ANALYSIS OF CORRECTIVE ACTION REQUESTS FROM FOREST STEWARDSHIP COUNCIL AUDITS OF NATURAL FOREST MANAGEMENT IN INDONESIA

Transcription

1 ANALYSIS OF CORRECTIVE ACTION REQUESTS FROM FOREST STEWARDSHIP COUNCIL AUDITS OF NATURAL FOREST MANAGEMENT IN INDONESIA By HERMUDANANTO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2017

2 2017 Hermudananto

3 To the greatest Allah S.W.T. who can listen, see, and assert power over all things, to my beloved wife Aldila Paramita, my daughter Quinsha Devina Yasmin, and my son Hanif Rasyid Yasin who patiently struggled alone in Indonesia while I studied in the USA.

4 ACKNOWLEDGMENTS I thank the USAID PRESTASI program for funding my graduate studies at the University of Florida (UF) and the Tropical Conservation and Development (TCD) for additional funds for my thesis research in Indonesia. I acknowledge the constant support of Professor Francis E. Jack Putz, my advisor, Claudia Romero and Ruslandi, my other principal advisors, Wendell Cropper and Karen Kainer, members of my advisory committee. I also acknowledge the support given to me and my young family by my parents (Sumardjono and Suharyanti) and my wife s parents (Eko Herwiyono and Nuryandari). I also thank to PT. Mutuagung Lestari for fully support to my family in Bogor while I studied at UF. I was helped by colleagues from the Forest Stewardship Council s (FSC) head office in Germany, FSC representatives in Indonesia, and numerous forest management auditors who work in Indonesia. Additionally, my colleagues in the Putz-Romero lab, students and faculty in the School of Forest Resources and Conservation, Denis Valle and James Colle for guidance on multinomial logit regression, and especially Pak Ruslandi and his family all helped me a great deal during my studies and made Gainesville feel like a second home. Other members of the Gainesville community helped make it a really nice place to live and provided me much inspiration and motivation. Finally, I thank my God (Allah S.W.T.). 4

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS... 4 LIST OF TABLES... 7 LIST OF FIGURES... 9 LIST OF ABBREVIATIONS ABSTRACT CHAPTER 1 INTRODUCTION BACKGROUND ON CERTIFICATION AND CORRECTIVE ACTION REQUESTS Forest Management Certification FSC Certification Forest Certification in Indonesia FSC Audit Process Corrective Action Requests - CARs Previous Analyses of CARs Objectives METHODS CAR Data CAR Data Extraction and Definitions Time Taken for Compliance FMU Characteristics and CARs CAR Analysis for Indonesia CAR Comparisons between Indonesia and Mexico RESULTS FSC Certification of Natural Forest Management in Indonesia Temporal Evolution FSC Natural Forests CAR Analysis Results for Indonesia CARs Issued by Category Number of CARs by Category Average Number of CARs per Audit by Category Time Required to Close CARs by Category

6 CAR Focus as a Function of FMU Characteristics Time-To-Close CARs The Auditors Indonesia-Mexico CAR Comparison DISCUSSION AND CONCLUSIONS Discussion The Growth of FSC Forests Certification Common Problems in FSC-Certified Forests in Indonesia Time Taken for Closure Influence of FMU Characteristics on CARs Auditors in Natural Forest Management Benefits and Limitations of Public Summaries of Audit Reports Conclusions LIST OF REFERENCES BIOGRAPHICAL SKETCH

7 LIST OF TABLES Tables page 2-1 The 10-FSC principles CABs for FSC natural forest management certification in Indonesia CAR metacategories and issues CARs sorted by metacategory and issue Variables used for the studies. Variables marked with asterisks were available for this research but not for the study in Mexico by Blackman et al. (2014) FSC-certified natural forests in Indonesia (Adapted from Ruslandi and Romero, 2015; for a key to the abbreviations and definitions, see Tables 3-1 and 3-4) CARs in Indonesia by metacategory and issue Percentage of CARs that were preconditions, major, and direct conditions in Indonesia by metacategory (n=933) Compliance with CARs assigned during pre- and post-certification audits in Indonesia by issue metacategory Average number of CARs issued to FMUs in Indonesia annually since CAR code changes Marginal effects of FMU characteristics on the metacategory on which CARs focused (the complete variables) determined by multinomial logit regression [standard errors noted in parentheses and significant effects noted with asterisks; n=933] FMU characteristics related to CAR focus as determined by multinomial logit regression with different issue metacategories used as the reference (n=933) Relationships between FMU characteristics and time required to close CARs as revealed by multiple linear regression [standard errors noted in parentheses and significant effects noted with asterisks; n=715] Proportions of international and Indonesian auditors for FSC natural forest management in Indonesia

8 4-11 Characteristics of the 35 FSC-certified natural forest and plantation FMUs in Mexico CARs by metacategory and issue in Mexico (Mex) and Indonesia (Ind) Percentage of CARs that are preconditions, major, and direct in Mexico (Mex; n=1,162) and Indonesia (Ind; n=933) by type of issue in each metacategory CAR compliance by issue metacategory Marginal effects of FMU characteristics as a function of the metacategory on which CARs focused based on multinomial logit regression [standard errors noted in parentheses and significant effects noted with asterisks] HCVF toolkit for Indonesia (Consortium to Revise the HCV Toolkit for Indonesia, 2008) FMUs that changed CABs while FSC certified by August 2016 (for a key to the abbreviations, see Table 3-1)

9 LIST OF FIGURES Figures page 4-1 Cumulative area and numbers certificates of natural forest in Indonesia certified by FSC over time (source: Number of CARs issued by metacategory (n=933) Number of CARs issued by issue category (n=933; for a key to issue categories, see Table 3-2) Number of CARs issued by type of audit (n=933) Number of CARs issued per audit by type of CAR (n=933) Number of CARs issued per audit by metacategory Number of CARs issued per audit type and metacategory Number of CARs issued per audit by CAB (ME=main evaluation; RE=reevaluation; for a key to the abbreviations, see Table 3-1) Time required to close CARs by CAB (ME=main evaluation; RE=reevaluation; for a key to the abbreviations, see Table 3-1) Time required to close CARs by metacategory Time required to close CARs by type of audit and metacategory Description of FSC auditors with respect to their background, education, and nationalities New FSC forest certificates in Mexico and Indonesia by year (source: data were processed from database on the FSC's website ( and adapted from Blackman et al., 2014)

10 LIST OF ABBREVIATIONS ASI CAB CAR FMU FSC GoI HCVFs Accreditation Services International Conformity Assessment Body: the organization that provides conformity assessment services and that can be the object of accreditation (also referred to as certification bodies) Corrective Action Request: description of non-conformities and requests to take actions with standard system requirements in audit reports and their public summaries prepared by auditors from a CAB (also referred to as non-conformity reports) Forest Management Unit: a spatial area or areas submitted for FSC certification with clearly defined boundaries managed to a set of explicit long-term management objectives described in a management plan Forest Stewardship Council Government of the Republic of Indonesia High Conservation Value Forests: areas within FMUs with particularly noteworthy environmental characteristics that warrant a higher degree of protection Main evaluation An assessment of an FMU for FSC certification after any preparatory (i.e., pre-evaluation) audits NGO NTFPs PHPL Postcertification Non-Governmental Organization Non-Timber Forest Products Pengelolaan Hutan Produksi Lestari: sustainable management of natural production forests The period when FMUs are already FSC certified and undergo annual surveillance audits or re-evaluation audits at 5-year intervals Pre-certification The period when FMUs are not yet FSC certified but after the first visit made by a CAB Pre-evaluation Re-evaluation An assessment to determine an FMU s readiness for their main evaluation Assessment at 5-year intervals for re-certification or renewal of the FMU s FSC certificate 10

11 SFM Surveillance audit SVLK Verification audit Sustainable forest management Systematic repetition of conformity assessment activities as a basis for maintaining the validity of FSC certification; annual audits Sistem Verifikasi Legalitas Kayu: Indonesia s national timber legality assurance system, which is a mandatory legality certification system for all Indonesian forests and timber industries Assessments to verify that CARs were closed 11

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ANALYSIS OF CORRECTIVE ACTION REQUESTS FROM FOREST STEWARDSHIP COUNCIL AUDITS OF NATURAL FOREST MANAGEMENT IN INDONESIA Chair: Francis E. Putz Major: Forest Resources and Conservation By Hermudananto May 2017 I used Corrective Action Requests (CARs) issued by conformity assessment bodies (CABs) working under the Forest Stewardship Council (FSC) umbrella in Indonesia to explore differences among audited forest management units (FMUs). Specifically, I evaluated how FMU characteristics influenced the types of CARs issued and the time elapsed before their closure. I used 933 CARs from 22 FSC-certified FMUs reported by six CABs in 99 public summaries. I also compared the results of my study with those from a similar study carried out in Mexico that used the same variables and analytical approach. In Indonesia, the average number of CARs issued did not vary with type of audit or CAB. The time that elapsed before CAR closure differed among CABs and type of audit, but time-to-compliance decreased over time. Additionally, large FMUs established before 1998 that employed large numbers of workers and subcontracted logging took longer to close CARs than FMUs with the opposite characteristics. In both Indonesia and Mexico, FSC-certified areas increased over time and most CARs represented only minor infractions that required only procedural changes. While the issues raised in Mexico were predominantly social in focus, both social and 12

13 environmental issues were common in Indonesia. In contrast to Mexico, no characteristics of Indonesian FMUs helped explain the focus of the CARs assigned. Finally, conclusions based on this analysis should be made in light of the limitations of this indirect approach to management assessment based on reports from auditors. 13

14 CHAPTER 1 INTRODUCTION Tropical forest degradation by unnecessarily destructive and often illegal selective logging is a major global environmental concern (e.g., Contreras-Hermosilla et al., 2008; Hosonuma et al., 2012; Lawson and MacFaul, 2010; Tacconi, 2012; Vidal et al., 2005). Forest abuse continues despite increasing recognition that they deliver vital environmental services such as clean air and water, support economic development through their production of wood and non-wood products, and generate local, regional, and global social benefits. One intervention used to stem this tide of destruction is voluntary third-party certification of products harvested from responsibly managed forests (e.g., Auld et al., 2008; Bishop et al., 2012; Ebeling and Yasué, 2009; Rametsteiner and Simula, 2003; Romero et al., 2013). Forest management certification is a private, market-based, and voluntary intervention that provides a mechanism to evaluate whether forest management is economically viable for timber or non-timber forest products (NTFPs), socially and culturally beneficial, and environmentally sound (Romero et al., 2013; Shanley et al., 2005). Certification aims to create economic and other incentives for responsible forestry through commercialization of consumer preferences for products harvested from responsibly managed forests. As a voluntary, market-based environmental initiative, the future of forest certification depends on consumer demand for certified products and on financial or other incentives, such as price premiums and improved market access. To a great extent, these financial benefits are needed to cover the substantial costs of certification (Ruslandi et al., 2014). 14

15 Voluntary third-party certification is based on labelling and then tracing products from forests managed according to responsible-management standards to their end-use consumers. To foster communication and transparency, a wide range of stakeholders participate in the process of certification. The evaluation process itself involves independent third-party conformity assessment bodies (CABs) that assess the quality of forest management and issue written assurances that the audited forest management unit (FMU) meets the requirements specified by the certification system s standards (Auld et al., 2008; Rametsteiner and Simula, 2003). As of 2016, the two most prominent forest management certification schemes at the global level were the Forest Stewardship Council (FSC; and the Programme for the Endorsement of Forest Certification (PEFC; that together certified about 12.4 per cent (495 million hectares) of the total global forest area. The next chapter explains in more detail how FSC certification operates and provides general background on forest management certification. Despite increased use of this conservation tool, some analysts question how and under what conditions its adoption affects forest management outcomes (Blackman and Rivera, 2011; Miteva et al., 2012; Romero et al., 2013). There have been numerous efforts to evaluate the environmental, social, and economic impacts of forest certification in various countries, with many focused on FSC certification (e.g., Cubbage, 2010; Ebeling and Yasué, 2009; Marx and Cuypers, 2010; Miteva et al., 2015; Moore, 2012; Tikina and Innes, 2008). Of particular relevance to this study, Romero et al. (2015) provides a comprehensive background study of the context of natural forest management and FSC certification in Indonesia. 15

16 Forest certification impacts are especially hard to evaluate due to the intervention being voluntarily, which increases the likelihood of positive selection bias. For this reason, naïve comparisons of certified and non-certified FMUs may lead to erroneous conclusions regarding the impacts of the intervention. Until well-designed, field-based evaluations of certification by independent assessors are conducted, certification s impacts can be indirectly inferred by analysis of Corrective Action Requests (CARs) reported by CAB in public summaries of their audits, as demonstrated by Blackman et al. (2013; 2014), Nebel et al. (2005), Peña-Claros et al. (2009), and Rametsteiner and Simula (2003). Evaluations based on analyses of CARs suffer numerous limitations, which are discussed below, but the rationale is that by tracking how CARs change over time, insights about changes in forest management practices associated with certification can be revealed (Newsom et al., 2006). Two basic assumptions that underpin CAR analyses are that audit reports faithfully represent the quality of management practices and that any revealed changes over time are due to the certification intervention (Newsom et al., 2006). I analyzed the public summaries of CARs issued by CABs for FSC certification of natural forest management in Indonesia to reveal how operations changed over time and differed among certified FMUs. I then compared the results from Indonesia with those reported for Mexico by Blackman et al. (2013; 2014) using exactly the same methods developed and used by those authors. This analysis is restricted to natural forests because, in comparison to plantations, they contain more biodiversity, deliver more critical ecosystem services, 16

17 and have higher priority for conservation (Peña-Claros et al., 2009). The focus is on FSC because of the various certification schemes around the world, it is the oldest, most strongly endorsed by international civil society organizations, and the most widely used for certifying natural forest management in the tropics (Auld et al., 2008; Peña- Claros et al., 2009; Ruslandi et al., 2014). The focus is on Indonesia because it supports the third largest tropical natural forest in the world, has >15 years of experience with forest certification, has received substantial funding for certification adoption, and it is where I have 7 years of experience as an auditor. The thesis includes five chapters, the first of which was this general introduction. The second chapter provides background about certification, CARs, and public summaries of audit reports. It also provides information about the history and implementation of forest management certification by the FSC in Indonesia. Chapter 3 describes the methods used, the results are presented in Chapter 4, and the final chapter discusses the results and presents conclusions. 17

18 CHAPTER 2 BACKGROUND ON CERTIFICATION AND CORRECTIVE ACTION REQUESTS Forest Management Certification Forest certification was initially developed and promoted by environmental activists who were frustrated by the failure of public policy and intergovernmental processes to curb forest destruction (e.g., Auld et al., 2008). Cashore et al. (2006) opined that it was the major failures of previous conservation interventions that motivated forest certification. Those failures included the failure of ITTO to promote good management practices in the tropics, lack of progress towards a global forest convention at the Rio Earth Summit, and growing concerns about the potentially deleterious impacts of boycotts of tropical timber. Starting in the 1980s, civil society standard-setting organizations emerged and started to develop national and global standards and monitoring schemes for forest management. After the founding of the FSC in 1993, a number of other environmental certification schemes emerged including the Indonesia Ecolabel Institute (LEI; 1993), the Sustainable Forest Initiative in the U.S.A. (SFI; 1994), the Canadian Standards Association System (CSA; 1996), the international PEFC scheme (1999), CERFLOR in Brazil (Certificação Florestal; 2002), and several others (Cashore et al., 2006). Currently the FSC and PEFC, both private forest certification schemes, dominate globally. Whilst FSC sets specific standards, PEFC is more of an umbrella organization that endorses national certification schemes. Both systems are designed to protect forests and ensure end users that the forest products they purchase are from responsibly-managed forests. The process of forest certification (reviewed by Peña-Claros et al., 2009) starts when a representative of an interested FMU contracts with an approved CAB for a pre- 18

19 evaluation visit. Later, if the FMU decides to continue towards certification, the CAB makes a public announcement about the upcoming initial assessment and gathers information from various stakeholders about the performance of the FMU. Interested stakeholders may respond via letter or , or can request to meet the audit team in person to discuss issues related to the FMU. The initial team of 2-4 auditors generally consists of trained individuals with backgrounds in ecology, sociology, or forestry. Their job is to use field visits, stakeholder interviews, and document reviews to assess the performance of the FMU against the standards provided by the certification organization. The team generates an assessment report that provides general information about the FMU and the audit process in addition to a detailed account of FMU compliance or non-compliance with the standards. This report is reviewed by at least two independent experts, one with the experience and technical knowledge necessary to evaluate the adequacy of the report and the validity of the proposed certification decision and the other with specialist knowledge (e.g., indigenous peoples rights or High Conservation Value Forests-HCVF). The independent reviewers either local or international experts are typically selected and paid by CABs. Although the reviewers may be personally acquainted with audit team members and are paid for their reviews (US $75-US$150 per report), their independence is expected. This independence and other aspects of the certification process are monitored regularly by Accreditation Services International (ASI; Once both reviewers and the CAB are satisfied by the performance of the forest operator, a five-year certificate is granted, which requires annual audits to monitor FMU performance. CABs also publish public summaries of each audit on the forest 19

20 certification schemes or CAB s website to maintain transparency of the audit processes. Certified FMUs are entitled to apply the FSC logo to distinguish their products in the market from non-certified ones. To extend the certificate for another five years, the process just described is repeated. FSC Certification FSC is an international, independent, membership-based, non-governmental organization (NGO) dedicated to promoting environmentally appropriate, socially beneficial, and economically viable management of the world s forests (FSC, 2016). Officially founded in 1993, the scheme brings together a wide range of stakeholders including international timber traders, representatives of environmental NGOs, indigenous groups, and forest worker organizations, as well as other stakeholders including consumers of timber products (Auld et al., 2008). To achieve its goals, FSC developed 10 principles at the international level (Table 2-1) with multiple criteria each. On this base, locally adapted forest management standards are developed by national working groups that specify one or more indicators for each criterion. FSC s international standards for forest management are adapted at the regional or national level to reflect the variety of legal, social, and environmental conditions in different parts of the world. Efforts are made to assure that stakeholders representing the ecological, economic, and social sectors collaborate in an equitable fashion (Auld et al., 2008). FSC is the leading natural forest eco-labelling initiative in the tropics; by January 2017 it had certified 194 million hectares of natural, semi-natural and plantation forests in 82 countries, with 1,476 total forest management certificates (FSC, 2017). 20

21 Table 2-1. The 10-FSC principles. FSC Principles Requirements 1 Compliance with laws and FSC principles 2 Tenure and use rights and responsibilities 3 Indigenous people rights 4 Community relation and workers rights 5 Benefits from the forest 6 Environmental impact 7 Management plan 8 Monitoring and assessment 9 Maintenance of high conservation value forests (HCVFs) 10 Plantations The types of audits performed by CABs are main evaluation audits, annual surveillance audits, re-evaluation audits at 5-year intervals after initial certification, and verification audits to field-check on CAR closure. To determine an FMU s readiness for forest certification and to help it move towards that goal, pre-evaluation audits are strongly recommended but only required for FMUs >50,000 ha and those that include HCVFs. The results of the pre-evaluation are valid for 24 months after completion and should be included in the main evaluation report. If an FMU postpones the main evaluation for >2 years, then another pre-evaluation is required. Given major CARs should be closed within 3 months to avoid suspension or termination of an FSC certificate, instead of annual surveillance audits that are typically conducted within 12 months since previous audit, verification audits are carried out to verify the progress of corrective actions on the major issues either through field visits or office-based audits. The governance structure of FSC consists of a General Assembly of members with three chambers (environmental, social, and economic) subdivided with equal representation of stakeholders from the global North and South. This decision-making body meets every three years. A Board of Directors of nine individuals who serve 3-year 21

22 terms is selected from all chambers to maintain balanced voting power (Auld et al., 2008). To run daily operations, a Director General is appointed. To maintain FSC s independence, employees of government agencies are not allowed to join as members; since 2002, individuals and organizations with commercial interests in forest management and forest products can become members, but only in the economic chamber (Meidinger, 2006; Schepers, 2010). To ensure that CABs are competent, impartial, and capable of providing the auditing services on which certification relies, they are assessed periodically by ASI. Its conducts both office and field audits of CAB performance. Suspension or termination mechanisms may be applied to certified FMUs as well as CABs if they fail to meet ASI s requirements. FSC also deals directly with complaints and appeals from interested stakeholders. Forest Certification in Indonesia With the third largest expanse of tropical forest in the world (120 million hectares; Ministry of Environment and Forestry, 2015), Indonesia s forests contribute directly or indirectly to the well-being of 255 million people. Therefore, maintaining the capacity of its forests to provide society with environment, economic, and social benefits is critical. Certification in Indonesia emerged in the 1990s as a result of the Government of the Republic of Indonesia s (GoI) participation in ITTO and the 1992 Earth Summit (Muhtaman and Prasetyo, 2006). This voluntary environmental intervention emerged in response to poor management practices in natural forests (Maryudi, 2015; Tacconi et al., 2004). In 1990, prior to the founding of FSC in 1993, Rainforest Alliance used its own certification scheme called SmartWood to certify the management of a state-owned teak plantation in Java called Perum Perhutani (Muhtaman and Prasetyo, 2006). Soon 22

23 after the founding of FSC, also in 1993, Emil Salim a professor, economist and former Minister of State for Population and the Environment established the Indonesia Ecolabel Institute (LEI; an independent certification scheme with many similarities to FSC (Muhtaman and Prasetyo, 2006). In 2000, LEI and FSC signed a Joint Certification Protocol to promote international credibility of the LEI certification scheme (Muhtaman and Prasetyo, 2006; Tacconi et al., 2004), but that agreement lasted only until In 2002 the Ministry of Environment and Forestry created a mandatory national certification system called Sustainable Management of Natural Production Forests (PHPL; More recently, another voluntary and international certification scheme called PEFC endorsed a voluntary national forest certification system, the Indonesian Forestry Certification Cooperation (IFCC; which was established in September 2011 to promote SFM. By January 2017, four forest certification schemes were in operation in Indonesia: LEI, FSC, IFCC/PEFC, and the mandatory PHPL. The first three of these schemes also provide chain-of-custody certification for wood products that verify that certified material is identified or kept segregated from non-certified or non-controlled material through this chain. At that time, LEI certified 2.4 million hectares of natural, plantation, and community forests in 55 FMUs, and also granted five chain-of-custody certificates for timber industries (LEI, 2017). At the same time, FSC certified 2.8 million hectares of natural, semi-natural forests, and plantations with 244 chain-of-custody certificates (FSC, 2017). The newest voluntary scheme, IFCC, hold 2.3 million hectares certified natural, semi-natural forests, and plantations in 45 FMUs and also granted 23 chain-of-custody certificates. While PHPL is a mandatory scheme run by the GoI that 23

24 applies to all natural forests in Indonesia, by February 2016 it had granted 235 certificates for 18 million hectares of forests (Kementerian Lingkungan Hidup dan Kehutanan, 2016). Besides forest certification, a Voluntary Partnership Agreement on Forest Law Enforcement Governance and Trade (FLEGT-VPA) was developed between Indonesia and the European Union to combat trade of illegal timber. On 30 September 2013 the two parties signed a commitment to trade only legal timber and timber products through a mechanism using the Indonesian timber legality verification system (Sistem Verifikasi Legalitas Kayu; SVLK; which became mandatory for all FMUs and timber industries in 2009 (Maryudi, 2016). By January 2017, 1,812 timber industries were verified as complying with SVLK standards (LIU, 2017), with 12 million hectares SVLK certified by April 2015 (EUFLEGT, 2017). FSC Audit Process The CARs extracted from public summaries of audit reports and used in this analysis resulted from a long process that I will here describe. Audits start with a review of documents provided by the FMU related to implementation of FSC standards. Then the auditors interview forest managers, workers, and other relevant stakeholders (e.g., representatives of NGOs, governments, local communities), followed by one or more field visits that can last from four to eight days, to verify implementation of the FMU s stated procedures. For annual surveillance audits, which are supposed to be scheduled at random, the audit team with different backgrounds (ecology, sociology, forestry) typically spends less than a week assessing the FMU s performance against FSC standards. Any non-conformities revealed against FSC standards are described in written audit reports by a lead auditor, who leads the audit and makes a decision among 24

25 members. The evidence provided by the auditors is evaluated by the independent peer reviewers of the audit report, especially close scrutiny is applied when the auditors recommend suspension of a certificate. During audits each team member is free to collect information related to the FMU s implementation of FSC standards, and to discuss their observations with forest managers. When audit team members disagree about the FMU s compliance with FSC standard (e.g., whether a CAR should be major or minor), the lead auditor makes the final decision. The lead auditor also coordinates the activities of the team and otherwise leads the audit. On the last day of audit, the audit team reports its preliminary findings to the FMU (e.g., describes potential CARs) in a meeting with forest managers and representatives of the workers. The lead auditor is then responsible for compiling the audit report and then submitting it to the forest management coordinator in their CAB for a peer review by selected experts who are given two weeks to complete their review. During the review process, the lead auditor can be asked for further clarification. Lastly, the final audit report is sent to the FMU and a public document containing a summary of the audit is uploaded onto FSC s website ( Corrective Action Requests - CARs All assessments of FSC certification are required to be reported in public summaries that describe the audit (i.e., names of auditors and dates of audit visits), provide information about the FMU (e.g., socio-economic, cultural, and environmental conditions), and the certification decision, including descriptions of any CARs that were issued. The goal of issuing CARs is to identify a problem that requires action to solve and to prevent its recurrence. By tracking CARs over time, information of an indirect nature is provided that can be used to monitor the changes made in forest management 25

26 practices to satisfy certification requirements. Issuance and subsequent closure of CARs can be considered important steps in the process of improving the performance and achieving the goals of certified forest management operations. All CARs refer to specific principles, criteria, or indicators in the FSC standards and are classified as major or minor. A major CAR (prior to 2006 known as a precondition) is a fundamental non-compliance that precludes achievement of the objectives of the standard. A number of minor failings that have a major cumulative effect can collectively be assigned a major CAR. Additionally, minor CARs can be upgraded to major CARs if they are not dealt with adequately and in a timely manner. Minor CARs (prior to 2006 known as conditions) are assigned for temporary, unusual, or non-systematic failing at the indicator level with limited deleterious effects and do not prevent certification or invoke its suspension. Since 2006 major CARs need to be resolved before FMUs can be certified or within 3 months for FMUs that are already certified. Major CARs that are not closed in the time allotted result in certificate suspension or even termination if the problem persists. Minor CARs must be closed within 12 months or they are upgraded to major CARs (Peña-Claros et al., 2009). Researchers that analyze CARs often classify them as either direct, which require onthe-ground changes, and indirect when only procedural changes are required (Blackman et al., 2013; 2014; Newsom et al., 2006). It should be noted that prior to 2006, FMUs were permitted to close CARs in stages over multiple years, and CARs were not classified as minor and major (Blackman et al., 2013; 2014). Previous Analyses of CARs Although studies on the impacts of FSC certification based on CARs can be criticized because they are not based on field-based observations by independent 26

27 researchers (WWF, 2014), the information in public summaries can nevertheless be used for indirect assessments (Romero and Castrén, 2013; Sheil et al., 2010). By analyzing CARs, a list of topics that needed improvement can be assembled (Peña- Claros et al., 2009). Any analysis based on CARs assumes that the auditors observations and decisions reflect actual conditions in the field, that any problems encountered were described accurately, and that changes in CARs over time represent FMU compliance with FSC standards. As in this thesis, several published analyses of the effects of FSC certification are based on analyses of CARs extracted from public summary reports. For example, Rametsteiner and Simula (2003) analyzed 140 CARs from 32 certified forests in Europe. They categorized the CARs into several issues to assess changes in forest management practices but reported certification had no perceptible positive direct impact on management practices or biodiversity, nor on planning and monitoring. In a study of CARs in Bolivia, Nebel et al. (2005) used 255 conditions issued over the period in ten FSC-certified forests. They reported that documentation, monitoring, and environmental issues were the most common concerns, but they concluded that certification did little to curb forest degradation. In a CAR study in six countries in Europe (Estonia, Germany, Latvia, Russia, Sweden, and the United Kingdom), WWF- EFP (2005) compared the trends of 2,817 non-conformities through categorizing them as either ecological, economic, or social. On this basis they reported that forest certification was associated with improvements in all aspects of forest operations. Newsom and Hewitt (2005) argued that systematic issues were the most dominant problems in certified FMUs based on 2,099 findings from 129 randomly 27

28 selected certified forests in 21 countries. They categorized CARs as environmental, social, economic, forest management, and system-related issues and concluded that most certified operations were required to make on-the-ground changes, which they interpreted as an indication that certification caused changes in the behavior of forest operators. Similarly, Newsom et al. (2006) examined 1,120 CARs (preconditions and conditions) categorized into 34 issues from 80 certified forests in the U.S.A. They suggested that FSC certification provided an effective impetus to improve forest operations. In a study in tropical Latin America and the Caribbean, Peña-Claros et al. (2009), analyzed 3,952 CARs from 123 public summaries. They linked each CAR with one or more specific criterion in the FSC standards. Among the non-conformities, no one type (i.e., social, economic, or environmental) dominated, but the authors reported that they perceived improvements in forest management over time based on the number of CARs issued in surveillance and subsequent recertification audits. Halalisan et al. (2016) carried out a study based on 253 non-conformities issued to 31 FMUs in Bosnia and Herzegovina, Estonia, Romania, Slovenia, and the United Kingdom. They reported that most CARs were minor and pertained to environmental impacts. Additionally, they reported that the total number of CARs issued increased with FMU area and that non-conformities related to HCVFs were particularly common. The exemplary approach of Blackman et al. (2013; 2014) to the study of the types and numbers of CARs issued in Mexico was adopted for my study in Indonesia. Those authors analyzed 1,162 non-conformities from 233 public summaries of FSC audit reports in managed natural forest and plantations. They reported that most CARs pertained to social and forest management issues, especially related to communication 28

29 and conflict resolution and regeneration and reforestation issues, respectively. They observed that the CARs issued were more often related to environmental than economic issues in large FMUs. They also reported that most CARs were minor and procedural (i.e., indirect) and that time used to close CARs declined over time. The authors provided a more robust research protocol for work on CARs analyses than used in previous studies. The improvements in their analytical procedures include better categorization of CARs and more systematic treatment of assigning variables (e.g., major versus minor). Objectives This study employs the method developed by Blackman et al. (2013; 2014) to study CAR evolution and the relationships between FMU characteristics and CAR issuance and time-to-closure in Indonesia. The intention is to reveal, to the extent possible with this indirect approach, under what conditions and how FSC certification affected forest management outcomes. I predicted that the types of CARs issued (environmental, social, economic/legal, forest management) varied with FMU characteristics (area, age, number of workers, permit duration, vertical integration, principal market destination, and subcontracted logging). I also predicted that these characteristics influence the time required to close CARs. My second objective is to compare CARs in Mexico and Indonesia using only the FMU characteristics available for both countries. This comparison seems reasonable insofar as the two countries are similar in their development status and experiences with FSC certification. 29

30 CHAPTER 3 METHODS I employed the CAR analysis method developed by Blackman et al. (2013; 2014) who modified the CAR issue categories of Newsom and Hewitt (2005) and discriminated among preconditions and conditions, major and minor CARS, direct and indirect CARs, and CAR compliance status. As in the Blackman et al. study, I also dropped the systems issue used in previous studies (i.e., management plan, monitoring), which is difficult to categorize. For the in-depth analysis of Indonesian CARs, I used the information about FMUs from Ruslandi and Romero (2015), which provides more detail than available to Blackman et al. I also compared the numbers of CARs per audit and time required to close them by issue category, type of CAR, type of audit, and CAB, including the compliance with CARs during the pre- and postcertification periods. As in the Blackman et al. study, I assessed how FMU characteristics affected the issue category on which CARs focus, and added with their influence on the time to close CAR. Unlike the Mexico study, I did not include plantations nor did I consider CARs related to reduced carbon emissions. For the Indonesia-Mexico comparison, I examined the growth of FSC certificates and the most issue on CARs. I also analyzed the number of CARs by their types (precondition, major, direct), the compliance with CARs, and the influence of FMU characteristic (area) on the type of CAR issues. CAR Data Data for the studies were mostly extracted from public summaries of audit reports published on the FSC website ( from March 2001 through August I restricted my purview to natural forest FMUs in Indonesia with valid, suspended, 30

31 or terminated FSC certificates. Public summaries missing on the FSC website were requested from its head office in Germany, FSC representatives in Indonesia, or the CABs themselves. The 99 public summaries collected represent different types of audits carried out by all six CABs that operate in Indonesia (Table 3-1). Included were 22 main evaluations, 49 annual surveillance audits, 3 re-evaluations, and 25 verification audits. Table 3-1. CABs for FSC natural forest management certification in Indonesia. Code CUC CAB Control Union Certification (PT. PCU Indonesia) Started working Office in Indonesia in Indonesia 2009 AD Premier Building 8 th floor, suite Jl. TB. Simatupang No. 5 Ragunan, Pasar Minggu, Jakarta Website RA Rainforest Alliance 1990 Asia Pacific Regional Office. Jl. Tantular Barat No. 88, Denpasar, Bali SGS Société Générale de Surveillance (PT. SGS Indonesia) 1985 Cilandak Commercial Estate #108C. Jl. Raya Cilandak KKO, Jakarta Selatan SCS Scientific Certification Systems (PT. SCS Indonesia) 2009 Mayapada Tower 11 th FIoor. JI. Jend Sudirman Kav 28, Jakarta /southeast-asia NEP Nature, Ecology and People Consult SA Soil Association Woodmark (PT. Mutuagung Lestari) 2007 Jl. Raya Bogor No.19 KM 33,5 Cimanggis, Depok

32 CAR Data Extraction and Definitions I extracted from each public summary the information related to non-conformities (i.e., precondition and condition, major and minor), which resulted in 933 CARs. Following Blackman et al. (2013; 2014), I divided the CARs into four metacategories (environmental, social, economic/legal, forest management), and then divided each metacategory into several issue categories (Table 3-2). CARs not in any of these issue categories (i.e., chain-of-custody and FSC trademark/labelling) were disregarded. I also considered how CARs varied by year, type of audit (main evaluation, surveillance, reevaluation, and verification audits), and CAB. To test for differences in the number of CARs per audit by type of CAR, metacategory, type of audit, and CAB, I used t-tests or analysis of variance (ANOVA) followed by Tukey s post-hoc tests in R (DeCoster, 2006). Standardizing by the number of CARs per audit makes it possible to compare FMUs that underwent different numbers of audits. Table 3-2. CAR metacategories and issues. Environmental issues Economic and legal issues 1 Aquatic and riparian areas 13 Profitability 2 Sensitive sites and HCVFs 14 Compliance with state, federal, and international laws 3 Threatened and endangered species 4 Landscape-level considerations of forest management 5 Woody debris, snags, legacy trees 6 Soil damage and erosion 15 Illegal activities and trespassing 16 Long-term tenure 32

33 Table 3-2. Continued. Social issues 7 Communication and conflict resolution Forest management issues 17 Roads and skid trails 8 Training 18 Regeneration and reforestation 9 Worker safety 19 Chemical use and inorganic waste management 10 Non-Timber Forest Products (NTFPs) 11 Worker wages and living conditions 12 Special cultural sites 20 Exotic species and pests 21 Conversion to non-forest uses For the 74 (of 933) CARs for which the appropriate issue category was not immediately evident, I assigned a category after consulting with two other scientists who conduct research on FSC certification (C. Romero and Ruslandi) to an agreed-upon category. I kept detailed records of the choices made and strived for consistency in their classification (Table 3-3). The categorized CARs were used to examine how CAR focus varied with FMU characteristics, over time, type of audit, and CAB. Table 3-3. CARs sorted by metacategory and issue. Metacategory Issue Examples of CARs ENVIRONMENTAL 1 Insufficient monitoring of aquatic species Aquatic species not considered in the management plans for HCVFs 2 HCVFs assessments did not cover all protected or conservation sites Lack of measures to protect HCVFs in the management plan 3 Insufficient information provided about the identification and monitoring of rare, threatened, and endangered species and their habitats Lack of awareness of species protected under CITES 4 Issues about area boundaries No site-specific assessments of the environmental impacts of forestry operations 5 Post-harvesting monitoring did not assess wood waste and site damage No policies to reduce damage/waste from harvesting 6 No plan for monitoring forest conditions (i.e. skid trail erosion) Lack of adequate skid drainage structures to reduce erosion 33

34 Table 3-3. Continued. Metacategory Issue Examples of CARs SOCIAL 7 Lack of communication or consultation with communities and other stakeholders about management plans, monitoring, HCVFs Failure to provide public summaries to interested stakeholders related to monitoring or forest operation activities 8 Poor training and supervision of field staff and no formal training plan Inadequate policies and procedures to set standards of competence based on worker qualifications, skills, and experience as a basis for staff recruitment, placement, training, and advancement 9 Unawareness of laws and/or regulations related to worker health and safety and/or has not fully implemented a health and safety management system 10 Lack of control of NTFPs collection by communities No inventory of NTFPs 11 Workers not provided adequate information about the terms and conditions of their employment Lack of periodic reviews of worker welfare 12 Sites of cultural and religious significance not demarcated in the field No local consultation about sites of special cultural, historical, ecological, economic, or religious significance ECONOMIC/ LEGAL FOREST MANAGEMENT 13 No indication of expected profits in the forest operation s budget An annual budget with estimated revenues is not available 14 Used Genetic Modified Organisms (GMOs) Insufficient evidence that the company recognizes and respects the ILO Convention related to discrimination and minimum wages 15 Conservation areas not regularly monitored Inadequate measures to protect the forest from illegal settlement and other unauthorized activities 16 Management plan does not describe long-term forest operations Best logging practice guidelines not consistently implemented 17 Inadequate planning of roads or skid trails Insufficient monitoring of the environmental impacts of harvesting and road-related activities 18 Inadequate program for restoration of degraded sites Insufficient monitoring of plantings in restoration areas 19 Chemicals prohibited by FSC used Chemicals, containers, and solid non-organic wastes not disposed in environmentally appropriate manners 20 Exotic species not controlled and/or monitored for adverse ecological impacts No monitoring of pests and diseases 21 The management plan indicated reduction of the protected area No assurance that the FMU will remain forested 34

35 Time Taken for Compliance CABs determine whether a CAR was appropriately and fully closed within the designated timeframe as specified in FSC s Forest Management Evaluation Standard clause 8.11 (FSC-STD version 3-0). In my study, the time required to close a CAR is measured from the date of its issuance during closing audit meeting to the date the CAR was reportedly closed, with the caveat that some CABs use the date of final audit report in which the CAR closure is described. Of the 933 CARs, 715 were closed by August 2016 (note that CARs issued prior 2006 were excluded because no completion time was specified). To test for differences in the time taken for closure by type of CAR, metacategory, type of audit, and CAB, I used either t-tests or analysis of variance (ANOVA) followed by Tukey s post-hoc tests. Original time taken for CAR closure refers to the time taken to close a CAR without an extension, whereas total time taken for CAR closure includes granted extensions. Percentage of CARs closed on time does not include extensions, whereas percentage closed does. FMU Characteristics and CARs I tested how the numbers and types of CARs and the time needed for their closure varied with FMU characteristics (Table 3-4). For this analysis, I followed Blackman et al. (2013; 2014) and employed a multinomial logit regression model. Such models are used to predict a nominal dependent variable given one or more independent variables that can be nominal or continuous, plus interactions between independent variables. I report marginal effects that indicate how a one-unit change in the independent variable affects the probability that a CAR will focus on a particular issue category instead of an arbitrarily chosen base category, for which I followed Blackman et al. (2014) and selected the economic/legal category. For this analysis, I 35