and agricultural soils in Hawaii. Soil Science Society
Stanners, D., Bourdeau, Ph. (Eds.), 1995. Europe's Environment: 41, 132–136.
The Dobris Assessment. Office for Official Publications of the Woodward, R.T., Wui, Y.-S., 2001. The economic value of wetland
European Communities, Luxembourg. services: a meta analysis. Ecological Economics 37, 257–270.

View publication stats