REPORTS
Supply Market Analysis for Certification of Forest Ecosystem Services Forest Certification Bodies’ Preferences and Audit Capacity International Market Assessment Part I
Center for International Forestry Research March 2013
R e p o r t s
Supply Market Analysis for Certification of Forest Ecosystem Services Forest Certification Bodies’ Preferences and Audit Capacity
International Market Assessment Part I
Wanggi Jaung Louis Putzel
Reports © 2013 Center for International Forestry Research All rights reserved Jaung, W. and Putzel, L. 2013. Supply Market Analysis for Certification of Forest Ecosystem Services: Forest Certification Bodies’ Audit Capacity and Preferences. Report. CIFOR, Bogor, Indonesia. CIFOR Jl. CIFOR, Situ Gede Bogor Barat 16115 Indonesia T +62 (251) 8622-‐622 F +62 (251) 8622-‐100 E
[email protected]
cifor.org
This report and work is part of the Forest Stewardship Council’s ForCES project (Forest Certification for Ecosystem Services) coordinated by FSC and funded by UNEP through a grant of the GEF. Co-‐financing of the work is provided by CIFOR. The views expressed herein are those of the authors and can in no way be taken to reflect the official opinion of UNEP-‐GEF.
Table of contents
Summary
v
Abbreviations
vi
Acknowledgements
vi
1. Introduction
1
2. Certification bodies and certification
3
3. Material and methods
5
3.1. Data collection
5
3.2. Survey design
6
3.3. Factor analysis
10
3.4. Discrete choice models
10
3.5. Three types of supply market structures
11
4. Results
13
4.1. Capacity to audit ecosystem services provision
13
4.2. Capacity to audit bundled ecosystem services
14
4.3. Preferences for certification attributes
15
4.4. Supply market structures analysis
18
5. Discussion
19
5.1. Three enabling conditions for certification
19
5.2. Supply market structures of certification
20
6. Conclusions
22
7. References
23
Appendix I: ecosystem services selected for the market assessment
26
Appendix II: discrete choice models
28
Random utility models
28
Model specifications
29
iii
List of figures and tables
Figures 1. Certification bodies in a certification system and market
3
2. Hypothetical relationship between certification costs and CBs’ profits
9
3. A framework for analyzing certification’s supply markets
11
4. Estimated capacity of CBs to audit ecosystem services
13
5. Scree plot of the factor analysis
14
6. Component plot of the factor analysis showing the three significant groupings of variables corresponding to identified ecosystem services and the relative auditing capacity of CBs
15
Tables 1. 2. 3. 4. 5. 6. 7. 8. A1.
iv
Survey participants Selected ecosystem services for audit capacity analyses Attributes and levels used for discrete choice experiments An example of an experimental design set Result of factor analysis Descriptions of Factor 1, 2, and 3 Results of the logit models Result of the supply market structure analysis A result of the analysis of ecosystem services projects
5 6 7 10 14 14 16 17 27
Summary Certification for forest ecosystem services has the potential to underpin market-‐based mechanisms governing ecosystem services, such as PES and REDD+ schemes. This study examines this potential by identifying enabling conditions for certification and analyzing certification’s current supply market structures. The enabling conditions were identified by discrete choice models. The supply market structures were defined and analyzed by a framework that is based on certification bodies (CBs)’ estimated capacity to audit ecosystem services and the result of discrete choice experiments. The study surveyed accredited CBs of the Forest Stewardship Council (FSC) and Programme for the Endorsement of Forest Certification (PEFC). Findings consist of three components: enabling conditions, CBs’ audit capacity, and ecosystem services that are adopted and not yet adopted into CBs’ auditing businesses. First, three identified enabling conditions for certification of forest ecosystem services include: secured ownership of ecosystem services, low certification cost, and high credibility of certification. Second, watershed protection services require audit capacity building of CBs in order to develop a new certification scheme. Third, the supply market structure analysis shows which ecosystem services have been more or less adopted into CBs’ auditing businesses. On the one hand, ecosystem services, including soil conservation and ecotourism for scenic beauty and cultural experience, have been less adopted into auditing services of CBs. These services might be associated with lack of demand from forest owners. On the other hand, ecosystem services of carbon storage and ecotourism for biodiversity have already well integrated into CBs’ auditing business. To secure CBs in a new certification scheme that target these ecosystem services, the new scheme needs to provide better advantages over the existing schemes, such as a more efficient standard, lower certification cost, and higher credibility. These ecosystem services adopted by CB’s auditing business also signal the presence of forest owners’ demand for certification schemes for these ecosystem services.
v
Abbreviations CARs: Corrective action requests CB: Certification bodies CCB: Climate, Community and Biodiversity CDM: Clean Development Mechanism CIFOR: Center for International Forestry Research CSA: Canadian Standards Association ES: Ecosystem Service(s) ForCES: Forest Certification for Ecosystem Services FSC: Forest Stewardship Council GEF-‐UNEP: Global Environment Facility-‐ United Nations Environment Programme GIS: Geographic information system ICPs: Integrated Conservation and Development Projects ICRAF: the World Agroforestry Centre IIA: Independence of irrelevant alternatives InVEST: the Integrated Valuation of Ecosystem Services and Tradeoffs MA: Millennium Ecosystem Assessment PAs: Protected Areas PES: Payment for Ecosystem Services PEFC: Programme for the Endorsement of Forest Certification PRESA: Pro-‐Poor Rewards for Environmental Services in Africa REDD+: Reducing Emissions from Deforestation and Forest Degradation RUPES: Rewarding Upland Poor for Environmental Services programme VCS: Voluntary Carbon Standard
Acknowledgements The authors thank the FSC who supported this market study and GEF-‐UNEP for its financial support. We also appreciate all the participants of the survey.
vi
1. Introduction Since its emergence in the early 1990s, forest certification has been growing globally. It was initially designed as a market-‐oriented tool that governs sustainable forest management by increasing consumer demand for products obtained from better-‐managed forests (Upton and Bass, 1995; Overdevest and Rickenbach, 2006). The scope of a certification label was used to remain mainly in timber. In addition to timber, however, forests provide other diverse services of great benefit to society and human well-‐being, such as watershed protection, climate regulation, maintenance of cultural heritage, improvement of soil and erosion prevention (Costanza et al., 1997; MA, 2005). In view of this fact, there have been increasing global initiatives to protect and enhance provision of ecosystem services through regulatory and market-‐based schemes. Some major initiatives undertaken include payment for ecosystem services (PES), voluntary carbon markets, Clean Development Mechanism (CDM), and Reducing Emissions from Deforestation and Forest Degradation (REDD+). However, many of these schemes are still in their infancy and face various challenges. Challenges that PES and REDD+ face include asymmetric information between buyers and sellers (Ferraro, 2008), high transaction costs (van Noordwijk et al., 2008; Ghazoul et al., 2009; Vatn, 2010), weak institutional capacity (Corbera and Brown, 2008), and insufficient safeguards of the interests of local communities (Jagger et al., 2012). In addition adequate systems are lacking to measure, report, and verify ecosystem services (or MRV systems, Herold and Skutsch, 2009) as well as to ensure the provision of the promised services upon which the condition of payment or reward is based (or conditionality, Ghazoul et al., 2009). Theoretically forest certification has the potential to alleviate these challenges (Ghazoul et al., 2009; Entenmann, 2010; Merger et al., 2011; Kanowski et al., 2012; Pettenella and Brotto, 2012). First, certification can reduce asymmetric information between ecosystem services sellers and buyers if its credibility (or “credence”) is secured (Rametsteiner, 2002; Zorn et al., 2009). Once certification gains a trust of the buyers by its sound performance, the buyers can readily access to information that adequate quality (or provision) of ecosystem services is guaranteed by simply checking a label of certification. In turn certification is able to balance information on the quality of certified services between the buyers and sellers. Second, theoretically certification can reduce transaction costs through several of its inherent qualities. Service buyers can reduce the costs of finding or sourcing ecosystem services (termed the “search qualities” of the certification). Buyers can also reduce the cost of identifying their expected experience with the certified product or services (through the “experience qualities” of certification) and validate claimed values of the product of services (through the “credence qualities” of the certification) (Zorn et al., 2009). With environmental quality, the benefits of credence qualities associated with certification are especially high since it is hard for consumers to directly observe the generation of credibility (Harbaugh et al., 2011); certification can save a high amount of consumers’ transaction costs in trading ecosystem services. Through certification, ecosystem services sellers also can reduce their costs of demonstrating the value of the service to buyers. Using certification can be an effective and cost efficient way of promoting their ecosystem services since certification with credibility can more readily convince consumers than sellers’ own demonstration about service quality. These reduced transaction costs can benefit the development of PES and REDD+. Third, certification has the potential to strengthen the management capacity of forest owners when they go through basic certification procedures such as applying standards and fulfilling corrective
1
actions required after third-‐party assessments. Improving forest management capacity is one of the main motivations of forest owners to seek certification (Overdevest and Rickenbach, 2006; Araujo et al., 2009; Johnstone and Labonne, 2009). Also some certification bodies provide capacity building for forest owners who desire to obtain a certificate of forest management although they do not directly train those audited by them for transparent auditing. This type of capacity building can be potentially expected from a certification scheme of ecosystem services, fostering the capacity of forest ecosystem owners. Fourth, forest certification has the potential to safeguard the interests of local communities and indigenous people whose livelihoods are vulnerable to emerging ecosystem services markets. In addition to ecological and technical forest management, many forest certification schemes incorporate social requirements (Vogt et al., 2000). This is because sustainability is a key concept embedded in forest certification and achieving it requires an interdisciplinary approach that takes into accounts of social, cultural, and political interactions in forests (Vogt et al., 2000). For instance, FSC certification requires forest managers to recognize and respect indigenous peoples’ rights (Principle 3) and community relations and worker’s rights (Principle 4) (FSC, 2012), so that forest owners are required to go through stakeholder consultation during certification assessments (Nussbaum and Simula, 2005). Canadian Standards Association (CSA) explicitly requires incorporating local values in decision-‐making procedures of assessing certification and asks forest owners to demonstrate an effort to seek participation of local stakeholders (Vogt et al., 2000). Fifth, certification has the potential to strengthen monitoring, reporting, and verifying provision of ecosystem services. Certification assessments consist of procedures of auditing, reporting, and complying to corrective action requests (CARs), key procedures in obtaining and maintain forest certification (Upton and Bass, 1995; Nussbaum and Simula, 2005). For this reason many forest certification schemes have developed, improved, or experienced their certification procedures. These procedures would be applied to ecosystem services management and benefit management. Also once certification is placed in managing ecosystem services, these procedures will be no longer voluntary but become mandatory because certificates cannot be extended without regular certification assessments. Last but not least, lessons from the development and implementation of forest certification have the potential to inform the development of REDD+ and other similar mechanisms (Kanowski et al., 2012). Kanowski et al. (2012) address lessons from forest certification that can be applied to implementation of REDD+. First, certification has enhanced engagement of policy networks and shifted power relations from the business sector to pluralistic stakeholders including environmental NGOs and local communities. Second, certification standards are generally inter-‐dependent on governmental policies (McDermott et al., 2008), emphasizing the importance of government policies that support market-‐based mechanisms including both certification and REDD+. In addition to these lessons, there are other lessons from certification that can benefit the development of REDD+ such as enabling conditions for market-‐based mechanisms. However, are these theoretical benefits of certification achievable in practice and under what conditions? Answering this question requires scrutinizing the various elements of certification schemes for ecosystem services. Assessing the potential to establish and sustain a certification system is critical. In an attempt to address this issue this study analyzes the enabling conditions and current supply market structures for certification. The analysis was conducted by analyzing forest certification bodies’ capacity to audit ecosystem services and their preferences for specific characteristics of certification schemes.
2
2. Certification bodies1 and certification In certification for forest ecosystem services, certification bodies (CBs) are expected to play similar roles as to those they play in forest certification. In forest certification, CBs play various vital roles. First, CBs perform third-‐party assessment upon which the functioning and credibility of certification relies. Although some certification schemes adopt first and second party assessments such as ISO 9000 and ISO 14001, schemes with third party assessment are considered the most credible due to independence (Nussbaum and Simula, 2005; Taylor, 2005). Second, CBs are in charge of auditing procedures of certification such as interpreting certification standards. Owing to this role, their auditing capacity affects the consistency of certification requirements and performance at national and global levels. In addition, CBs are empowered to determine specified regional requirements for certification2 (Upton and Bass, 1995; Auld and Bull, 2003; Rametsteiner and Simula, 2003). Third, CBs connect a certification system with certificate holders and external agents such as environmental NGOs. In other words, in a certification system, it is CBs that directly interact with forest owners and certification supporters on the ground (Kiker and Putz, 1997). Fig. 1. demonstrates how CBs connect a system of certification of ecosystem services and ecosystem services markets. A certification system represents the supply-‐side of certification, and ecosystem services markets represent the demand-‐side of certification. Even though both sides have various agents, they rarely have a direct interaction with agents in the other side of the certification market, except for CBs. Thus, the market demand for certification can be conditioned by CBs’ business activities and performance. For instance, a case study of forest certification in British Columbia in Canada demonstrates how trust and distrust of local supporters in CBs result in certification development and growth (McDermott, 2011).
Figure 1. Certification bodies in a certification system and market
1 Depending on literature, a certification body is also referred as a certifying body (Cashore et al., 2006), certifying organization (Kiker and Putz, 1997), certifier, registration body, or registrar (Nussbaum and Simula, 2005). 2 The proper interpretation of rules of CBs is critical because negative consequences are expected from CBs when their rules are not adequately played. Some literature, for example, reports cases in which certification requirements were leveraged in such a way to serve the self-‐interests of CBs (Camino and Alfaro, 1998; Cerutti et al., 2011; Charles, 2001).
3
This study assumes that CBs’ audit capacity and preferred characteristics of a certification scheme are highly likely to determine the feasibility of developing a certification scheme and the business sustainability of that scheme. First, if the capacity of CBs to audit ecosystem services is not sufficient, establishing a certification system with a sound third-‐party assessment mechanism is not feasible; the capacity of CBs determines their ability to perform objective third-‐party assessments, interpret certification standards, and monitor ecosystem management. Second, understanding CBs’ preferred characteristics of certification is a necessary condition for the success of certification since sustaining a certification scheme depends in large part on their demand. In addition, there should be demand from multiple CBs for a given scheme in order to assure competitive service provision and pricing. If the number of CBs for a scheme is too limited, their bargaining power would likely result in higher audit costs. This would end up restricting the market for a certification scheme since high certification costs would suppress demand from forest owners. The preferences of CBs also indicate the current demand for certification from forest owners since CBs closely work with forest owners and in turn they are aware of the demand of forest owners. The more demand for certification is detected from forest owners (or demand from the demand-‐side in Fig. 1.), the higher chance for CBs to joint auditing business for certification. In this perspective, their audit capacity and preferences play significant roles in building up a certification system and affect the certification market. Certification’s potential to support PES and REDD+ is also subject to these factors.
4
3. Material and methods
3.1.
Data collection
We surveyed the accredited certification bodies (CBs) of Forest Stewardship Council (FSC) and Programme for the Endorsement of Forest Certification (PEFC) as well as a certification body of carbon verification which audits forest management. The FSC has 28 and the PEFC has 48 accredited CBs for serving their forest management schemes as of 2012. These CBs were identified from the websites of the FSC, FSC Network Partners, and the PEFC. CBs only auditing chain of custody certification were excluded from the survey because the objectives of the survey were to analyze the current capacity of CBs to audit provision of forest ecosystem services and to ascertain their preferences for different hypothetical characteristics of certification schemes for forest ecosystem services. Also one certification body auditing carbon verification was identified from CIFOR’s database. Based on this information, a total of 158 contact email addresses were obtained, including the regional offices of CBs in different countries. The survey was conducted from March 12 to 26, 2012. Of the 158 contacts, 20 contacts were not valid and 44 contacts participated in the survey, corresponding to a response rate of 32% (=44/138). Of the 44 participants, 41 currently work in accredited CBs of the FSC and PEFC, one is a consultant auditor who regularly works with these CBs, and two work in CBs for carbon verification. Table 1 identifies the participants. The survey covered 64.29% of accredited CBs of the FSC and 28.26% of the counterparts of the PEFC. Table 1. Survey participants FSC PEFC
Accredited CBs for forest management identified 28 48
CBs participating in the survey 18
Percentages
13
28.26%
64.29%
5
3.2.
Survey design
3.2.1 Survey design for audit capacity analysis To estimate certification bodies (CBs)’ capacity to audit ecosystem services, the survey asked the participants to rate their organizations’ capacity to audit ecosystem services in a five-‐point Liker scale (“not at all”, “low capacity”, “medium capacity”, “high capacity”, and “already in business”). In addition to these five options, an option of “don’t know” was also provided. In the survey we provided descriptions of these five points such that: “not at all” indicated inability to audit; “low capacity” indicated difficulty in auditing; “medium capacity” indicated the possibility of auditing; “high capacity” indicated a high ability to audit; and “already in business” indicated that services were already being provided by their organizations. The ecosystem services included in the survey were selected by reviewing targeted ecosystem services from a total of 175 projects that protect or trade in ecosystem services. The selection procedure is demonstrated in Appendix I. The selection resulted in 12 selected ecosystem services shown in Table 2. Although the result includes ecosystem goods such as timber and non-‐timber forest products, we maintained them in the survey not only because we planned to analyze their potential to be bundled with ecosystem services in terms of CBs’ auditing capacity, but also because they are in the Millennium Ecosystem Assessment (MA) framework which was used to enumerate the ecosystem services included in the projects. Table 2. Selected ecosystem services for audit capacity analyses
Variable name 1 Water quality 2 Water quantity 3 Water risk 4 Carbon 5 Biodiversity 6 Ecotourism: scenic 7 Ecotourism: culture 8 Ecotourism: biodiversity 9 Soil conservation 10 Agriculture goods 11 NTFP 12 Timber
Variable description Watershed protection in forests for provision of high water quality Watershed protection for provision of certain quantity of water Watershed protection to reduce water-‐related risks, such as floods Sequestrating and storing carbon in forests to alleviate climate change Conserving biodiversity Providing scenic beauty through ecotourism Providing cultural experiences through ecotourism Providing biodiversity experiences through ecotourism Conserving soil Providing agricultural goods from forest ecosystems Providing non-‐timber forest products from forest ecosystems Providing timber from forest ecosystems
3.2.2 Scenario design for discrete choice models Discrete choice models are part of stated preference techniques commonly used for market research, transportation studies, and non-‐market good valuation (Bateman et al., 2002; Hoyos, 2010). Discrete choice models elicit individuals’ demand for goods or services described in terms of their attributes or characteristics. This approach originates from the traditional microeconomic theory of consumer behavior postulating that the purpose of goods are to produce specific attributes (Louviere et al., 2000). These attributes are expressed in terms of function, design, and cost efficiency. A choice experiment design was selected for this analysis among survey designs of discrete choice models. Discrete choice experiments have advantages over other discrete choice models (Bateman et al., 2002), as well as over contingent valuation techniques (Adamowicz et al., 1998; Hanley et al., 1998). This is because discrete choice experiments consistently support the economic welfare theory, while other survey designs do not, such as contingent rating and paired comparisons. Discrete choice experiments provide a choice of “none” or “status quo” to survey respondents, so 6
that when they do not face any welfare changes from the given scenarios, they are able to choose the status quo option. Compared to contingent valuation, discrete choice experiments provide a deeper description of trade-‐offs among attributes, offer advantages for transferring benefits, and allow additional statistical tests. Five attributes of certification for forest ecosystem services were established to build scenarios of a discrete choice experiment by examining motivations of forest certificate holders and enabling conditions of forest certification: economic benefits, administrative service, ownership of the ecosystem services, credibility of certification, and total certification costs (Table 3). Each attribute’s level was designed considering a required factorial design. The more levels the attributes have, the more survey questions required to respondents, which in turn reduces a response rate; the attribute levels are limited to three at maximum. Table 3. Attributes and levels used for discrete choice experiments Attributes 1 Economic benefit
Level 1 unknown
Level 2 extremely high
Level 3
2 Administrative service
yes
no
3 Ownership of the ecosystem services
secured
not secured
4 Credibility of certification
50% or less
70%
90% or more
5 Total certification cost
same as FSC forest management certification
50% more than FSC forest management certification
100% more than FSC forest management certification
These attributes are embedded in the assumption that whether choosing a certification scheme for forest management or for ecosystem services, forest owners would face similar motivations because both schemes take place in forests and the main customers of the schemes are forest owners or managers. Similar business enabling conditions are also expected because both schemes are subject to forest policies and regulations as well as the opportunity costs of developing forests commercially. Therefore, scenarios were constructed based on the attributes of certification (Table 3) to analyze CBs’ preferences for certification of ecosystem services, and the certification attributes were derived largely from forest management certification but also analysis of ecosystem service projects. The first attribute, economic benefit, is a main motivation of forest owners to obtain forest certification (Overdevest and Rickenbach, 2006). Forest certification is often promoted as a market-‐ based tool and forest owners are expected to benefit from certification by price premiums or new market access (Cashore et al., 2005; Durst et al., 2006). For those getting a certificate of ecosystem services are expected to have the same motivation. Koellner et al. (2010) demonstrate that for firms investing in ecosystem services, economic benefits are one of the main motivations, which supports that our identification of this attribute as a key attribute in certification of ecosystem services. The second attribute, the administrative service, is based on a potential business value of certification of ecosystem services. This business value implies certification’s potential to build capacity of forest ecosystem owners via certifying procedures since the lack of institutional capacity is considered as a challenge to implementing PES and REDD+ (Corbera et al., 2009; Romijn et
7
al., 2012). Among various types of capacity building for forest owners, registering credits in international registries was selected and tested in this analysis. Examples of these registries include Markit Environmental Registry and NYSE Blue/APX where credits of carbon storage and the improvement of water temperature are traded. When forest owners aim at selling the provision of ecosystem services to the international markets, registering credits of ecosystem services is an essential prerequisite. The registration generally demands administrative procedures such as demonstration of ecosystem services and their management that are complicated and challenging for forest ecosystem owners. Even finding information of the registries would be a challenge for them. For this reason, the registration is dominantly a task of PES and REDD+ developers in implementing PES or REDD+ projects. However, this support is only available to a few forest owners in the world who participate in the development of these projects; the potential demand for this administrative service is expected from many other forest owners interested in trading in ecosystem services. The third attribute, secured ownership of ecosystem services, is not only an enabling condition of forest certification, but also an essential condition for resolving any tenure conflicts among stakeholders in forest ecosystems. First, security about tenure and resources right to forests is an essential condition for achieving good forest management, determining the success of forest certification (Nussbaum and Simula, 2005). Rights include not only the rights of forest owners or managers but also customary rights of local communities and indigenous people. Second, without secured tenure, forest owners have little incentives to invest in their forests including getting a certificate of forest management (Durst et al., 2006). This lack of incentives would result in decreasing demand for certification and preventing the growth of certification schemes of both forest management and ecosystem services. Third, secured ownership of ecosystem services plays a vital role in reducing conflicts among stakeholders of ecosystem services; it is a required condition in developing market-‐based mechanisms associated with ecosystem services including certification of ecosystem services (Durst et al., 2006; Meijaard et al., 2011), PES (Wunder et al., 2008), and REDD+ (Larson, 2010; Yasmi et al, 2012). The fourth attribute, credibility of certification, is a fundamental requirement for certification to reduce asymmetric information between buyers and sellers (Rametsteiner, 2002). Without credibility there will be no demand for certification and consequently no benefit for certified services; this attribute greatly matters to both CBs and certificate holders. The fifth attribute, cost, is a main attribute in the discrete choice experiments as well as in implementing certification schemes. In utilizing the discrete choice experiments, analyzing costs of a good or service is the main interest of economists so that this attribute was also applied to this analysis. In addition, high certification cost is often addressed as one of the main challenges in implementing certification schemes of both forest certification and eco-‐labelling (Durst et al, 2006; Simula et al., 2004; Tikina et al., 2008; Meijaard et al., 2011). Here, total certification cost is a sum of direct and indirect costs for forest owners to obtain certification of ecosystem services (Simula et al., 2004). In this cost analysis, costs to forest owners were analyzed rather than costs to CBs for running auditing business of certification because estimating a cost range for auditing business of CBs is highly challenging. This cost setup based on forest owners is rooted in two assumptions. First, increasing total certification costs results in a decrease in the demand for audits and in profits of CBs (Fig. 2). When certification cost decreases from a high cost (or ph) to a low cost (or pl), the demand of forest owners for certification is expected to increase since more forest owners can afford a certificate.
8
This increasing demand for certification will in turn increase the amount of CBs’ audits from ql to qh. This increasing demand for CBs’ audits will increase or decrease profits of CBs, depending on a profit zone of CBs related to the second assumption. The second assumption is that there are two different profit zones of CBs: increasing and decreasing profit (or π) zones, and CBs are in the increasing π zone. On the one hand, when CBs are in the “increasing π zone,” the certification cost (or certification price) to forest owners is higher than the operating cost of CBs per one audit. In this case, the lower certification cost, the higher demand for CBs’ audits, and the more profits CBs can generate. One the other hand, when CBs are in the “decreasing π zone,” the certification cost to forest owners is lower than the operating cost of CBs. Thus, decreasing certification cost results in decreasing profits of CBs. In this assumption, CBs were expected in the increasing π zone and the assumption is verified in Section 4.3.
Figure 2. Hypothetical relationship between certification costs and CBs’ profits
Based on these five attributes of certification for forest ecosystem services, eight scenario sets were generated (Table 4). The eight scenarios were generated by a fractional factorial design. AlgDesign package of R was utilized, following the method explained by Louviere et al. (2000) (Aizaki and Nishimura, 2008).
9
Table 4. An example of an experimental design set Feature 1. Economic benefits The probability that certification holders will receive economic benefits such as price premiums or access to markets from the ecosystem services certification scheme is: 2. Registration of transactions Certifying bodies register ecosystem services transactions (e.g. trades or swaps) with relevant market oversight agencies. 3. Ownership of the ecosystem services Property rights over ecosystem services that might be targeted for certification in your region are: 4. Credibility Evidence that the certified ecosystem service is actually produced or delivered is: 5. Cost The costs of certifying the service(s), including all direct and indirect costs, are:
Scheme A
Scheme B
unknown
extremely high
No
No
secured
secured
50% proven
90% proven
same as FM certification costs
50% more than FM certification costs
Neither
(I am interested in neither scheme A nor B.)
3.3.
Factor analysis
Factor analysis was applied to examine potential bundles of audit capacity of CBs. In conducting factor analysis, the “don’t know” responses were combined with the “low capacity” responses with the assumption that the survey respondents are experts so that if they are not aware of the capacity, CBs might not yet have auditing experience with these ecosystem services and have low capacity to audit them. After testing various rotations, a varimax rotation was selected for the factor analysis. R 2. 10.1 was also used for the statistical computation.
3.4.
Discrete choice models
To examine discrete choice models, random utility models were adopted (Train 2009). The models examine three groups of variables and their impacts on CBs’ preferences for joining auditing business for certification of ecosystem services. Three variable groups include attributes of certification of ecosystem services, sociodemographic variables of CBs, and CBs’ capacity to audit ecosystem services. With these variables, five logit models were specified: two logit models (Model 1 and 2) and three mixed logit models (Model 3, 4, and 5). Model 1 and 3 had the variables of the five certification attributes only, Model 2 and 4 had the variables of the certification attributes and sociodemographic variables, and Model 5 had the variables of the attributes and audit capacity of CBs. Detailed logit models and their specifications used in the analysis are demonstrated in Appendix II.
10
3.5.
Three types of supply market structures
A framework was developed to analyze the supply market structures of certification for forest ecosystem services by employing two criteria (Fig. 3): CBs’ capacity to audit ecosystem services and its engagement with certification schemes. The first criterion was verified by the result of the audit capacity analysis. The second criterion was verified by the result of Model 5 with the assumption that if CBs are in high audit capacity and already engaged with certification schemes for ecosystem services, they have negative log-‐likelihoods with ecosystem services. This assumption is based on logic that high audit capacity of CBs can be explained by two main factors: internal organizational capacity and capacity supported by certification systems. First, internal organizational capacity comes from internal elements of CBs such as auditors and organizational experience with auditing ecosystem services. For example, CBs’ audit capacity is increased by auditors having strong experience in measuring and monitoring ecosystem services. Second, certification systems can increase the audit capacity of CBs. For instance, if certification adopts standards technically easy to be verified or cost efficient and provides effective guidelines for CBs, audit capacity of CBs can be improved. As a result, high audit capacity can be built by either of these criteria, or a combination of the two.
-‐ CBs: certification bodies -‐ ES: ecosystem services
Figure 3. A framework for analyzing certification’s supply markets
With the two criteria, the framework defines three supply market structures for certification for ecosystem services: type 1, 2, and 3 (Fig. 3). Type 1 defines a supply market structure where CBs do not have sufficient capacity to audit selected ecosystem services. For these ecosystem services, CBs are not yet able to engage with certification schemes due to their limited auditing capacity. Type 2 defines a supply market structure where CBs already have sufficient capacity to audit selected ecosystem services. This sufficient capacity mainly comes from institutional capacity derived from expertise in certification. For these ecosystem services, however, not many certification schemes are adopted by the CBs’ auditing business. This potentially signals that the selected ecosystem services are not in the high demand for certification from forest owners (or the demand-‐side of the
11
certification market in Fig. 1) as well as that a new certification scheme does not have high competition to be integrated into auditing businesses of CBs. Type 3 defines a supply market structure where CBs already have the capacity to audit selected ecosystem services as in type 2. In type 3, however, the capacity was mainly developed specifically for certification schemes with which CBs are currently engaged. The existence of this capacity indicates the potential demand for certification schemes for selected ecosystem services. Also, CBs might join a new certification scheme only if this new scheme provides distinguishable advantages over the schemes that they are engaged with, such as higher profits or demand from forest owners: high competition might exist for a new certification scheme.
12
4. Results
4.1.
Capacity to audit ecosystem services provision
Most of the forest certification bodies (CBs) indicated that their capacity is near or above medium capacity although there are variations among ecosystem services (Fig. 4.). The highest audit capacity was observed in CBs certifying ecosystem goods such as timber products, non-‐timber forestry product (NTFP), and agricultural product. Their capacity to audit the services of biodiversity conservation and carbon storage were also evaluated as high, though not as high as for goods. In comparison, medium-‐level capacity was observed in CBs or auditing the services of ecotourism for cultural experiences, biodiversity conservation, and scenic beauty, and soil conservation. The lowest audit capacity observed was for services reducing water risk and improving water quality and quantity.
(Capacity descriptions) 1. not at all: not feasible to audit 2. low capacity: challenging to audit 3. medium capacity: might be possible to audit 4. high capacity: highly feasible to audit 5. already in business: services are already in business
Figure 4. Estimated capacity of CBs to audit ecosystem services
13
4.2.
Capacity to audit bundled ecosystem services
Three factors were identified, i.e. groupings of significantly related variables corresponding to the ecosystem services and the associated relative audit capacity of CBs. Each factor selected comprised closely related ecosystem services corresponding to watershed protection, ecosystem goods, and ecotourism, and no capacity of CBs to audit services across those categories was observed from the survey. The three factors were selected based on a scree plot of eigenvalues demonstrating the distinctive slope change between Comp. 3 (or Factor 3) and Comp. 4 (Fig. 5.). The sum of Factor 1, 2, and 3 represented 79.49 % of the total data variance. Of rotation methods, the varimax rotation method was chosen because it provided the most distinguishable factor loadings. This rotation is also an orthogonal rotation which indicates that each factor’s correlation is zero (Kaiser, 1958). In order to determine variables of Factor 1, 2, and 3, cut values of 0.8, 0.7, and 0.9 were employed respectively (Table 5). In Factor 1, variables of water quality, quantity, and risk were selected; in Factor 2, variables of agricultural goods, NTFP, and timber product were selected; and in Factor 3, ecotourism based on scenic beauty and cultural experiences were selected (Table 6). According on the selected variables, Factor 1, 2, and 3 were named as watershed protection, ecosystem goods, and ecotourism respectively. Table 5. Result of factor analysis Factor analysis: varimax (factanal) Factor Factor Ecosystem services 1 2 Water quality 0.94 0.16 Water quantity 0.85 0.07 Water risk 0.92 0.12 Carbon 0.63 0.36 Biodiversity 0.44 0.50 Ecotourism: scenic 0.22 0.15 Ecotourism: culture 0.19 0.17 Ecotourism: 0.37 0.41 biodiversity Soil conservation 0.47 0.67 Agriculture goods 0.30 0.83 NTFP 0.09 0.88 Timber 0.01 0.70 Cut v alues >0.8 >0.7
Factor 3 0.21 0.35 0.09 0.25 0.51 0.95 0.96 0.64 0.25 0.25 0.16 0.07 >0.9
Figure 5. Scree plot of the factor analysis
Table 6. Descriptions of Factor 1, 2, and 3 Factors Factor1 Factor2 Factor3
Selected variables Water quality, Water quantity, and Water risk Agriculture goods, NTFP, and timber Ecotourism: scenic, and Ecotourism: culture
Factor names Proportion variance Watershed protection 30.00% Ecosystem goods 25.43% Ecotourism 24.06% Final communality estimates Total = 79.49%
A component plot was drawn from the factor analysis, demonstrating patterns of ecosystem services in terms of CBs’ audit capacity (Fig. 6.). The plot shows services of carbon sequestration or storage is close to Factor 1 (or watershed protection), services of soil conservation is close to Factor 2 (or
14
ecosystem goods), and services of providing ecotourism through biodiversity is close to Factor 3 and to the variable service of biodiversity conservation.
Figure 6. Component plot of the factor analysis showing the three significant groupings of variables corresponding to identified ecosystem services and the relative auditing capacity of CBs
4.3.
Preferences for certification attributes
The result of the logit model analysis including Model 1 and 2 are shown in Table 7. Both models had significant goodness of fit in the likelihood ratio tests. The McFadden R-‐squares demonstrate that Model 2 (0.103) had overall model fitness higher than Model 1 (0.084). In Model 2 certification attributes meeting 5% significant level were secured ownership of ecosystem services, low certification costs, and market benefits. Unlike the other attributes, the cost attribute had a negative log-‐likelihood ratio supporting the assumption that the utility of CBs would decrease by increasing total certification costs to forest owners. Although credibility had a high log-‐likelihood ratio (1.19), its significance level remained in between 5% and 10 %. Sociodemographic variables achieving 5% significant level were “years-‐in-‐business” and “auditing only in developed countries.” However, the log-‐likelihood ratio (0.017) of business years was negligible. The result of the Hausman-‐McFadden test shows that the p-‐value (0.90) of the test was close to 0.10; the IIA assumption of the logit models was rejected with 5% significant level.
15
Table 7. Results of the logit models Logit models ASC (Std. E) Market benefits (Std. E) ES market registration service (Std. E) Secured ES ownership (Std. E)
Model 1 -‐1.213540 ** (0.01192) 0.540140 ** (0.01047)
Model 2 -‐4.61E-‐01 (0.39831) 5.55E-‐01 *** (0.00935)
0.198850 2.04E-‐01 (0.32747) (0.31919) 0.743540 *** 7.44E-‐01 *** (0.00020) (0.00022) Credibility 1.190660 * 1.19E+00 * (Std. E) (0.06728) (0.07107) Cost -‐0.622710 *** -‐6.27E-‐01 *** (Std. E) (0.00013) (0.00013) -‐1.74E-‐02 ** Year (Std. E) (0.03919) Employee 1.14E-‐05
(Std. E) (0.37656) Developing countries -‐1.18E-‐01 (Std. E) (0.75007) Developed countries -‐8.47E-‐01 (Std. E) (0.00168) -‐315.68 Log-‐Likelihood: -‐309.33 Likelihood ratio test Chisq 58.492 71.188 Pr(>Chisq) 2.49E-‐11 *** 8.90E-‐12 McFadden R-‐square 0.08479 0.10319 test Hausman-‐McFadden chisq = 2.1323, df = 6, p-‐value = 0.9071 alternative hypothesis: IIA is rejected
*** ***
***0.01 significant level, ** 0.05 significant level, *0.10 significant level with two-‐tailed tests. -‐ ASC: Alternative Specific Constant
The result of the mixed logit analysis including Model 3, 4, and 5 are described in Table 8. All the models met 0.01% significant level in their likelihood ratio tests and were not restricted to the independence of irrelevant alternatives (IIA) assumption. All the mixed logit models had McFadden R-‐squares higher than the logit models, and Model 5 had the highest McFadden R-‐square (0.31). In Model 5 certification attributes meeting 5% significant level were secured ownership of ecosystem services and total certification cost, while the market benefits attribute was only significant with 10% significant level. Sociodemographic variables meeting 5% significant level were ecotourism for scenic beauty and cultural experiences (Factor 3), ecotourism for biodiversity, and soil conservation.
16
Table 8. Results of the mixed logit models Mixed logit models Model 3
ASC (Std. E)
-‐1.48039 (0.08103)
Model 4 *
Market benefits (Std. E)
0.65552 (0.12584)
ES market registration service (Std. E)
0.52968 (0.13089)
Secured ES ownership (Std. E)
1.47485 (0.00005)
***
Credibility (Std. E)
2.34911 (0.06375)
*
Cost (Std. E)
-‐1.68578 (0.00010)
***
-‐1.02E+00 (0.38034)
Market benefits (Std. E)
6.41E-‐01 (0.14185)
Secured ES ownership (Std. E)
1.28007 (0.00038)
***
Credibility (Std. E)
2.02998 (0.10267)
Cost (Std. E)
-‐1.37466 (0.00054)
***
Watershed (Factor 1) (Std. E)
0.20827 (0.40851)
Ecosystem goods (Factor 2) (Std. E)
-‐0.17370 (0.59281)
Ecotourism (Factor 3) (Std. E)
0.78395 (0.02125)
**
Carbon (Std. E)
-‐0.41677 (0.08548)
*
Biodiversity (Std. E) Ecotourism: biodiversity (Std. E)
-‐0.63106 (0.15135)
-‐1.33097 (0.01009)
***
Soil conservation (Std. E)
1.43875 (0.00000)
***
-‐237.23
-‐1.68E+00 (0.00014)
-‐5.93E-‐03 (0.84865)
Employee (Std. E)
-‐1.20E-‐05 (0.76477)
Developing countries (Std. E)
9.09E-‐01 (0.16657)
Developed countries (Std. E)
-‐9.96E-‐01 (0.09188)
***
Year (Std. E)
0.29978
0.36953 (0.27662)
Cost ((Std. E)
McFadden R-‐square
ES market registration service (Std. E)
Likelihood ratio test Chisq 206.81 Pr(>Chisq) 2.20E-‐16
*
1.99E+00 (0.13734)
-‐241.52
0.78402 (0.06507)
Credibility (Std. E)
Log-‐Likelihood:
Market benefits (Std. E)
***
1.40E+00 (0.00007)
-‐0.59063 (0.73666)
Secured ES ownership (Std. E)
ASC (Std. E)
4.55E-‐01 (0.19110)
Model 5
ES market registration service (Std. E)
ASC (Std. E)
*
***
-‐239.95 209.95 2.20E-‐16
0.30435
***
215.38 2.20E-‐16
0.31222
***
***0.01 significant level, ** 0.05 significant level, *0.10 significant level with two-‐tailed tests. -‐ ASC: Alternative Specific Constant
17
4.4.
Supply market structures analysis
The framework analysis identified that watershed protection (Factor 1) is in type 1 of the supply market structure (no audit capacity), ecotourism (Factor 3) and soil conservation are in type 2 (institutional audit capacity), and carbon storage and ecotourism for biodiversity experience are in type 3 (audit capacity developed for a certification scheme) (Table 8). The first criterion (presence/absence of high audit capacity) was verified with all the variables of ecosystem services in the audit capacity analysis. On the other hand, due to some insignificant results from the mixed logit model, the second criterion (institutional vs. certification-‐specific audit capacity) was available for the supply market structure analysis only for the variables of ecotourism (Factor 3), carbon storage, ecotourism (for biodiversity), and soil conservation.
18
5. Discussion
5.1.
Three enabling conditions for certification
In order for certification to support REDD+ and PES schemes, establishing a functional certification system would be a logical prerequisite. For such a system to work in a rational market, a certain number of enabling conditions is required. This study identified three enabling conditions likely to be required in order to attract the adherence of CBs. There are secured ownership of ecosystem services, low cost of certification, and high credibility of certification. First, secured ownership is critical as shown in the mixed logit models (log-‐likelihood ≥ 1.28; significance < 1%). It showed that participating CBs consider this attribute highly important when choosing a certification scheme. This finding indicates that conflicting or unclear ownership of ecosystem services would work as a barrier to implementing certification for ecosystem services. Although forest certification is sometimes used as a strategy to reinforce forest rights or tenure especially by local communities and indigenous peoples (Bass et al., 2001; Taylor, 2005), issuing a certificate at the first place is not feasible if forests involve deep social conflicts and ambiguous ownerships of ecosystem services. In such circumstances identifying an owner of ecosystem services is especially challenging. Meijaard et al. (2011) point out that for less tangible eocsystem services such as watershed protection and carbon storage, the relationships between resources, resource onwer, and stakeholders become less clear compared to goods such as timber, which highlights the importance of clear and secured ownership in implementing certification for ecosystem services. Many REDD+ studies also show that weak and insecure forest and carbon ownership is one of the main challenges in developing REDD+, corroborating the significance of this enabling condition (Sunderlin et al., 2008; Sunderlin et al., 2009; Larson, 2010; Larson et al., 2012; Yasmi et al., 2012). Second, low cost is a requried enabling condition: the cost of certification of ecosystem services must not be higher than the cost of forest management certification. In the mixed logit models, the cost attribute (-‐1.68 ≤ log-‐likelihood ≤ -‐1.37; significance < 1%) determined that the probability that CBs would join a certification scheme decreased when certification costs increased, which is likely due to the inability of forest owners to pay for the certification. In forest certification, high certification cost has been frequently identified as a constraint to expand forest certification, especially for small forest owners and in developing countries (Butterfield et al., 2005; Durst et al., 2006). The costs of certification of ecosystem services can be divided into direct and indirect costs, as in certification of forest management (Simula et al., 2004). The direct costs are the auditing costs from CBs including initial assessment costs and annual auditing costs, and the indirect costs are the costs to meet certification standards such as costs for improving forest management. In the case of certification for ecosystem services, indirect costs are expected to be higher than those of forest certification because ecosystem services management is not only a less well-‐known concept to forest owners but is also more complex, compared to forest or timber management (Kiker and Putz, 1997; Kremen, 2005; Patterson and Coelho, 2009). Therefore, more corrective actions are to be expected with certification of ecosystem services, resulting in higher indirect costs. In this manner, certification of ecosystem services should take strategies to reduce costs, such as developing cost-‐ effective standards and building the management capacity of forest owners. Third, high credibility is required. Compared to the previous two conditions, this attribute received less significant levels (between 6% to 13%) but higher log-‐likelihood ratios (above 1.99) from the
19
mixed logit models. Credibility is the main value of certification. Certification credibility is subject to the level of asymetric information in the market of certified services, reputations of certifcate holders, and uncertain standards (Harbaugh et al., 2011). Certification gives credence to the impact of sustainable management, the environmental impact which often cannt be directly observed by consumers, so that certification can generate a greater value if certificate holders are associated with activities with potentially negative environmental effects (Harbaugh et al., 2011). Without credibility, certification loses its demand from services buyers, forest owners, and CBs and cannot survive in the market; building credibility is continuously to sustain a certification system.
5.2.
Supply market structures of certification
Three types of supply market structures were defined by CBs’ audit capacity and preferences for the ecosystem services (Fig. 3, and Table 8). Our analysis identified that type 1 includes the services of watershed protection for improving water quality and quantity and reducing water-‐related risks. Type 1 is defined as the supply market structure where CBs do not have sufficient capacity to audit ecosystem services. In other words, capacity of the supply-‐side of certification (Fig. 1) not sufficiently equipped to provide auditing of ecosystem services in forests. Thus, strengthening the capacity of CBs is required to establish a certification system that covers ecosystem services in this market structure. Capacity strengthening would be motivated once CBs sense increasing market demand for certification from forest owners. This result dovetails with findings from other literature. For example, Bond and Mayers (2010) claim that awareness of market opportunities and private sector demand are still low for watershed protection services. Experienced in certification projects, Orrego (2005) also asserts that the watershed market faces challenges in that many watersheds are restricted to the local scale and that it is often challenging to identify payers and suppliers of watershed. Our analysis shows that type 2 includes ecotourism (Factor 3) and soil conservation. Type 2 is defined as the supply market structure where CBs have sufficient audit capacity but are not engaged with many certification schemes. These CBs are likely to adopt a new certification scheme into their audit business if market demand is detected from forest owners because they have high capacity and may wish to add a new scheme to their potentially incomplete portfolio. Since these CBs are not working with many certification schemes, there is little pre-‐existing competition to a new scheme. On the other hand, this low engagement suggests a lack of market demand or low profits from auditing services. Weak market demand from forest owners can drive CBs to abandon their auditing businesses related to these certification schemes. Another possibility is that the market demand for certification exists, but that the auditing business associated with this certification scheme is not profitable for CBs. Identified as type 2, ecotourism included in Factor 3 is ecotourism schemes based on scenic beauty and cultural experience (Table 5 and 6). For ecotourism, currently there are more than 70 similar certification schemes (Font, 2003). However, despite their increasing number, their economic viability is not yet certain (Medina, 2005), and in many cases a certificate heavily depends on third-‐ party funding (Font, 2007). This uncertain economic viability partially explains why the study result indicates that not many forest CBs are involved with ecotourism certification schemes despite the large number of the certification schemes available (with the exception of one non-‐profit CB). Conserving soil quality or quantity is required by a few agricultural certification schemes (e.g., Global-‐GAP) and the FSC solely requires minimizing erosion risk (van Dam et al, 2010), which explains why not many CBs are engaged with auditing services for soil conservation.
20
Our results indicate that type 3 includes ecotourism for biodiversity experience and carbon storage. Type 3 is defined as the supply market structure where CBs have sufficient audit capacity and provide auditing services for specific certification schemes. On the one hand, type 3 confirms existing market demand from forest owners for a certification scheme for ecosystem services. For this reason, many CBs already provide auditing services for these certification schemes. On the other hand, type 3 signals competition among certification schemes not only to obtain the auditing services of CBs, but also to gain more certificate holders. Both CBs and forest owners will choose a certification scheme only if it has advantages over the other schemes, which may include higher credibility, more sufficient standards, low costs, and higher economic benefits. Identified as type 3, ecotourism for biodiversity experience was included in type 3 because CBs consider that the FSC scheme covers biodiversity and that services of biodiversity conservation and ecotourism for biodiversity experience are similar to each other. The principle 9 of the FSC requires forest owners to consider maintenance of high conservation value forests (FSC, 2012), so that many CBs might consider that they audit biodiversity conservation via the FSC scheme. Also the factor analysis demonstrates that audit capacity for biodiversity conservation was similarly high to that for ecotourism for biodiversity experience (Fig. 6). Carbon storage related auditing capacity is also high because this service is already covered by various certification schemes including Climate, Community and Biodiversity (CCB) and the Voluntary Carbon Standard (VCS) (Merger et al., 2001), and audited by many forest CBs for CDM and REDD+.
21
6. Conclusions This market analysis focuses on the supply-‐side of certification (Fig. 1), identifying three enabling conditions and three supply market structures. For certification to support other market-‐based mechanisms successfully such as PES and REDD+, the fundamental enabling condition that must be met are: secured ownership of ecosystem services, low certification costs, and high credibility of certification to forest owners as well as CBs. In the identification of supply market structures, we found that CBs have relatively low capacity (type 1) to audit watershed protection services. If a new certification scheme that includes watershed services is developed, it would need to be accompanied by major capacity building of CBs to audit those services. CBs have high capacity to audit services of soil conservation and ecotourism for scenic beauty and cultural experience, but their business does not currently cover those services (type 2). Therefore, these may be a potentially untapped opportunity provide certification of those services without a large involvement in auditing capacity. On the other hand, the lack of coverage of these ecosystem services may be associated with lack of demand from forest owners, which would need to be established before launching a scheme including these ecosystem services. Finally, CBs have high capacity to audit carbon storage and ecotourism for biodiversity experience, and there are already many CBs that already engage in auditing of schemes covering those services (type 3). The fact that CBs are already engaged in auditing these schemes is indicative of existing demand from forest owners. On the other hand, for a new certification scheme to succeed in the areas of carbon storage and ecotourism, it would need to provide advantages over the existing schemes already audited by these CBs. Such advantages might include a more efficient standard, lower certification cost, and higher credibility. These findings would benefit the development of a scheme for certification of ecosystem services by demonstrating which strategies should be taken into account and which ecosystem services likely face stronger market opportunities or challenges in terms of the supply market of certification. In future studies, it is also necessary to analyze the demand-‐side of certification (Fig. 1), which includes forest owners and stakeholders of PES and REDD+ schemes. These demand-‐side studies would assist in completing the picture of certification’s potential to underpin PES and REDD+ schemes.
22
7. References Adamowicz, W., Boxall, P., Williams, M., & Louviere, J. 1998. Stated preference apporaches for measuring passive use values: choice experiments and contingent valuation. American Agricultural Economics, 80, 64-‐75. Aizaki, H., & Nishimura, K. 2008. Design and analysis of choice experiments using R: A brief introduction. Agricultural Information Research, 17(2), 86-‐94. Araujo, M., Kant, S., & Couto, L. 2009. Why Brazilian companies are certifying their forests? Forest Policy and Economics, 11, 579-‐585. Auld, G., & Bull, G. Q. 2003. The institutional design of forest certification standards initiatives and its influence on the role of science: the case of forest genetic resources. Journal of Environmental Management, 69(1), 47-‐62. Bass, S., Thornber, K., Markopoulos, M., Roberts, S., & Grieg-‐Gran, M. 2001. Certification’s Impacts on Forests, Stakeholders and Supply Chains: Instruments for sustainable private sector forestry series. London: International Institute for Environment and Development Bateman, I. J., Carson, R. T., Day, B., Hanemann, M., Hanley, N., Hett, T., et al. 2002. Economic Valuation with Stated Preference Techniques: A Manual. Cheltenham: Edward Elgar Publishing Limited. Bond, I., & Mayers, J. 2010. Fair deals for watershed services: Lessons from a multi-‐country action-‐learning project. London: International Institute for Environment and Development. Butterfield, R., Hansen, E., Fletcher, R., & Nikinmaa, H. 2005. Forest Certification and Small Forest Enterprises: Key Trends and Impacts -‐ Benefits and Barriers. Washington D.C.: Forest Trends. Camino, R. d., & Alfaro, M. 1998. Rural Development Forestry Network: Certification in Latin America: Experience to Date. London: Rural Development Forestry Network. Cashore, B., Kooten, C. v., Vertinsky, I., Auld, G., & Affolderbach, J. 2005. Private or self-‐regulation? A comparative study of forest certification choices in Canada, the United States and Germany. Forest Policy and Economics, 7, 53-‐69. Cashore, B., Gale, F., Meidinger, E., & Newsom, D. 2006. Confronting Sustinability: Forest Certification in Developing and Transitioning Countries (No. 8). New Haven: Yale School of Forestry & Environmental Studies. Cerbu, G. A., Swallow, B. M., & Thompson, D. Y. 2011. Locating REDD: A global survey and analysis of REDD readiness and demonstration activities. Environmental Science & Policy, 14(2), 168-‐180. Cerutti, P. O., Tacconi, L., Nasi, R., & Lescuyer, G. 2011. Legal vs. certified timber: Preliminary impacts of forest certification in Cameroon. Forest Policy and Economics, 13(3), 184-‐190. Charles, E. 2001. Auditing and Governance in the forestry industry: Between protest and professionalism. Critical Perspectives on Accounting, 12(5), 647-‐671. Corbera, E., & Brown, K. 2008. Building Institutions to Trade Ecosystem Services: Marketing Forest Carbon in Mexico. World Development, 36(10), 1956-‐1979. Corbera, E., Soberanis, C. G., & Brown, K. 2009. Institutional dimensions of Payments for Ecosystem Services: An analysis of Mexico's carbon forestry programme. Ecological Economics, 68(3), 743-‐761. Costanza, R., d'Arge, R., Groot, R. d., Farber, S., Grasso, M., Hannon, B., et al. 1997. The Value of the World's Ecosystem Services and Natural Capital. Nature, 387, 253-‐260. Durst, P. B., Mckenzie, P. J., Brown, C. L., & Appanah, S. 2006. Challenges facing certification and eco-‐labelling of forest products in developing countries. International Forestry Review, 8(2), 193-‐200. Entenmann, S. 2010. Certification of REDD+ pilot projects for biodiversity conservation. In D. Sheil, F. E. Putz & R. J. Zagt (Eds.), Biodiversity Conservation in Certified Forests. Wageningen: Tropenbos International. FSC. 2012. FSC Principles and Criteria for Forest Stewardship (FSC-‐STD-‐01-‐001 V5-‐0 D5-‐0 EN). Bonn. Ferraro, P. J. 2008. Asymmetric information and contract design for payments for environmental services. Ecological Economics, 65(4), 810-‐821.
23
Font, X., Sanabria, R., & Skinner, E. 2003. Sustainable Tourism and Ecotourism Certification: Raising Standards and Benefits. Journal of Ecotourism, 2(3), 213-‐218. Font, X. 2007. Ecotourism certification: potential and challenges. In J. Higham (Ed.), Critical Issues in Ecotourism (pp. 386-‐405). Oxford: Elsevier Butterworth Heinemann. Ghazoul, J., Garcia, C., & Kushalappa, C. G. 2009. Landscape labelling: A concept for next-‐generation payment for ecosystem service schemes. Forest Ecology and Management, 258(9), 1889-‐1895. Hanley, N., Wright, R. E., & Adamowicz, V. 1998. Using choice experiments to value the environment. Environmental and Resource Economics, 11(3-‐4), 413-‐428. Harbaugh, R., Maxwell, J. W., & Roussillon, B. 2011. Label Confusion: The Groucho Effect of Uncertain Standards. Management Science, doi: 10.1287/mnsc.1110.1412. Hausman, J., & McFadden, D. 1984. Specification Tests for the Multinomial Logit Model. Econometrica, 52(5), 1219-‐1240. Herold, M., & Skutsch, M. M. 2009. Measurement, Reporting and Verification for REDD+: Objectives, Capacities and Institutions. In A. Angelsen (Ed.), Realising REDD+: National Strategy and Policy Options (pp. 85-‐100). Bogor: Center for International Forestry Research (CIFOR). Hoyos, D. 2010. The state of the art of environmental valuation with discrete choice experiments. Ecological Economics, 69(8), 1595-‐1603. Jagger, P., Lawlor, K., Brockhaus, M., Gebara, M. F., Sonwa, D. J., & Resosudarmo, I. A. P. 2012. REDD+ safeguards in national policy discourse and pilot projects. In A. Angelsen, M. Brockhaus, W. D. Sunderlin & L. Verchot (Eds.), Analysing REDD+: Challenges and choices (pp. 301-‐316). Bogor, Center for International Forestry Research (CIFOR). Johnstone, N., & Labonne, J. 2009. Why do manufacturing facilities introduce environmental management systems? Improving and/or signaling performance. Ecological Economics, 68(3), 719-‐730. Kaiser, H. 1958. The varimax criterion for analytic rotation in factor analysis. Psychometrika, 23(3), 187-‐200. Kanowski, P. J., McDermott, C. L., & Cashore, B. W. 2011. Implementing REDD+: lessons from analysis of forest governance. Environmental Science & Policy, 14(2), 111-‐117. Kiker, C. F., & Putz, F. E. 1997. Ecological certification of forest products: Economic challenges. Ecological Economics, 20(1), 37-‐51. Kremen, C. 2005. Managing ecosystem services: what do we need to know about their ecology? Ecology Letters, 8, 468-‐470. Larson, A. M. 2010. Forest Tenure Reform in the Age of Climate Change: Lessons for REDD+. Global Environmental Change, JGEC-‐827, 1-‐10. Larson, A. M., Brockhaus, M., & Sunderlin, W. D. 2012. Tenure matters in REDD+: Lessons from the field. In A. Angelsen, M. Brockhaus, W. D. Sunderlin & L. V. Verchot (Eds.), Analysing REDD+: Challenges and Choices. Bogor: Center for International Forestry Research (CIFOR). Louviere, J. J., Hensher, D. A., & Swait, J. D. 2000. Stated Choice Methods: Analysis and Applications. Cambridge: Cambridge University Press. McDermott, C. L. 2011. Trust, legitimacy and power in forest certification: A case study of the FSC in British Columbia. Geoforum, 43(3), 634-‐644. McDermott, C. L., Noah, E., & Cashore, B. 2008. Differences That ‘Matter’? A Framework for Comparing Environmental Certification Standards and Government Policies. Journal of Environmental Policy & Planning, 10(1), 47-‐70. Medina, L. K. 2005. Ecotourism and Certification: Confronting the Principles and Pragmatics of Socially Responsible Tourism. Journal of Sustainable Tourism, 13(3), 281-‐295. Meijaard, E., Sheil, D., Guariguata, M. R., Nasi, R., Sunderland, T. C. H., & Putzel, L. 2011. Ecosystem services certification: Opportunities and constraints (pp. 57p.). Bogor, Indonesia: Center for International Forestry Research (CIFOR). Merger, E., Dutschke, M., & Verchot, L. 2011. Options for REDD+ Voluntary Certification to Ensure Net GHG Benefits, Poverty Alleviation, Sustainable Management of Forests and Biodiversity Conservation. Forests, 2, 550-‐577.
24
Millennium Ecosystem Assessment (MA). 2005. Ecosystems and Human Well-‐being: A Framework for Assessment: Island Press. Nussbaum, R., & Simula, M. 2005. The Forest Certification Handbook (2nd ed.). London: Earthscan. Orrego, J. 2005. The Plan Vivo experience with carbon service provision and the potential lessons for watershed service projects. London: International Institute for Environment and Development. Ottaviani, D., & Scialabba, N. E.-‐H. 2011. Payment for Ecosystem Services and Food Security. Rome: FAO. Overdevest, C., & Rickenbach, M. G. 2006. Forest certification and institutional governance: An empirical study of forest stewardship council certificate holders in the United States. Forest Policy and Economics, 9(1), 93-‐102. Patterson, T. M., & Coelho, D. L. 2009. Ecosystem services: Foundations, opportunities, and challenges for the forest products sector. Forest Ecology and Management, 257(8), 1637-‐1646. Pettenella, D., & Brotto, L. 2012. Governance features for successful REDD+ projects organization. Forest Policy and Economics, 18, 46-‐52. Rametsteiner, E. 2002. The role of governments in forest certification -‐ a normative analysis based on new institutional economics theories. Forest Policy and Economics, 4, 163–173. Romijn, E., Herold, M., Kooistra, L., Murdiyarso, D., & Verchot, L. 2012. Assessing capacities of non-‐Annex I countries for national forest monitoring in the context of REDD+. Environmental Science and Policy, 19-‐20, 33-‐48. Simula, M., Astana, S., Ishmael, R., Santana, E. J., & Schmidt, M. L. 2004. Report on Financial Cost-‐Benefit Analysis of Forest Certification and Implementation of Phased Approaches. Yokohama: International Tropical Timber Council. Sunderlin, W. D., Hatcher, J., & Liddle, M. 2008. From Exclusion to Ownership? Challenges and Opportunities in Advancing Forest Tenure Reform. Washington DC: Rights and Resources Initiative. Sunderlin, W. D., Larson, A. M., & Cronkleton, P. 2009. Forest tenure rights and REDD+: From inertia to policy solutions. In A. Angelsen (Ed.), Realising REDD+: National Strategy and Policy Options (pp. 139-‐150). Bogor: Center for International Forestry Research (CIFOR). Taylor, P. L. 2005. In the Market But Not of It: Fair Trade Coffee and Forest Stewardship Council Certification as Market-‐Based Social Change World Development, 33(1), 129–147. Tikina, A., Kozak, R., & Larson, B. 2008. What factors influence obtaining forest certification in the U.S. Pacific Northwest? Forest Policy and Economics, 10(4), 240-‐247. Train, K. 2009. Discrete Choice Methods with Simulation. Cambridge: Cambridge University Press. Available: http://elsa.berkeley.edu/∼train Upton, C., & Bass, S. 1995. The Forest Certification Handbook. London: Earthscan Publications Limited. van Dam, J., Junginger, M., & Faaij, A. P. C. 2010. From the global efforts on certification of bioenergy towards an integrated approach based on sustainable land use planning. Renewable and Sustainable Energy Reviews, 14(9), 2445-‐2472. van Noordwijk, M., Purnomo, H., Peskett, L., & Setiono, B. 2008. Reducing emissions from deforestation and forest degradation (REDD) in Indonesia: options and challenges for fair and efficient payment distribution mechanisms (No. 81). Bogor: ICRAF. Vatn, A. 2010. An institutional analysis of payments for environmental services. Ecological Economics, 69(6), 1245-‐1252. Vogt, K. A., Larson, B. C., Gordon, J. C., Vogt, D. J., & Fanzers, A. 2000. Forest Certification: Roots, Issues, Challenges, and Benefits. New Haven: School of Forestry and Environmental Studies, Yale University. Wertz-‐Kanounnikoff, S., & Kongphan-‐apirak, M. 2009. Emerging REDD+: A Preliminary Survey of Demonstration and Readiness Activities. Bogor: CIFOR. Wunder, S., Engel, S., & Pagiola, S. 2008. Taking stock: A comparative analysis of payments for environmental services programs in developed and developing countries. Ecological Economics, 65(4), 834-‐852. Yasmi, Y., Kelley, L., Murdiyarso, D., & Patel, T. 2012. The struggle over Asia’s forests: An overview of forest conflict and potential implications for REDD+. International Forestry Review, 14(1), 99-‐109. Zorn, A., Lippert, C., & Dabbert, S. 2009. Economic Concepts of Organic Certification: Deliverable 5: Economic Analysis of Certification Systems in Organic Food and Farming.
25
Appendix I: ecosystem services selected for the market assessment Ecosystem services for the market assessment were selected by examining 175 ecosystem services projects whose main objectives are either conserving or selling ecosystem services. The project schemes include payment for ecosystem services (PES), protected areas (PAs), and integrated conservation and development projects (ICDPs). The project data were collected online via project websites. The analyzed projects include Rewarding Upland Poor for Environmental Services (RUPES) programme, Pro-‐Poor Rewards for Environmental Services in Africa (PRESA), payment for ecosystem services (PES) projects of the CIFOR, and a database for the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) project. RUPES and PRESA are PES projects run by the World Agroforestry Centre (ICRAF). InVEST is a tool created by the Natural Capital Project, both mapping and modelling ecosystem services in the world by Geographic Information System (GIS). InVEST built a database3 that contains information of 159 ecosystem services projects. The projects in the database of InVEST are schemes of PES, Protected Areas (PAs), and Integrated Conservation and Development Projects (ICDPs). The database, for instance, includes a project site of payment for watershed services in Lombok, which is Pilot 4 of the ForCES project, and a project site of PRESA in the Usambara Mountains in Tanzania. In order to count ecosystem services targeted by the projects, an analytic framework with 24 categories of ecosystem services was adapted from the Millennium Ecosystem Assessment (MA) (Table A1). Some of the categories were combined together due to projects not distinguishing ecosystem services as the analytic framework categorizes. For example, a project target described as water quality and quantity improvement was counted as the watershed “regulating service” (ESG14) although it might belong to the fresh water “provision service” (ESG6). Those targeting soil conservation to improve an agricultural system were counted as a “supporting service” (ESG23) even though it might have been counted as a “regulating service” (ESG13). The selection result demonstrates that the most commonly targeted ecosystem services by the projects are watershed protection (152 sources = ES12 + ES14), carbon storage4 (27), biodiversity conservation (42 = ES4 + ES9), cultural service (48) and ecotourism (79)5. Soil conservation (41) also includes production of nutrients (17) so the service of nutrient projection was combined into the service of soil conservation.
3 The database is available at: http://www.naturalcapitalproject.org/database.html 4 REDD+ projects were not targeted in this analysis because carbon sequestration is obviously within the eventual scope of the project and all supporting research. As a result, carbon storage only received a count of 27. However, carbon storage has an obviously strong potential demand for the ecosystem services certification due to increasing demonstration and readiness activities of REDD+ globally (Wertz-‐Kanounnikoff & Kongphan-‐apirak, 2009; Cerbu et al., 2010); it is included in the analysis. 5 As an ecosystem service, ecotourism does not stand alone: it entails conserving the other services of scenic beauty, cultural values, and biodiversity. However, these ecosystem services are examined separately due to different categorizations used in project reports.
26
Table A1. A result of the analysis of ecosystem services projects ES category 1. Providing services
2. Regulating services
#
Ecosystem services
4. Supporting services
Comments
ESG1
Agriculture
54
Eco-‐certifications certify them.
ESG2
NFTPs
5
ESG3
Timber
29
FSC and eco-‐certifications certify them. FSC certifies them.
ESG4
Biodiversity-‐Genetic
32
ESG5
NFTPs-‐Medicines
22
ESG6
Fresh water
0
Ecosystem service market (PES) exists. FSC and eco-‐certifications certify them.
ESG7
Aesthetic
0
ESG8
Ecotourism
79
Eco-‐certification schemes exist.
ESG9
Biodiversity
10
ESG10
Air quality
0
Ecosystem service market (PES) exists.
ESG11
Carbon
27
ESG12
Watershed –risk management Soil
41
ESG14
Watershed – Water quality and quantity
111
ESG15
Health
0
Ecosystem service market (PES) exists. A FSC case exists in Mexico.
ESG16
Pest
0
ESG17
Pollination
5
ESG18
Disasters
3
ESG19
Culture
48
Relevant to ecotourism
ESG20
Spiritual
0
ESG21
0
ESG22
Traditional ecological knowledge. Education /research
0
ESG23
Soil
41
Relevant to agriculture
ESG24
Nutrient
17
Relevant to agriculture
ESG13
3. Cultural services
Count
0
Ecosystem service market (PES, REDD+, A/R CDM), and carbon certification schemes exist. Ecosystem service market (PES) exists.
27
Appendix II: discrete choice models
Random utility models Following Train (2009), random utility models were employed to examine both impacts of the given attributes of certification and of respondents’ sociodemographic variables on their decisions. Both logit and mixed logit models were tested in this analysis. Let’s say there is a certification body n who face J alternatives, or certification schemes. A certification body takes the utility ( U nj ) from its choice, and the utility can be divided into two parts: utility known ( Vnj ) and unknown ( ε nj ) to the researchers: U nj = Vnj + ε nj ∀j = V( xnj , sn ) + ε nj
(1)
ε nj is treated as random. With a logit model ε nj takes the assumption that each unknown utility is extreme value that is distributed independently and identically, or called the independence from irrelevant alternatives (IIA) property. In order to test the IIA property, the Hausman-‐McFadden Test was utilized by generating an artificial nested logit model having the second and third alternatives as subset alternatives (Hausman and McFadden, 1984). With a mixed logit model, however, this assumption is relaxed, a strong advantage of a mixed logit model (Hoyos, 2010; Train, 2009). Based on this setting, the probability that a certification body n chooses alternative scheme i over scheme j is: Pni = Prob( Vni + ε ni > Vnj + ε nj ∀j ≠ i ) (2) The logit probabilities of the logit models become (Train, 2009):
Pni =
expVni ∑ expVnj
(3)
j The mixed logit probabilities use the integrals of logit probabilities over parameter densities, becoming:
Pni = Lni (β ) f (β )dβ (4)
∫
where Lni ( β ) is the utility observed by researchers, and f ( β ) is a parameter’s density function. Simulation techniques are used to calculate mixed logit probabilities. Computing logit and mixed logit probabilities were done by R 2. 10.1 installed with mlogit and Support.CEs packages.
28
Model specifications Five logit models were specified: two logit models (Model 1 and 2) and three mixed logit models (Model 3, 4, and 5). Model 1 and 3 had the variables of the five certification attributes only, Model 2 and 4 had both the variables of the certification attributes and sociodemographic factors, and Model 5 had the variables of the attributes and estimated audit capacity of certification bodies. Model 1 and 3 were specified as: V1nj = ASC + β1 x1nj + β 2 x2 nj + β 3 x3nj + β 4 x4 nj + β 5 x5 nj (5) where, ASC is alternative specific constant that captures the average effect on utility of all factors not included in the choice model (Train, 2009); x1nj is the variable of economic benefits; x 2 nj is the variable of administrative service; x3nj is the variable of ownership of the ecosystem services; x 4 nj is the variable of credibility of certification; and x5 nj is the variable of total certification costs to forest owners. Model 2 and 4 were specified as: 4
5
4
k =1
i =1
k =1
V 2 nj = V1nj + ASC (∑ γ k sknj ) = ASC + ∑ β i xinj + ASC (∑ γ k sknj ) (6)
where, s1nj is how many years certification bodies have been in their business; s2 nj is the number of employees in certification bodies; s3 nj is a dummy variable that indicates certification bodies only auditing in developing countries; and s4 nj is a dummy variable that indicates certification bodies only auditing in developed countries. When both s3 nj and s4 nj are zero, it indicates certification bodies auditing internationally. Model 5 was specified as:
V 3nj = V1nj + ASC (
11
∑γ
s ) (7)
k knj
k =5
where, s5 nj , s6 nj , and s7 nj are Factor 1, 2, and 3 from the factor analysis. Each factor variable was built by averaging selected capacities of certification bodies. s8 nj , s9 nj , s10 nj and s11nj are estimated capacities of certification bodies to audit ecosystem services of carbon storage, biodiversity conservation, ecotourism for biodiversity experience, and soil conservation respectively.
29
This research was carried out by CIFOR as part of the CGIAR Research Program on Forests, Trees and Agroforestry. This collaborative program aims to enhance the management and use of forests, agroforestry and tree genetic resources across the landscape from forests to farms. CIFOR leads the program in partnership with Bioversity International, CIRAD (Centre de coopération internationale en recherche agronomique pour le développement), the International Center for Tropical Agriculture and the World Agroforestry Centre.
cifor.org
blog.cifor.org
Center for International Forestry Research CIFOR advances human wellbeing, environmental conservation and equity by conducting research to inform policies and practices that affect forests in developing countries. CIFOR is a CGIAR Consortium Research Center. CIFOR’s headquarters are in Bogor, Indonesia. It also has offices in Asia, Africa and South America.
