NATURE AND EXTENT OF FARMER PARTICIPATION IN COCOA CERTIFICATIONIN GHANA, WEST AFRICAByEbenezer Offei AnsahA THESISSubmitted toMichigan State Universityinpartialfulfilment of the requirementsfor the degree ofCommunity Sustainability•Master of Science2016ABSTRACTNATURE AND EXTENT OF FARMER PARTICIPATION IN COCOA CERTIFICATION INGHANA, WEST AFRICAByEbenezer Offei AnsahCertification of cocoa producers isanexample of market-based mechanismsthatpromotesustainable agricultural practices.Suchmechanismswould benefit from better understanding ofcircumstances underlining the participation of farmers in certification programs.This studyexamines farmer participation in and stakeholder assessment of cocoa certification in Ghana. Italsoassessesdeterminantsoffarmer participation. Data was collected from farmers and otherstakeholders using a household survey, focus group discussions, and in-depth interviews.Descriptive and inferential statistics as well as a binaryprobitregressionmodelwere used toanalyse and present the quantitative results. Coding of the qualitative data was used to analysethe results of the focus groups and in-depth interviews. The results reveal thatthefarmers aresatisfiedwith how decisionsrelated to cocoa certificationare madeand largely agree withthosedecisions and certification requirements. However, farmers† level of knowledge of certificationas wells as their involvement in compliance inspections are not as expected.Importantdeterminants of participationwere:channel for implementation;farmer being a leader; age offarmer; and number of adult household members.Cost of membership registration, time forattending meetings, existing relationship with licensed buying companies, and inadequateinformation about certificationwere found to beimportant barriers to participation. Keyimplicationsof the findingsinclude the need to intensify the campaign for certificationtoimprove farmers†knowledge of the program.Also, innovative strategies to increase theefficiency of compliance inspectionswouldpotentiallybe beneficial.ACKNOWLEDGEMENTSMy sincere appreciation goes to The MasterCard Foundation (MCF) for funding my Master†sDegree through the MCF Scholars Programat Michigan State University (MSU). This study wassponsored by The Department of Community Sustainability (CSUS) at Michigan StateUniversity andwas alsosupported by The MCF Scholars Program at MSU.I am grateful to myadvisory committee for the guidance. I also thank my field enumerators Kenneth Quansah, AddoKingsley, Francis Essandoh, and Vida Owusu Mensah for theirhard work.I appreciate theassistanceofstaff of the Social Science and StatisticsUnit (SSSU)of the Cocoa ResearchInstitute of Ghana (CRIG).I thankMrFrank Otchere, PhD Candidate at The University of NorthCarolina at Chapel Hill for all the technical assistance and support.TABLE OF CONTENTSLIST OF TABLES.........................................................................................................................viLIST OF FIGURES......................................................................................................................viiKEY TO ABBREVIATIONS......................................................................................................viiiINTRODUCTION.........................................................................................................................1CHAPTER 1..................................................................................................................................3FARMER PARTICIPATION IN AND STAKEHOLDER ASSESSMENT OF COCOACERTIFICATION PROGRAMS IN GHANA, WEST AFRICA............................................3Introduction...............................................................................................................................3The Concept of Cocoa Certification........................................................................................5The State of Cocoa Certification..............................................................................................7Methods......................................................................................................................................9Study Design...........................................................................................................................9Sample and Sampling Techniques & Procedures...............................................................10Data Collection.....................................................................................................................12Variables and Measurement.................................................................................................12Data Analyses........................................................................................................................14Results......................................................................................................................................14In-depth Interviews andFocus Group Discussions............................................................15Structures and Procedures for Implementing Certification...............................................15Training and Compliance Verification...............................................................................17Deciding on Price Premiums and Inspectors.....................................................................18Sustainability Assessment...................................................................................................20Survey Results.......................................................................................................................23Knowledge of Certification................................................................................................23Participation in Certified Farmer Groups.........................................................................26Assessment of Certification Requirements and ExpectedBenefits.....................................30Discussion.................................................................................................................................33Farmers• Knowledge of Certification..................................................................................33Participation in Certified Farmer Groups...........................................................................34Stakeholder Assessment of Certification.............................................................................36CHAPTER 2................................................................................................................................38DETERMINANTS OF FARMER PARTICIPATION IN COCOA CERTIFICATIONPROGRAMS IN GHANA, WEST AFRICA............................................................................38Introduction.............................................................................................................................38Cocoa Production in Ghana...................................................................................................41Cocoa Certification Programs................................................................................................42Previous Studies ofParticipation in Sustainable Agricultural Practices Programs.........46Methods....................................................................................................................................51Study Design and Setting......................................................................................................51Sample, Sampling Techniques & Procedures, and Sample Weights..................................52Data Collection.....................................................................................................................54Variables and Measurements...............................................................................................54Sample Characteristics.......................................................................................................54Regression Variables..........................................................................................................55Data Analysis and Estimation Model...................................................................................56Results......................................................................................................................................57Characteristics of Survey Sample........................................................................................58Determinants of Farmer Participation................................................................................61Descriptive Statistics..........................................................................................................61Regression Estimates..........................................................................................................62Certification Standard and Implementation Channel......................................................62Individual Characteristics................................................................................................62Household and Farm Characteristics...............................................................................63Barriers to FarmerParticipation.........................................................................................63Discussion.................................................................................................................................66Determinants of Participation..............................................................................................66Certification Standard and Implementation Channel........................................................66Individual Characteristics..................................................................................................68Household and Farm Characteristics................................................................................69Barriers to Participation.......................................................................................................70CONCLUSIONS AND IMPLICATIONS FOR POLICY AND RESEARCH.....................72APPENDICES.............................................................................................................................74APPENDIX A:SAMPLING FRAMEWORK......................................................................75APPENDIX B:LOCATION OF STUDY COMMUNITIES..............................................77APPENDIX C:DETAILS OF SAMPLE WEIGHT CALCULATION.............................79APPENDIX D:DESCRIPTION AND HYPOTHESES OFINDEPENDENTVARIABLES................................................................................................81APPENDIX E:HOUSING CHARACTERISTICS.............................................................83APPENDIX F:DESCRIPTIVE STATISTICS OF REGRESSION VARIABLES..........85APPENDIX G:QUALITATIVE INTERVIEW AND DISCUSSION GUIDE.................87APPENDIX H:HOUSEHOLD QUESTIONNAIRE...........................................................91BIBLIOGRAPHY.....................................................................................................................108LIST OF TABLESTable 1.1: Trends in Production and Sale of Certified Cocoa (1000 Tonnes)‡‡‡‡‡‡‡‡.8Table1.2:Farmers† Awareness of the Main Objectives of Cocoa Certification‡‡‡‡‡‡...25Table 1.3: Time Lag between Implementation of Certification and Farmer Participation‡‡‡27Table 1.4: Farmers† Level of Agreement with Decision-Making in Certified Farmer Groups(1•strongly to 5•strongly disagree)‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡...29Table1.5: Internal Audit and External Verification for Compliance‡‡‡‡‡‡‡‡‡‡‡30Table 1.6: Farmers† Assessment of Certification Requirements and Expected Benefits(1•strongly to 5•strongly disagree)‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡...32Table 2.1:Selected Householdand Farm Characteristics‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡.60Table 2.2: Binary Logistic Estimates for Determinants of Participation‡‡‡‡‡‡‡‡‡...64Table 2.3: Barriers to Farmer Participation in Cocoa Certification‡‡‡‡‡‡‡‡‡‡‡...66Table A.1: Sampling Framework‡‡‡‡‡‡‡‡‡...‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡77Table C.1: Details of Sample Weight Calculation‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡.81Table D.1:Description and Hypotheses of Regression Variables‡‡..‡‡‡‡‡‡‡...‡‡83Table E.1:Housing Characteristics‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡‡85Table F.1: Descriptive Statistics ofRegressionVariables‡‡‡‡..‡‡‡‡‡‡‡‡‡‡...87LIST OF FIGURESFigure 1.1: Farmers†StatedLevel of Knowledge of Cocoa Certification‡‡‡‡‡‡‡‡‡.24Figure 1.2:Farmers that know the name of the standard they are certified with‡‡‡‡‡‡..25Figure B.1: Location of Study Communities‡‡‡‡‡...‡‡‡‡‡‡‡‡‡‡‡‡‡‡..79KEY TO ABBREVIATIONSCAPIComputer Assisted Personal InterviewsCHEDCocoa Health and Extension DivisionCMCCocoa Marketing CompanyCOCOBODGhana CocoaBoardCRIGCocoa Research Institute of GhanaCSProCensus and Survey Processing SystemDOIDiffusion of InnovationFLOFair Trade Labeling OrganizationICSInternal Control SystemIFOAMInternational Federation of Organic Agricultural MovementsPPPsPublic Private PartnershipsQCCQuality Control CompanySAN-RASustainable Agricultural Network•Rainforest AllianceTIBTheory of Interpersonal BehaviorTPBTheory of Planned BehaviorTRATheory of Reasoned ActionUTZUTZCertifieINTRODUCTIONSimilar to third-party certificationprograms across agriculture and food systems, certification ofcocoa producers is a means topromotefarmers†adoption ofa comprehensive package oftechnical and institutional innovationsrecommended for sustainable cocoa production (TCC2010; KPMG 2012).Certificationprograms are gaining recognition within the internationaldevelopmentsettingat a fast rate because,in many cases,theyappear to addressconcerns aboutenvironmental andsocial issues associatedwith agricultural production(Saltiel, Bauder, &Palakovich 1994; Mandaet al.2015).In the case of cocoa certification,there are specificrequirementsintended to addressissues such as: soil, water, forest and biodiversity management;integrated pests and diseases management and crop protection;health and safety practices(including waste management);and workers† right (including issues on child labor)(CA 2013;FLO 2014;Mahrizalet al. 2012; UTZ 2009).An underlining principle ofcertificationprograms is that producers who participate in themtypically incur extra costs(as compared to those of conventional producers) associated withadopting sustainable agricultural practices. Thenthese programsrewardor compensate forproducers for those increased costs viahigher pricespaid for certified produce/productsand insome cases increasedlevels ofoutputs.In this way,marketforces areused aspart ofamechanismthatdeterminesand controlssome of the negativeexternalitiesof traditional (non-certified) practicesincluding those thatare difficult to manage through regulations.The marketfor produce/products,in turn,benefits throughincreasedproduct differentiationand creation ofnewmarket niches(Waldman & Kerr 2014;Lebel 2012;Mahrizalet al.2012).Privatecompaniesandinternational development organizationsare promoting the programwith thepromise of price premium as an incentive for participation(KPMG 2012; Melisa & Don 2012;TCC 2012).While some features of cocoa certification in Ghana appear to be visible anduniform, it turns out that in practice there is a high degree of variability and mostlyunsubstantiated claims about the program.Thefeatures ofcertificationprogramscan vary and be modified and as suchthere is theneedforresearch to deepenourunderstanding ofthe contexts and design elements that can increasecertification programimplementationin smallholder production systems(Gockowskiet al.2013).Likeothercertificationprograms,cocoacertification isvoluntaryandcertifies onlygroupsoffarmers whowant to be certified. Thusunderstanding what goes intofarmers†decisionsto participate in such groupsas well as the realities of their participation isveryimportant(Van Herzeleet al.2013).There isa small body ofliterature onthe topic of farmerparticipation in certification programs, including cocoa certification.Thisstudyaims to add tothat literature byfocusingondeveloping an understanding ofthe nature and extent of farmerparticipation in cocoa certification in Ghana.In order to achievethis objective, the study focusedon the followingspecificresearch questions:·In what ways and to what extent are farmersin Ghanaengaged inthe activities andprocesses of implementing cocoa certification?·What are stakeholders† assessments of cocoa certificationin Ghana?·What are the factors that determine participation in cocoacertificationin Ghana?The nexttwo chaptersarestructured asseparateresearch papers/manuscripts.Following thosechapters there is a chaptersummarizing someconclusionsand an appendixwith detailedresearch-related information and documents.CHAPTER 1FARMER PARTICIPATION IN AND STAKEHOLDER ASSESSMENT OFCOCOACERTIFICATION PROGRAMS IN GHANA, WEST AFRICAIntroductionVoluntarythird-partycertification of primary agricultural commodity producers is arecentstrategy through which the market is used to promote the adoptionof sustainable agriculturalpractices(Gonzalez & Nigh 2005; Auld 2010; Menozzi, Fioravanzi & Donati 2014; Teyet al.2014;).Acareful review of the history of voluntary third-party certification programsindeveloping countriesreveals that they wereoriginally used for plantation crops(Lebel 2012;Mahrizalet al.2012).Obviously, there are major differences in plantation production systemsand smallholder production systems that wouldhave significant implications on how third-partycertification would work in each case.Therefore,understandingthe contextofpromotingtheadoption ofsustainable agricultural practicesamong smallholder farmersthrough third-partyvoluntary certification wouldenhance the assessment of such programs.Empiricalevidence on these contextual circumstances is however lacking.This lack of evidenceexplainswhysome conditionsrelated to certification programs make it difficult to empiricallylink them with tangible benefits to farmers and society (Barham & Weber 2012; Elder, Zerriffi &Billon 2012).Issues regarding proliferation of certification standards and related multiplecertifications, inadequate marketsfor certified produce, and difficulty in the management ofprice premiums are important in this regard(Auld 2010;Hainmueller, Hiscox & Tampe, 2011;Barham& Weber 2012).Another feature of certifying smallholders thatmakes adifference isthe practice of groupcertifications.Carefully consideringgroup certifications raisessomequestions that have barely been addressed.This paperinvestigates somecontextual conditions for implementing third-party certificationprograms that promote smallholders† adoption ofsustainable agricultural practices. To achievethis, the paperexaminesthe participation of farmers in cocoa certification in Ghana as well asstakeholder assessment of the program. Cocoa certification is an example of voluntarythird-partycertification schemes that promote the adoption ofsustainable agricultural practicesamongsmallholdercocoafarmers. Cocoa producers are certified after they have been verified as havingadopted comprehensive packages of technical and institutional innovations recommended forsustainable cocoa production (TCC 2010; KPMG 2012).Cocoa production in Ghana is important to the world cocoa economy as well as the Ghanaianeconomy (Smithers & Furman 2003; Ntiamoah & Afrane 2008; Quarmineet al. 2012; GSS2013; Läderachet al. 2013). It is however confronted bysome economic, social, andenvironmentalissues that bring its sustainability into question (Ntiamoah & Afrane 2008; Kyei,Foli & Ankoh 2011; KPMG 2012; Melisa & Don 2012; Quarmineet al. 2012; TCC 2012;Darkwah & Verter 2014). For about a decade now, cocoa certification has been implemented inGhana to promote sustainable cocoa production (KPMG 2012; Melisa & Don 2012; TCC 2012).The lack of empirical research on the contexts of implementing certification programs asmentionedabove is very typical of the literature on cocoa certification. As a result,similarmethodological and other limitationshave made it difficult for research on cocoa certification tomake conclusive statements on its impacts(Gockowskiet al.2013; KPMG 2012; Melisa & Don2012; TCC 2012; and Verinaet al.2014).The Concept of Cocoa CertificationVoluntary standards and certifications have evolved through several means since the 1920s andhave been in several sectors including agricultural and food trade as a governance tool (Auld2010; Herzfeldet al.2011).For producer groups, standards and certifications are means ofestablishing acceptable production practices among members. They offer a bridge to theglobalized commodity chain and its standardized quality attributes for consumers. Businessesalong the commodity supply chain use standards and certifications to capture niches of producersand consumers in order to ensure consistent demand and supply (Lebel 2012;Mahrizalet al.2012). Generally, certification schemes train farmers, get them verified by third-party auditorsand then certify them by putting a label on their products to indicate adherence tosustainableagricultural practices. Businesses along the supply chain use theselabels to trace certifiedproducts in order to make claims on them to the consumer. Consumers then pay some extraamount (usually referred to as price premium) on such products to support farmers† efforts inadoptingsustainable agricultural practices(Elderet al.2012 & 2013; Jenaet al.2012; Rueda &Lambin 2013).Cocoa certification is a program that came as a result of initiatives by stakeholders within thecocoa economy to promote sustainable cocoa production (KPMG 2012; Melisa & Don 2012;TCC 2012).The program aims at alleviating the economic, social and environmental issuesconfronting sustainable cocoa production. Cocoa certification comprises a set of principles ofsustainable cocoa production practices spelt out in codes of conduct of various certificationschemes or standard bodies (KPMG 2012; TCC 2010). Cocoa is currently being certified inGhana by Fair Trade Labeling Organization (FLO), UTZ Certified, Sustainable AgricultureNetwork•Rainforest Alliance (SAN-RA), and International Federationof Organic AgriculturalMovements (IFOAM) (KPMG 2012; Mahrizalet al. 2012; Melisa & Don 2012; TCC 2012 &2015).In most cases, promoters of certificationpartner with licensed buying companiesand usetheexisting structuresof thelicensed buying companiesto organize farmers for groupcertifications. The group certifications are necessitated by the need to reduce transaction cost andinformation asymmetries as well as the need to consolidate verification responsibilities(Gockowski et al., 2013 and TCC,2015).A farmer organizationseeking certificationfirsthas to indicate its interestand get registered. Itthen trains its members on the requirements detailed in the codes of conduct for the particularscheme it wants to certify with. Most of the changes that result from the implementation of therequirements happen at the farm level. The organization therefore conducts aninternal auditthrough an Internal Control System (ICS) to check for its readiness for the certification. This isdone by visiting farms of members and checking for the various control points in the codes ofconduct. If the organization is convinced by theresults of the internal audit, it invites an externalauditor usually specified by the certification scheme to verify the group for certification. Thework of the external auditor results in either approval for granting certification and orrecommendations for improvements. Not all of the requirements are to be met before a group iscertified, groups usually have a 3-year transitioning period. After certification, groups areverified every 3 years to ensure compliance. Thelicensed buying companiesobtainpricepremiums through the liaisons with the standard bodies and pay them to farmers (Gockowskietal.2013; KPMG 2012;Mahrizalet al. 2012;TCC 2015).The State of Cocoa CertificationThe share of certified cocoa has been increasing steadily since the introduction of the program.In 2009, certified cocoa was about 3% of world cocoa sales, this rose to 6% in 2010 and 30% in2013. There are however issues ofmultiplecertification as wellas leakage (selling certifiedcocoa as conventional) that have the potential of affecting these figures (KPMG, 2012; and TCC,2010, 2012, & 2015). Fairtrade International and Fairtrade Africa†s report on West Africa for2014 indicates that there are 11 certified producer organizations in Ghana with 95,900 farmersproducing45,800 metric tonnes of certified cocoa on 146,800 hectares of land. Information onthe website of UTZ certified indicate that there are currently 16 producer groups certified inGhana.SAN-RA had about 7,000 farmers with about 34,000 hectares of land as of 2011 (Melisa& Don, 2012). KPMG, 2012 reported market shares of 39%, 25%, 20% and 15% forFLO, UTZ,SAN-RA, and organic certified respectively. Table 1 below shows the amount of cocoaproducedand sold as certified under FLO, UTZ and SAN-RA standards in 2009, 2011 and 2013.Though the characteristics of the production process for certified cocoa productionarenot toodifferent from the traditional process, the certification process alone comes with a lot of cost.This together with the small amounts of changes made in the production process and the costsassociated with them means the gains in cocoa certification need to be able to offset such cost inorder to get producers incentivized. Benefits to the producer in this sense depend on a sufficientand stable price premium and an increase in productivity (Waldman & Kerr 2014;Gockowski etal., 2013; andMahrizal et al., 2012). These have been the bases for most of the impactassessmentstudies conducted on cocoa certification. Generally, certified cocoa attracts a pricepremiumof$150 to $200 per tonne, which is about 10% of the price of cocoa. Variedproportions of this amount (depending on certification scheme and producer group) go into directpayments and services tofarmers, community development,and organization and administration.However, theredoes notseem to beagreat impact ofthepremium on the actual income offarmers because of transaction costs and the fact that some certified cocoa beans end up beingsold as conventional cocoa (KPMG, 2012;Mahrizal et al., 2012; and TCC, 2015).Table1.1: Trends in Production and Sale of Certified Cocoa (1000 Tonnes)200920112013FLOProducedSold(%)650(0)12446(37)17660(34)UTZProducedSold(%)50(0)21443(20)691297(43)SAN•RAProducedSold(%)130(0)9865(66)571279(49)AllProducedSold(%)830(0)436154(35)1438636(44)Source: TCC 2015Amidst methodological and otherlimitations, some evidence suggeststhat cocoa certificationleads to increased productivity (with the exception of organic certification which is associatedwith a decline in productivity especially in the initial years), capacity building and communitydevelopment. Specifically, better access to farm inputs, training, and credit as well as increasesin yield, productivity and product quality have been identified at the farm level. Financialviability, strengthened organizational capacity,and improved political representativenesshavebeen found at the group level. Improved livelihoods, improved labor conditions, efficient use ofagrochemicals, improved management of natural resources, conservation and restoration of localecosystems and biodiversity, group involvement in productive infrastructure, and generation oflocal employment have also been seen at the community level (Gockowski et al., 2013; KPMG,2012; Melisa & Don, 2012; TCC, 2012; and Verina et al., 2014). Despite these prospective gainsfromcertification, there are also a number of difficulties that have been identified. Significantamong them are the issues of discontent with premiums and pricing, credibility of auditing,functioning of farmer organizations, discrimination against …non-certifiable farmers†, persistenceof gender inequality, leakage and high compliance cost (KPMG, 2012; and TCC, 2010, 2012 &2015).MethodsThis section presents themethods employed in the study design,including thesampling ofcommunities andparticipants,and data collection and analyses.Study DesignThe study was designed to cover farmers in communities where any of the three majorcertification schemes currently involved in cocoa certification (FLO, UTZ and SAN-RA) isoperating exclusively. Cocoa certification is implemented in Ghana either through a farmercooperative or alicensed buying company. Therefore, the study was also designed to coverfarmers in communities where certification has been implemented through a farmer cooperativeor alicensedbuying company. This design allowsfor comparative data analysis across theschemes and implementation channels.During the study period, FLO, UTZ and SAN-RAstandards had been implemented in a total of 39 cocoa districts in 5 out of the 6 cocoa growingregions of Ghana. Communities in the study are located across 5 cocoa districtsin4 regions.There were 8licensed buying companies, 15 farmer cooperatives and 2licensed buyingcompany-cum-farmer cooperatives involved in the implementation of certificationandinformation was available for 6 of thelicensed buying companiesand the 15 farmercooperatives.Based on information from the 2010 population and housing census reports, the study wasdesigned to cover farmers in communities with estimated number of households between 200and 300. There is however one survey community that did not have this information in thecensus report. Again, in order to compare farmers participating in certification with non-certifiedfarmers where necessary, the study was designed to include certified and non-certified farmers.Data collection was done in two phases: a qualitative phase of in-depth interviews and focusgroup discussions and a quantitative phase of a household survey. Details of the samplingframeworkandthe geographical location of the study communitiesare presented and shown inAppendixes A and Brespectively.Sample and Sampling Techniques&ProceduresFarmers used for the study were selected from 6 communities, 2 each for FLO, UTZ and SAN-RA. The 2 communities for each standard consistof 1 inwhich implementationwas through alicensed buying companyand the otherthrough afarmer cooperative implementation. A total of16 in-depth interviews were conducted: 4 with Ghana cocoa board (COCOBOD) officials; 3withofficials oflicensed buying companies; 4 with certification standard officials; 2 with capacitybuilding organizations officials; 2 with officials of external auditing firms; and 1 with an officialof an international development organization. A total of 56 farmers participated in 6focus groupdiscussions, with 1focus group discussionper community. There were 352 farmers selected forthe household survey, with 30 certified and 30 non-certified farmers from each community(except one community whichhad only 22 certified farmers). With an 88.6% response rate, 312(150 certified and 162 non-certified) of themwere surveyed.Officials for the in-depth interviews were selected purposively. Purposive sampling was alsoused to select communities. This was necessary to allow for selection of communities in whichonly one certification scheme has been implemented through either alicensed buying companyor a farmer cooperative as well as communities that havebetween 200 and 300 households.Communities of such sizesare typically within well-defined geographic areas and this enhancesefficient comprehensive household enumeration.Participants in thefocus group discussions werealso selected purposively based on availability and willingness while taking into account factorslike gender, certification status, and social status (age, community leadership, migrant/nativeetc). Certified and non-certified farmers in the household survey were selected using stratifiedrandom sampling.Officials for the in-depth interviews were identified through snowballing, starting with contactsobtained from the cocoa research institute of Ghana (CRIG). Information on cocoa districts inwhich the three certification standards had been implemented was obtained from in-depthinterviews. Then 2 districts were selected per standard, 1 for each implementation channel. Foreach selected district, a list of communities in which certification had been implementedexclusively under a particular standard was generated. Information on the estimated number ofhouseholds in the listed communities was obtained from districts 2010 population and housingcensus reports. Upon arrival of the survey team in each selected community, there was anenumeration of all households in thecommunity. The enumeration exercise collectedinformation to identify cocoa farming households as well as certified and non-certifiedhouseholds. Then from strata of certified and non-certified households, a sample is selected forcertified and non-certified households in each community.Data CollectionOpen-ended interview and discussion guides were used to conduct the in-depth interviews andfocus group discussions. They covered issues on organizing farmers for certification, training offarmers, auditing and monitoring farmers for compliance, issuance of certificate, purchasing ofcertified cocoa beans, and distribution and management of premiums. The in-depth interviewswere conducted in English while thefocus group discussionswere conducted in Twi. They wereaudio-recorded and transcribed in English. Through computer assisted personal interview (CAPI)programmed using thecensusandsurveyprocessing system (CSPro), a structured questionnairewas used to conduct the household survey. The questionnaire collected information onknowledge, participation and assessment of certification. Field interviewers were trained on thequestionnaire and survey procedures in a four-day training and pretest program.Variables and MeasurementThree items where used to examine farmers† knowledge of cocoa certification: farmers† generallevel of knowledge of the program, their awareness of main objectives of the program, andcertified farmers† awareness of the standard they are certified with. Farmers† general level ofknowledge was measured on a 3-point Likert scale of 1-very knowledgeable, 2-somewhatknowledgeable, and 3-not knowledgeable at all. Farmers wereasked to indicate their awarenessof some main objectives of cocoa certification as well as the standards they are certified with.Farmers were also asked to indicate what they believe are objectives of certification.Farmers† participation in certifiedfarmer groupswas examined in relation toonly certifiedfarmers. It was examined using the time lag between the implementation of certification in acommunity and individual farmer†s participation in the program,farmers† engagement incertified farmer group decision-making,and their involvement in internal and externalinspection. Time lag was measured with number of years. Farmers† engagement in decision-making wasmeasuredusing typical mode for choosingcertified farmer group leaders,percentage of farmers holding leadership positions in their farmer groups,percentage of groupmeetings held in the 12 monthsleading up to the surveythat farmer attended,farmers† level ofagreement that group decision-making engages all members,and their level of agreement withgroup decisions. Levels of agreement were measured on a scale of 1•strongly agree to 5•strongly disagree for decisions on specific topics. The topics were sale of certified beans andprice premiums, membership payments and meetings, and internal and external inspection.Farmers† involvement in internal and external inspections was measured by asking farmers thelast time they and their farms were inspected by internal and external auditors.Farmers in the entire sample were asked to indicate(on a scale of 1•strongly agree to 5•strongly disagree) their level of agreement with some requirements and expected benefits ofcertification. The requirements were those regarding production, harvest and post-harvestpractices; fertilization and crop protection; safety practices and workers† right (including childlabor issues); environmental and natural resources protection and waste management; and farmerorganization. The benefits were improving access to inputs and services; increasing farm outputand income; improving environmental conditions;improving social conditions; improving farmmanagement and safety practices; and community infrastructure development. For the benefits,farmers were asked to indicate the extent to which they expected certification to result in suchbenefits as well as the extent to which they agree that certification has resulted in such benefits.Data AnalysesCoding of the qualitative data was used to analyse the results of the focus groups and in-depthinterviews.Descriptive statistics (percentages) were used to analyze all items under farmers†knowledge, modes of choosing group leaders, and farmers holding group leadership positions.Participation time lag, group meeting attendance, and levels of agreement with group decisions,certification requirements and expected benefits were analyzed with means. For participationtime lag and group meeting attendance, one-way ANOVA was used to test the difference inmeans across the certification standards. Differences in mean scores for levels of agreement withgroup decisionsacross the certification standards were tested using the Kruskal-Wallis test. T-test was used to test the differences in mean scores for the levels of agreement with certificationrequirements and expected benefits for certified and non-certified farmers.ResultsThis section presents the results of the study in two subsections: results from the in-depthinterviews andfocus group discussions, and results from the household survey.The firstsubsection focuses on findings regarding thestructures andprocedures for implementingcertification;the conduct of training and verifying/inspecting farmers for compliance;decision-making onprice premiums andinspectors;and sustainability assessment of the program. Thesecond subsection is focusedon farmers†knowledge of cocoa certification, their participation incertified farmer groupsand activities, and their assessment of certification requirements andexpected benefits.In-depth InterviewsandFocus Group DiscussionsStructures and Procedures forImplementing CertificationThe regulatory body of Ghana†scocoa sector, COCOBODmanagescocoa certification throughpublic-private partnerships (PPPs). Through these partnerships, COCOBOD, internationaldevelopment organizations,and private companies in and outside Ghana collaborate inorganizing farmers and taking them through the processescertification. There is currently nosingle division or unit of the COCOBOD thatis exclusively responsible for dealing withcocoacertification. However, the Cocoa Health and Extension Division (CHED), Cocoa MarketingCompany (CMC), Quality Control Company (QCC) and Research, Monitoring and Evaluationareunits and divisions of COCOBOD that areclosely related to different aspects of cocoacertification by virtue of their work. COCOBOD has started a review of all certification manualswith an aim of putting together one comprehensive manual. Also, COCOBOD is reviewingmarket dynamics and PPPs to come up with a document to regulate cocoa certification. Trainingand sensitization of COCOBOD staff on cocoa certification is also ongoing.Recently,COCOBOD instituted a regulation to retain 20% of the price premiums on all certified cocoabeans.Companies and organizations in the PPPs first assess the needs of communities to determinewhere to implement certification. Once a community isidentified, officials oforganizations andcompaniesapproach the leaders of the community to explain the concepts and objectives of theprogram. After community leaders† approval, there is community-wide sensitization. Cocoafarmerswho are 18 years and overare then invited to join farmer groups to be taken through theprocesses of certification. These farmer groups are mainlyfarmer cooperative associations,groups associated with alicensed buying company, or ahybridof the two.They are in somecases newly created and in other casespre-existing for other purposes.They are mostly formedin a cluster of communities, with each community having a group of farmers.In the communitiesused for this study, interested farmers have to payan average ofaboutGH µ201as one-timemembership registration feeandattend meetings once or twice a month.The 2 FLO communitieshave farmer groups that have been certifiedfor 2 and 5 years; the UTZ communities have groupscertified for 2 and 3 years; and the SAN-RA communities have groups certified for 1 and 2years.Certified cocoa farmer groups are run on democratic principles with emphasis on transparencyand awareness.An official of a certification standard puts it this way—†you must have membership list,theremust be leadership, there must be evidence thatthose elections are done according to democratic principles, transparency and stuffs likethat, and then the membership must be aware as to what is it that theyare signingonto,so there must be informationsome evidence that the people understand what[CERTIFICATION STANDARD]is about,what their responsibility in it would be,whattheirexpectationsof it should be at the basic level‡(IN020201). µ3.96Some officials mentioned that farmer groups organized underthe cooperative system do better inthis regard. As a result, they have stronger leadership and bargainingpower and theirmembersarewell-informedand own group decisions.Training and Compliance VerificationOnce acertification project is establishedand groups of farmers are formed, training of farmerson certificationrequirementandobjectives is done using the training-of-trainers(TOT)approach.Some farmers are selected from each grouptobepart ofan internalcontrol/management system(ICS/IMS)of the larger farmer cooperative orlicensed buyingcompany. These farmers are trained by capacity building organizations working withcertification and they in turn train their colleague farmers in their respective communities underthe supervision of the capacity building organization. This process of training is typicallyexpected tolastfora period of at least six months.After the training, the ICS/IMSplans for internal and external inspection. Internal inspectionsshould be conducted atleast once a year and all farmers in the groups are supposed to beinspected during every internal inspection. These internal inspections should be very welldocumented.The following are quotes fromalicensed buying companyofficialandacertificationstandard official respectivelyon internal inspection:—‡the group administrator will select a number of internal inspectors this could be hiredor could come from the same community or could come from the group who are train oninternal inspections protocols. These inspectors will then go around doing inspection ofeach member…sfarms which hundred per cent of the farms must be internally inspected tomeet the requirement– (IN020301).—So you do the internal inspection, you document it andthen if and auditor comes, he willbe able to identify that you are working with 1000 farmers this is their farmcharacteristics or farm profile– (IN020101).Internal inspections are done and corrective measures taken until the ICS/IMSis comfortablewith the results and then an auditor is invited to conduct external inspection. In some cases theICS/IMSfirst invites an auditor to do a pre-audit before the external inspection.Each groupshould be externally inspected at leastoncewithin a certification cycle, which takes 3 or 4years.External inspections check the entire system of operation of the farmer groups (documentreview, farm visits, key personnel interviews). From each farmer group, some farmers arerandomly selected for farm visits and other inspection activities. If external inspectors findnonconformities, they are expected to be corrected in what is termed —continuous improvement–before the farmer group can be certified. The importance of both internal and externalinspections was echoed in the words of a private consultant who does auditing:—Auditing formsone of the core pillars of certification scheme– (IN050301).Inspection/auditing is the feature ofcertification that actually allows for claims to bemade of certified beans as been produced withsustainable practices.DecidingonPrice Premiums andInspectorsFor price premiums, the focus was on decision-making regarding the amount paid and thedistributionamong various uses.It wasrevealed that price premiums are mainly negotiatedbetween locallicensed buying companiesand offshore cocoabuyers(mainlyprocessors). This istypically done prior to the implementation of aparticular certification project. In other words,licensed buying companiesfirst establish contracts with external buyers for specified amounts ofcertified beans at specific price premiums before they implement a certification projectin thatregard. Once alicensed buying companysuccessfully does this andreceives the money for pricepremiums, the decision on how muchto giveto farmers lies solely within the powers of thatlicensed buying company.Typically, it is alicensed buying companythat funds the cost ofimplementing the certification project(including the cases where implementation is donethrough a farmer cooperative). Therefore, thelicensed buying companyfirst deducts the costofimplementation and then decideshow much to give to farmers.Howmoney to farmersis distributed largely depends on the structures of the certified farmergroup. Typically, in the case of farmer cooperative-based groups, the bulk of money is paid tothe group. Then thegroupleaders(popularly referred to as lead farmers)together with membersdecideonhow muchtobe givento individual farmers as cash and how muchtogo into otheruses, mostly purchase of inputs and community projects.In the case oflicensed buyingcompany-based groups, because the groups are seemingly inseparable from thelicensed buyingcompanies, the bulk of money to be given to farmers still gets to be in the possession of thelicensed buying companies. They therefore decide(sometimes in consultation withlead farmers)on how muchtogo to individual farmers as cash and how much should go into other uses, stillmostly inputs and community projects.One focus group participant had this to say:—Our organization does not decide the amount of premium to be paid to farmers. Alldecisions regarding the amount to be paid are made by the officers from[licensed buyingcompany]. They set the amount to be paid per bag of cocoa e.g., this yearGH µ15.00was paid as premium per bag of cocoa–(FG030701).Regardinginspectors, the focus was tolearn about who decides onthe inspector.Internalinspectors are usually lead farmerswho are chosenby thelicensed buying companyor farmercooperative officialsbased on literacy and experience.According to a certification officer forone licensed buying company,—‡the lead farmer should be somebody who can read and write and should knowsomething about the work that he is doing, that is the farming, he should understand thecode of conduct and the training that he is supposed to do. So basically we don…t putemphasizes on certificate and those kind of things, is about readings, is about youunderstanding the code of conducts, is about you understanding the cocoawork–(IN030101).These internal inspectors typically inspect farmers outside of their communities of residence inorder to avoid conflict of interest.Farmer groups working with SAN-RA and UTZ certificationsget to choose from a number of accredited external inspectors (referred to as certification bodies•CBs or auditors).This is typically done in consultation with thelicensed buying companiesandother organizations involved.Private consultants employed and assigned by FLO-Cert, amember of the Fair trade group, externally inspect groups certified by FLO.Sustainability AssessmentFocusgroup discussionand in-depth interview participants echoedtwoissues mentioned inliterature as weaknessesandthreats to certifications.The first issuehas to do withtheproblem ofcertified beans been sold as conventional.The survey data for instanceshows that certifiedfarmers on average sold 13%to 88%of their certified beans as conventionalin the12 monthsleading up to the survey.Majority of the certified farmers in the survey indicated that they soldtheir certified beans as conventional because of their relationships withlicensed buyingcompaniesother than thelicensed buying companydesignated to buy certified beans in theircommunity.Other certified farmers indicated thatsometimespurchasing clerks of designatedlicensed buying companiesdidnot have money to pay for beans that were ready to be sold.Thesecond issue is what some participants described as unhealthy competition between companiesand organizations involved in certification, which results in the duplication of efforts andmultiple certifications.An official of an internationaldevelopment organization involved incertification implementation had this to say:—Certification I will always say is a good thing but the implementers should becareful,people are working for beans and not the message to the people who are suppose toreceive and implementthemfor us all to benefit– (IN050401).Some issues were also revealed as threats to the sustainability of cocoa certification in Ghana.Farmers in thefocus group discussionsindicated that certified group members†lack ofpunctuality to meetings is not encouraging;sometimes there is discrimination against somemembers in thedistribution of inputs to certified groups;and there is lack of frequent visits bycertification officials.Farmers and officials expressed concern with lack of transparency andaccountability on the part oflead farmersand somelicensed buying companies. A privateconsultant that conducts external auditing/inspection said this:—In my opinion, I think that most of these buying companies are into this programbasically because of profit but not promoting sustainability of the program. Most of theselicensed buying companiesfind it difficult in disclosing the cost incurred in preparing thegrower groups so in terms of transparency, it has always being a problem for most oftheselicensed buying companies.  Therefore, visibility has been a big problem.Yes I havewitness an internal officer cheating. Some give recommendations that do not reflect withobservations, field officers giving falsehood information…s, and checklist not properly fillout. I have personally caught a project officer filling in forms for the farmers–(IN050301).Other issuesrevealed includepaperwork being too much; inconsistent interpretation of somecertification requirement among different stakeholders; and apathy on the part of farmers.Also,it was mentioned that becausemost third-party certifications were originally designed forplantation crops;itisdifficult to implement some certification requirements.Some certificationrequirements are practicable only on a large scalebut cocoa production is predominantlysmallholder-based.Finally,some see the practice of making price premiums the majormotivation message as a threat to the sustainability of the program. Asa private consultant putsit: —Technically, I think the communication that certification is for premium was a disaster. Thebiggest threat is the wrong communication that certification is for premium– (IN050301).Asindicated above, farmers end up selling very high proportions of their certified beans asconventional and do not receive price premiums for such beans.This erodes the benefits of pricepremiumto farmersandreasonablyexplains why it isnot a very good idea to have itas a majormotivation.These weaknesses and threats not withstanding, farmers and certification officials believecertification has several positive implications and there are measures that can be taken to evenavert the weaknesses and threats.It is believed that certification is improving access to inputs;reducing the incidence of child labor; increasing cocoa outputs and incomes; and facilitatingcommunity developmental projects. Also, certification is considered a very good instrumentfordriving knowledge in farm management improvement much more than traditional extension isable to do. The practice of certifying farmers in groups alsoenhances complianceto some extent.Thisis because if a major non-compliance is identified for an individual farmer, it may affect theentire group†s certification. Therefore, farmers sometimes assume the responsibility of enforcingcompliance among colleagues.In moving forward, there are callsfor more cooperation betweencompanies and organizations involved to enhance database management and also reduceduplication of efforts. It is also important to pay attention to other sectors and aspects of thecocoa economy, especially marketingwhich isthe backbone of certification. Majority of cocoafarmers are still not captured under certification and officials see this as an opportunity fortheprogram to keep expanding.Survey ResultsKnowledge of CertificationThe results as showin figure 1.1below indicatethatabout 38% of farmers in the entire samplesaid they are very knowledgeable whileabout 39% and 22% respectively said they are somewhatknowledgeable and not knowledgeable at all. For certified farmers, 71% and 29% respectivelysaid they are very knowledgeable and somewhat knowledgeable. For non-certified farmers, thepercentages are respectively about 7, 49 and 42 for very knowledgeable, somewhatknowledgeable and not knowledgeable at all.The results for farmers† awareness ofcertification objectives are presented in table1.2below. Itcan be seen from the table that 5 out of the 8 objectives have at least 15% of farmers in the entiresample being aware of them. On top of the list is improving farmers† output and income, which62% of the entire sample, 92% of certified farmers, and 33% of non-certified farmers are awareof. Next is improving working conditions of farm workers and this 46%, 61% and 33% of theentire sample, certified and non-certified farmers respectively are aware of. For the entiresample, certified and non-certified farmers respectively, 29%, 39% and 19% said they are awarethat it is an objective of cocoa certification to eliminate child labor. The percentage of farmerswho are aware that conserving/protecting natural resources is an objective of certification is 25for the entire sample and 40 and 12 for certified and non-certified farmers respectively. Fifteenpercent of the entire sample, 22% of certified farmers and 8% of non-certified farmers are awarethat community infrastructure development is an objective of certification.Figure1.1: Farmers•StatedLevel of Knowledge of Cocoa Certification38%71%7%39%29%49%22%0%42%Entire Sample (310)Certified Farmers (150)Non-Certified Farmers (160)Farmers' Stated Level of Knowledge of CocoaCertificationVery KnowledgeableSomewhat knowledgeableNot knowledgeable at allTable1.2: Farmers• Awareness of the Main Objectives of Cocoa CertificationMain ObjectiveFarmers who are Aware (Percent)EntireSampleCertifiedFarmersNon-certifiedFarmersImproving farmers output and incomeImproving working conditions of farm workersEliminating child laborConserving/protecting natural resourcesCommunityinfrastructure developmentImproving access to inputsGeneral farmer educationFinancial benefits of companies involvedN624629251584431292613940224341503333191281274162Results of certified farmers† awareness of thename of thestandards they are certified with areshown in figure 1.2below. It can be seen from the figure that30% of farmers certified with FLOknow that it is FLO that certifies them. For UTZ certified farmers, 19% know that they arecertified with UTZand 24% of SAN-RA certified farmers know that they are certified withSAN-RA.For certified farmers in groups organized throughlicensed buying companies, 29%know the standards they are certified with. For certified farmers in groups organized throughfarmer cooperatives, 20% know the standard they are certified with.Figure1.2:Farmers that know the name of the standard they are certified withFLOUTZSAN-RALBCCooperativeFarmers that know the name of the standard they are certified withParticipationinCertified Farmer GroupsResults for the time lag between the implementation ofcertification in a community andindividual farmer†s participation in the program are presented in table1.3below.The first part ofthe table haspercentages of farmers in five categories of time while the second part has meantime lags.The 2ndto 4thcolumns contain the results for the certification standards and the lasttwo columns are for the implementation channels.For the subsample of certified farmers, themean time lag was about 0.76 years (about 9 months). As seen from the table, the mean timelagwas 0.63 years (about 8 months) for FLO certified farmers, 1.30 years (about 14 months) forUTZ certified farmers, and 0.17 years (about 2 months) for SAN-RA certified farmers.A one-way ANOVA (withF= 15.39 andp-value = 0.000) indicates that there are significantdifferences in the mean lag times for the different certification standards. However, the results ofthe ANOVA also showed that the variances for the different certification standards were nothomogeneous (chi-square =21.43 andp-value = 0.000 for the Bartlett's test for equal variances).The differences in the variances mightthereforeexplain why the means are also different.Also,the lag time forlicensed buying companyorganized certified farmers is0.8 years (about 10months) and for farmer cooperative organized farmers it is 0.72 years (about 9 months).A t-testof means show no significant difference in the time lag forlicensed buying companyorganizedand farmer cooperative organized farmers.Table1.3: Time Lag between Implementation of Certification and Farmer ParticipationTime LagPercentage of FarmersCertification StandardImplementation ChannelFLOUTZSAN-RALBCCooperativeWithin 1 yearAfter 1 year but less than 2 yearsAfter 2 years but less than 3 yearsAfter 3 years but less than 4 yearsAfter 4 yearsMean time lag (number of years)N484111000.6346205318441.204583170000.1724325711000.8054562612330.7261Note: LBC-licensed buying companyAn average of about 69% of all certified farmers indicated that leaders of their group aretypically chosen by vote of all members and 20% said that village leaders typically choose groupleaders. Some of the remaining 11% said external agents typically choose leaders and some saidleaders typically self-nominate. Across the 3 certification standards, an average of about 15% ofthe farmers hold leadership positions in their certified farmer groups.Certified farmersonaverageattended about 64% of meetings held by their certified farmers groups in the 12 monthsleading up to the survey. A one-way ANOVA revealed that there is no significant difference inthe percentage of meetings attended across the certification standards.Table1.4below presents the results of farmers† level of agreement with group decision-making.The first part of the table presents and compares the results across the certification standards andthe second part does so across the implementation channels.The results show that, in general,farmers agree to a high extent that group members are engaged in making decisions. This iswiththe exception of decisions regarding inspection for the case of SAN-RA certified farmersandboth implementation channels, wherethe mean scores approach neutral. All other topics havemean scores of less than 2 across thecertificationstandardsas well as the implementationchannels. However, it is important to mention that, the Kruskal-Wallis test indicated that thereare significant differences between the mean scores for decisions regarding membershippayments and meetings as well as internal and external inspections across the different standards.The test showed chi-squares of 9.611 and 10.97 andp-values of 0.008 and 0.004 for membershippayments and meetings and inspections respectively.Again, for decisions regarding sale ofcertified beans and price premiums, there is a significant difference between the mean scores forlicensed buying company organized farmers and farmersorganized through farmer cooperative.For farmers† level of agreement with group decisions, the results as shown in table1.4againindicate thatacross the certification standardsfarmers agree to a high extent with decisionsregarding all the specific topics. Mean scores for all topics across all the standards are all below2. The Kruskal-Wallis test indicated that there are no significant differences in the mean scoresacross the certification standards.Also, it is important to note thatin all cases,farmers† levels ofagreement with decisions are higher than their levels of agreement that members are engaged indecision-making (with the exception of decisions regarding membership payments and meetingsfor the case of FLO certified farmersand farmersorganized through cooperative). Though inallcases farmers generally agree to a high extent as mentioned.Table1.4: Farmers• Level of Agreement with Decision-Making in Certified Farmer Groups(1†strongly to 5†strongly disagree)ItemMeanScore (SD)FLOUTZSAN-RAMembers are engaged in decision-making regarding‡Sale of certified beans and price premiumMembership payments and meetingsInternal and external inspection1.7 (1.0)1.4 (0.6)**2.6 (1.5)**1.9 (1.1)1.5 (0.9)**1.9 (1.1)**2.2 (1.5)1.9 (1.0)**2.8 (1.3)**Farmer agreeswith group decisions regarding‡Sale of certified beans and price premiumMembership payments and meetingsInternal and external inspectionN1.3 (0.7)1.4 (0.8)1.9 (1.4)531.3(0.5)1.4 (0.8)1.7(1.1)471.2 (0.4)1.6 (0.9)1.8 (1.1)49LBC (A)Coop. (B)B-A (Std.Err.)Members are engaged in decision-making regarding‡Sale of certified beans and price premiumMembership payments and meetingsInternal and external inspection1.8(1.2)1.6(0. 9)2.3(1.4)2.1(1.2)1.5 (0.8)2.5 (1.4)-0.4* (0.2)0.1 (0.1)-0.1(0.2)Farmer agreeswith group decisions regarding‡Sale of certified beans and price premiumMembership payments and meetingsInternal and external inspectionN1.2 (0.5)1.4(0.7)1.9 (1.3)791.4 (0.6)1.5(0.9)1.8 (1.1)71-0.1 (0.1)-0.1(0.1)0.1(0.2)150Note:*** p<0.01, ** p<0.05, * p<0.1; LBC•Licensed Buying Company; Coop.•FarmerCooperativeRegarding inspections, the results asshown in table1.5below indicate that47%, 38% and 37%of farmers certified by FLO, UTZ and SAN-RA respectively have never been inspected by aninternal auditor. For FLO, UTZ and SAN-RA certified, internal auditors respectively lastinspected 44%, 46% and 39% of farmers within the last year. For external verification, it can beseen that 68%, 66% and 76% of FLO, UTZ and SAN-RA certified farmers have never beeninspected. Also, external auditors last inspected 28%, 26% and 22% of FLO, UTZ and SAN-RAcertified farmers respectively, within the last year.Table1.5: Internal Audit and External Verification for ComplianceLast InspectedPercentage of FarmersInternal AuditExternal VerificationFLOUTZSAN-RAFLOUTZSAN-RANever473837686676Within 1 year444639282622Within 3 years but more than 1 year ago91424482More than 3 years ago020000Note: N (FLO)=53, N (UTZ)=47, N (SAN-RA)=50Assessment of Certification Requirements and Expected BenefitsTable 1.6 presents results on farmers† levelof agreement with certification requirements as wellas expected and realized benefits. The second and third columns have the mean scores andstandard deviations for certified and non-certified farmers and the last column has t-tests of thedifference in the means and the standard errors.The results as shownin table1.6below indicatethatboth certified and non-certified farmers agree to some extent with all certificationrequirements, as all mean scores are less than 2. While all farmers generally agree with allcertification requirements, t-tests of mean scores indicate that the levels of agreement of certifiedfarmers are significantly higher than those of non-certified.Regarding expected benefits of certification, certified farmers generally agree to some extent thatthey expected certification to be beneficial in those specific ways(mean scores below 2.5).However, forthe expectation for community infrastructure development, they disagree to someextent (mean score 3.23). Non-certified farmers also agreed to some extent that they expectedcertification to improve environmental conditions and farm management and safety practices(mean scores of 2.46 and 2.43 respectively). Their levels of agreement are closer to neutral forthe expectations for inputs and services, farm output and income, and social conditions. Similarto certified farmers; non-certified farmers disagreed to some extent (mean score of 3.35) thatthey expected certification to bring community infrastructure development. Again, for allexpected benefits except community infrastructure development, t-tests of mean scoresindicatethat certified farmers† level of agreement are significantly higher than non-certified farmers†.For realized benefits of certification, the results show that both certified and non-certifiedfarmers generally agree to some extent that certification has improved outputs and income,environmental and social conditions, and farm management and safety practices (with meanscores less than 2.2). For access to inputs and services, farmers† level of agreement is closer toneutral: 2.79 for certified farmers and 2.60 for non-certified farmers. Both certified and non-certified farmers disagree to a high extent that certification has brought some communityinfrastructure development (mean scores of 4.46 and 4.49). T-tests of mean scores indicate that,certified farmers† level of agreement are significantly higher than non-certified farmers† forbenefits regarding outputs and income, environmental and social conditions, and farmmanagement and safety practices. For inputs and services, certified farmers† level of agreementis significantly lower than that of non-certified farmers. There is no significant differencebetween the mean scores of certified and non-certified farmers for community infrastructuredevelopment.It also interesting to note the differences between the levels of agreements for expected andrealized benefits. With the exception of community infrastructure development, all the benefitshavehigher levels of agreement for realization than for expectation and this is the case for bothcertified and non-certified farmers.This means that per the judgment of farmers, certification hasbeen more beneficial in terms of the listed items than farmers actually expected it to be.It is alsoimportant to that for the case of infrastructure development it is rather the reverse: farmersexpected certification to be more beneficial in terms of infrastructure development than it hasbeen so far.Table1.6: Farmers• Assessment of Certification Requirements and Expected Benefits(1†strongly to 5†strongly disagree)ItemMean Score (SD)B-A(Std. Err.Certified(A)Non-certified(B)Certification requirements regarding‡Safety practices and workers' rightsEnvironmental andnatural resource protectionand waste managementProduction, harvest and post-harvest practicesFertilization and crop protectionFarmer organization1.2 (0.4)1.2 (0.4)1.2 (0.4)1.2 (0.3)1.3 (0.7)1.5 (0.7)1.5 (0.6)1.6 (0.7)1.6 (0.8)1.8(1.2)0.3 (0.1)***0.3 (0.1)***0.3 (0.1)***0.4(0.1)***0.5(0.1)***Farmer expected certification to‡Improved farm management and safety practicesImprove environmental conditionsIncrease farm output and incomeImprove social conditionsImproveaccess to inputs and servicesBring community infrastructure development1.9 (1.3)2.1 (1.6)2.1 (1.3)2.2 (1.7)2.4 (1.2)3.2 (1.8)2.4 (1.6)2.5 (1.7)2.6 (1.8)2.6 (1.6)2.8 (1.5)3.4 (1.8)0.6 (0.2)***0.4 (0.2)*0.5 (0.2)***0.4 (0.2)**0.4(0.2)**0.1(0.2)Farmer believes certification has‡Improved environmental conditionsImproved farm management and safety practicesImproved social conditionsIncreased farm output and incomeImproved Access to inputs and servicesBrought community infrastructure developmentN1.4 (0.6)1.5 (0.7)1.6 (0.9)1.8 (0.8)2.8 (0.6)4.5 (1.1)1501.9 (1.0)2.0 (0.9)2.1 (1.1)2.2 (1.1)2.6 (0.9)4.5 (1.1)1600.5 (0.1)***0.5 (0.1)***0.5 (0.1)***0.4 (0.1)***-0.2(0.1)**0.03(0.1)310Note:*** p<0.01, ** p<0.05, * p<0.1DiscussionThis section discusses the results presented above with focus on farmers† knowledge ofcertification,their participation in activities of certified groups, and stakeholder assessment ofcertification.Thediscussion draws on the resultsofthe in-depth interviews andfocus groupdiscussions as well as the survey.Farmers• Knowledge of CertificationFarmers† knowledge of certification is important for two main reasons. First,some of theobjectives of the program and their respective requirements have very detailed information thatdemand high level of awareness and close attention.The fact that farmer groups have to undergotraining for atleast six months before they can be inspected and verified for certification speaksto this point.Second, certified farmer groups as seen from the in-depth interviews andfocusgroup discussions, are toberun ondemocratic principles and this means to a large extent thatfarmers should know what they are signing up for.In these regards, two points from thesurveyresults are worth noting:there is much room for improving farmers† knowledge of certification,andthis needapplies tocertified and non-certified farmers alike.It is not surprising to find certified farmersindicateknowing about certification and its objectivesmore thandonon-certified farmers. This point becomes more relevant in the next chapter wherelack of knowledge about certification comes up as one of the important barriers to participation.The focus here is rather on the fact that though certified farmers considerthemselves moreknowledgeable than non-certified farmers do, the level of knowledge over all is quite low. Thisshows upmore visibly when farmers† awareness of the objectives of the programas well asstandards they are certified withisconsidered.Even among certified farmers, as much as about 60% to 80% are not aware that issues likeeliminating child labor, conserving/protecting natural resources, and community infrastructuredevelopment are objectives of the program. This should be of concern considering the fact thatthese are issues that featurequite prominently in the rhetoric of certification.Anotherresult thatraises concern is that across certification standards andimplementation channels about 70% to80% of certified farmers do not knowthe standards they are certified with.The question here is,if farmers do not know things as common as names of standards, then how can we be sure thatthey know what they sign up for in joining certified farmer groups and will be able to implementthem?Participation inCertified FarmerGroupsGenerally, farmers† participation in decision-making and activities of certifiedgroups raiselittleor no concerns except for inspections for complianceand how price premiums are distributed.For instance, thetime lag between the implementation of certification in communityandfarmers† participation is notfar from expected as majority of farmers join groups within 2 yearsof implementation.Also, farmers† participation in terms of choosing leaders of groups andattending group meetings is at levels that reflect what would be expected.Certified farmers alsoagreethat theyare engaged in groups†decision-making and also agree with decisions thatcertified groups work with.Decisions that are taking outside of the certified farmer groups(specifically, decisions on inspectors and amount ofprice premium) are also very reasonable.Considering the nature of the market for cocoa beans and how premiums work,it is reasonablethatthe market determines the amount to be paid as price premium.However, as revealed in theresults, the distribution of the price premiums do not seem to go according to whatisstipulatedin the codes of conduct of the standards.According to the variouscodes of conduct, the net amount of money that certified groups are toreceive as price premiums is to be distributed among various uses per the decision of the groupmembers.However,from thefocus group discussions and in-depth interviews,this is not foundto be the case. In most cases,the licensed buying company in question solely takeseverydecision regarding price premiums.Though this might be considered as a smart way for thelicensed buying companiestomaximize the returns on their investments in the certificationprocess, it is also a violation of the codes of conduct of certification.Afterlicensed buyingcompanieshave deducted the costs they incur in the process of getting farmers certified, there isno reason for them to decide on whattodo with themoney that is supposed to go tothe farmers.Inthe ideal situation, even the amount of money spent on the certification process, which getsdeducted,should not be unknown to the membership of certified farmer groups.Thefocus group discussionand in-depth interview results show that certified farmer groups areexpected to conduct internal inspections of all members at least once a year and invite externalauditors at least once in a certification cycle (which is 3 to 4 years for all standards).Thesurveyresults however show that on average, about 40% of certified farmers say theyhave never beeninspected internally and about 70% say external auditors have never inspected them.Thesituation of external inspection does not raise much concern considering the fact that suchinspections are not expected to cover every farmer andthereare also some farmer groupsthatareyet to exhaust the entire certification cycle of 3 to 4 years. The situation of internal inspectionhowever calls for critical consideration.Asverywell established, compliance verification is verypivotal in the whole concept of certification and it is actually the major difference thatcertification makes in terms of getting farmers to adoptsustainable agricultural practicesbesidespremiums.It is thereforedisturbing to find that farmers are not being inspected as expected,especially considering the fact that certification is relatively new in the survey communities andthe morale is expected to still be high in such situations.Stakeholder Assessment of CertificationThe results on farmers† assessment of the requirements and expected benefits of certificationcoupled with the results on sustainability assessment by other stakeholders point out two mainthings. First, the requirements and expectations of certification are largely acceptable anddesirable.What this means is thatfarmersare likelydo what iswithintheirlimitsto makecertification work. This may be what shows up inthe results onhow farmers participate incertified farmer groups as discussed above.It is important to point out that farmers† levels ofagreement with requirements are even higher than with expectations. This means that it isreasonable to assume that farmers are not necessarilyparticipatingbecause of the expectedbenefits but that they are really committed to makingchanges to theirfarming practices.The second point is thatcertification,as has been touted, has many prospects to impact thesustainability of cocoa production. Meanwhile, there are equally important reasons why thesustainability of the program itself should be of concern to stakeholders in the cocoa economy.One such issue that demandsemphasisis the sale of certified beans as conventional. Asmentioned in the results, majority of certified farmers say they sell certified beans asconventional because of relationships they have with otherlicensed buying companiesother thanlicensed buying companiesthat buy certified beans in their respective communities. This issuealso shows up in chapter two below asanimportant reason why some farmers are notparticipating in certification.Meanwhile, the role oflicensed buying companiesin theimplementation of certification is so critical that it is not easy to immediately prescribe possibleremedies.However,a reasonable starting point could be a careful reconsideration of theroleoflicensed buying companiesinthe payment and distribution of price premiums.This is because ofthe close relation between selling certified beans to a designated licensed buying company andreceiving price premium.CHAPTER 2DETERMINANTSOFFARMERPARTICIPATION INCOCOACERTIFICATIONPROGRAMS IN GHANA, WEST AFRICAIntroductionIncreasing agricultural productivity and sustainable development ofan economy†sagriculturalsector are major drivers of beneficial structural transformation. Furthermore, they are strategiesbelieved to enhance living standards as well as social and environmental conditions  (Johnston &Mellor 1961; Antle & Diagana 2003; Kassieet al.2013). Both sustainable agriculturaldevelopment and increasing agricultural productivity are driven by technical and institutionalinnovations throughout the food and agricultural system (Staatz 1994). In Sub-Saharan Africa,traditional farming practices and geographical conditions have been confronting challenges toincrease agricultural productivity and to become sustainable because of low soil fertility and highincidence of pests and diseases. It appears that conventional technological and institutionalinnovations meant to mitigate the effects of low soil fertility and deal with pests and diseasestend to pose social, environmental, and health risks (Saltiel, Bauder, & Palakovich 1994; Mandaet al.2015).In response, several practices and technologies have been recommended to better address theapparent trade-offs between productivity and damage to society and the environment. Suchpractices and technologies are considered jointly as sustainable agricultural practices (Saltiel,Bauder, & Palakovich 1994; Teyet al.2013).Sustainable agricultural practices, in general, sharefour key attributes or features:they conserve natural resources;do not degrade the environment;are technically appropriate; andareeconomically and socially acceptable or desirable (Kotile &Martin 2000; Kassieet al.2013).Sustainable agricultural practicesbroadly include suchapproaches as precision farming, enhanced nutrient management, and water-related management.Some specific examples ofsustainable agricultural practicesare conservation tillage,intercropping and crop rotation; the use of improved varieties or breeds; and the use of animalmanure and organic fertilizer (Kassieet al.2013; Teyet al.2013; Mandaet al.2015). It has beenobserved thatsustainable agricultural practicesconserve and enhance natural resources, improvesoil fertility, sequester carbon, and increase farm output and incomes (Kassieet al.2013; Mandaet al.2015).As a result,sustainable agricultural practiceshave become an increasingly important componentof developmentpolicy because policymakers and development agencies appreciatesustainableagricultural practicesability to improve living standards as well as maintain or improve socialand environmental conditions (Kotile & Martin 2000; Teklewold, Kassie & Shiferaw 2012; Teyet al.2014).At first,sustainable agricultural practiceswere primarily promoted by public sectorand non-governmental organizations (NGOs). Recently, they have captured the attention ofagribusiness firms. Food industry organizations promotesustainable agricultural practicesadoption through private and voluntary certification schemes. These certification programsenable the participating food industries/businessesto demonstrate corporate social responsibility,differentiatethemselves fromcompetitors, and meet retailer and consumer demands (Teyet al.2014; Menozzi, Fioravanzi & Donati 2014).Despite the claimed benefits ofsustainable agricultural practicesandtheir promotion in manypartsof the developing world, their adoption and diffusion in Sub-Saharan Africa is only justbeginning. This lag ofsustainable agricultural practicesadoption has been attributed to imperfectinformation, input constraints, andmarket failures, among other reasons (Caviglia-Harris 2003;Kassieet al.2013; Teklewold, Kassie & Shiferaw 2012). As a result, researchers and policyanalysts have been keen to learn about how to design mechanisms to incentivize developingcountry farmers, especially in places likesub Saharan Africa, to adoptsustainable agriculturalpractices(Antle & Diagana 2003; Kassieet al.2013).There is limited empirical evidence on determinants ofsustainable agricultural practices†adoption in developing countries (Mandaet al.2015). This gap is particularly apparent forsituationswheresustainable agricultural practicesare promoted as part of private and voluntarycertification schemes. As has beennoted, many smallholder farmers do not have the means toimplement required changes and therefore certification programsmay appear todiscriminateagainst such …non-certifiable farmers† (González & Nigh 2005; KPMG 2012). A prime exampleof the use of private and voluntary certification schemes to promotesustainable agriculturalpracticesamong smallholder farmers is cocoa certification.Certification of cocoa producers is a relatively new strategy to promote cocoa farmers† adoptionof comprehensive packages of technical and institutional innovations recommended forsustainable cocoa production (TCC 2010; KPMG 2012). However, likesustainable agriculturalpracticesin general,a dearth ofempirical research has been undertaken to understand thecharacteristics of certified and non-certified cocoa farmersand the barriers totheirparticipationin cocoa certification schemes. This paper focuses on identifyingindividual, household, farm,andcertificationprogramcharacteristics associated with farmers† participation in cocoacertification programs in Ghanaas well as barriers to farmer participation in cocoa certificationprograms.Cocoa Production in GhanaGhana is a world leader in the quality of premium cocoa beans and the second largest producerof cocoa in the world (Smithers & Furman 2003;Ntiamoah &Afrane 2008;Quarmineet al.2012; Läderachet al. 2013).Cocoa production employs about 60% of Ghana†s agricultural laborforce and contributed to about 2.2% to Ghana†s Gross Domestic Product (GDP) and about 16.5%of its total export earnings in 2013(Ntiamoah & Afrane 2008; GSS 2013). Due to its importanceto the economy, the governmenthas principally controlled the cocoa sectorthrough the GhanaCocoa Board (COCOBOD) since the 1930s. COCOBOD serves the cocoa economy mainly withresearch and development, quality control, subsidized farm inputs, agricultural extension, and thebuying and exporting of Ghana†s cocoa beans. The Structural Adjustment Program of the 1980sresulted in the liberalization of the buying of cocoa beans in Ghana and gave rise tothe licensingof companies by COCOBOD to purchase cocoa from farmers. Since 2001, the responsibility offixing producer prices shifted from COCOBOD to a multi-stakeholder producer price reviewcommittee (PPRC),comprisingrepresentatives from farmers, COCOBOD, Ministry of Food andAgriculture (MOFA), and recently haulers andlicensed buying companies(Darkwah & Verter,2014; KPMG, 2012; and Quarmine et al., 2012).The demand for cocoa on the world market continues to rise with this increase predicted tocontinue (TCC, 2012). Increasingly, several issues continue to pose challenges to the sustainableproduction of cocoa in West Africa, where cocoa dominates agriculture activity (TCC, 2012).Cocoa production in West Africa is predominantly (80•90%) an endeavor undertaken bysmallholder farmers (below 3 hectares) (Darkwah & Verter, 2014; KPMG, 2012; Melisa & Don,2012; and TCC, 2012). As such, these farmers face several economic, environmental, and socialchallenges. These problems are complex, interrelated,and have persisted overlong periodsoftime.They includelow earnings from cocoa farming attributed to low farm productivity andresulting low living standards for cocoa farmers; inadequate social and public services in cocoagrowing communities; environmental degradation; unfair labor conditions; and the pervasive useof child labor in cocoa farming (Ntiamoah & Afrane 2008;Kyei, Foli & Ankoh 2011;KPMG2012; Melisa & Don 2012;Quarmineet al. 2012; TCC 2012; Darkwah & Verter 2014).Ghana†scocoa production is less efficient than cocoa production in other West African countries (basedon yield per unit area) (Kyei, Foli & Ankoh 2011).Cocoa Certification ProgramsStakeholder initiatives for sustainable cocoa production and increasing consumer concernsresulted in increased interest in cocoa certification programs (KPMG 2012; Melisa & Don 2012;TCC 2012). The goals ofcertification of cocoa producers include promotion of good agriculturalpractices, healthy and safe production practices, workers† rights, and natural resource andbiodiversity conservation. Certification programs are seen as appropriate means for participantsto achieve sustainable cocoa production inways that improve the livelihoods of cocoa farmersand their communities (TCC 2010; KPMG 2012). Currently, cocoa production is certified inGhana by one of four internationally recognized organizations and their standards: Fair TradeLabeling Organization(FLO), UTZ-Certified(UTZ), Sustainable Agriculture Network•Rainforest Alliance (SAN-RA), and the International Federation of Organic AgriculturalMovements (IFOAM) (KPMG 2012; Mahrizalet al. 2012; Melisa & Don 2012; TCC 2012 &2015).The four certification standards (FLO, UTZ, SAN-RA, and IFOAM) share the general aim ofpromoting sustainable production of cocoa throughsustainable agricultural practices, improvingfarmer livelihoods, and capacity building. However, each of these standards has strong affiliationwith particular thematic areas. FLO is more concerned with trade relations; UTZ and SAN-RAare more focusedonproductivity issues; and IFOAM focuses more on food safety. The generalrequirements of the different certification schemes are similar in most ways (KPMG 2012;Mahrizalet al. 2012). What sets the different certification schemes apart lies not in thecharacteristics of the production process but rather in their certification approach and programrequirements. Most of the elements required for certifiable production of cocoa predate theconcept of certification. For instance, 50% of Ghanaian cocoa farmers do not use agrochemicalsand are in essencede factoorganic producers. These cocoa farmers in Ghana do not grow cocoaorganically as a matter of choice but rather their circumstancesnecessitate it. Inorganicfertilizers, fungicides and other inputs are too expensive for these small-scale farmers and thereis minimal opportunity to access credit for farming (Gockowskiet al. 2013;Mahrizalet al.2012).Each certification scheme has its own code of conduct required for certification. These codes ofconduct specify detailed requirements onsustainable agricultural practices, cocoa communities,natural resources and biodiversity management, effective implementation, product flow, socialresponsibilities, and internal control systems (ICSs). The certifying agent at either the producerlevel or the group level checks specific details sometimes referred to as …control points†. Forsustainable agricultural practices, certification standards specify what producers have to doregarding cocoa farm establishment and rehabilitation, cocoa farm maintenance, soilmanagement and fertilization, integrated pest management and crop protection, harvesting, andpost-harvest handling. For community and social responsibility issues, there are specifics onhealth and safety production practices, worker†s rights, accountable and transparent managementstructures, education, and first aid and emergency health care(CA 2013; FLO 2014;Mahrizaletal. 2012; UTZ 2009).Under natural resources and biodiversity, issues regarding soil, water, and forest and biodiversityare specified. Specifics on organization, risk-based implementation and producer training arespecified under effective implementation and there are also specific details on product flowcontrol.Topicsregarding management of ICSs, ICS staffs, internal standards and contracts,internal inspections and registration of producers, and record keeping are specified under ICSs.All of the specifications are basedon the international labor organization (ILO) conventions andnational standards onsustainable agricultural practicesfor cocoa production. For organiccertification, producers are expected to follow all the requirements for FLO and cocoa should begrownon land that has been free of prohibited substances for three years prior to harvest beforeit can be certified (CA 2013; FLO 2014;Mahrizalet al. 2012; UTZ 2009).As mentioned above, the different standards vary in some of their requirements because ofthedifferent thematic areas emphasized by each program. A recent report by KPMG on behalf of theInternational Cocoa Organization looked at the similarities and differences of cocoa certificationprograms (KPMG, 2012). The KPMG report points out variations in characteristics such as: feesto certification organization, audits, price premium paid to the certified growers, distribution ofprice premiums to growers, biodiversity and climate change undertakings,waste disposal, andlimits/bans on genetically modified organisms (GMOs). For fees paid to the certificationorganization, FLO only charges initial group fees and annual fees while SAN-RA and UTZcharge those fees per quantity of certified beans. FLO charges annual audit fees whether or notthere was audit whiles SAN-RA and UTZ price annual audit fees according to thenumber ofaudits. FLO pays a fixed-price premium and audits groups† distribution and use of premiums.SAN-RA and UTZ pay price premiums to certified growers based on the market and do not auditthe distribution and use of premiums by recipient groups(KPMG, 2012).All the certification schemes use local and regional wage legislation or agreements except forSAN-RA, whichuses wage guidelines and allows for deductions for housing and food providedto growers. FLO has no specific requirements regarding use of shade trees, while SAN-RAspecifies shade tree requirements, plans for carbon emission and sequestration, as well asenergyuse guidelines. UTZ also has specific requirements on use of shade trees. FLO and SAN-RAhave waste disposal and storage guidelines. FLO allows burning of waste if it is in accordancewith local legislation but SAN-RA does not allow such burning under any circumstance. UTZhas no guidelines on waste disposal and storage except for specifying that waste storage anddisposal areas should be in central locations. FLO says GMOs should not intentionally be used,SAN-RA says no GMOs may be used, and UTZ does not have any guidelines on GMOs sincethere are no GMOs for cocoa (KPMG, 2012).In order to reduce transaction costs and information asymmetries as well as consolidateverification responsibilities, certification programs organize smallholder cocoa farmers intogroups for certification.  However, cocoa farmer groups/organizations are not well-developedand established in Ghana. For this reason,Ghanaianlicensed buying companieshave oftenassumed the role of implementing cocoa certification schemes and securing farmers† cocoabeans. Thelicensed buying companiesbuy the certified cocoa beans and pay price premiumsthrough their liaisons with the standard bodies (Gockowski et al., 2013 and TCC, 2015).In general, to become a certified cocoa producer,cocoa growers must adhere to the production,administrative and social standards promulgated by the particular certification standard.  Theprocesses for cocoa certification are similar across all of the schemes and typically focus onregistration and training, compliance, and audit. Generally, the participating organizations andtheir members do not have tofulfilall of the certification requirements in the first year. There isusually a transitioning period of three or four years. The certification schemes typically havesome specific requirements that need to be met before certification can be granted and othercriteria that may be met over the transition period. Once a group is certified, there is generallyadherence verification every 3 years to ensurecompliance (Gockowski et al., 2013; KPMG,2012;Mahrizal et al., 2012;and TCC, 2015).Previous Studies ofParticipation in Sustainable Agricultural Practices ProgramsMany studies of adoption ofsustainable agricultural practiceshave been grounded onbehaviouraltheories developed by rural sociologists and psychologiststargetingimproveddissemination of technical and institutional innovations developed to increase agriculturalproductivity (Neill & Lee 2001; Padel 2001; Van Herzele et al. 2013; Teyetal.2014). Previousstudies have, to varying degrees, integrated the theory of diffusion of innovation (DOI), thetheory of plannedbehaviour(TPB), the theory of interpersonalbehaviour(TIB), the theory ofreasoned action (TRA), the pest-believe theory,and the structuration theory (Smithers & Furman2003; Teyet al.2014).Most previous work has focused on adoption decision-making with a fewstudies looking at abandonment of adopted innovation (Neill & Lee 2001; Läpple 2010; Sahm etal. 2013).In light of the previous research, this paper considers Ghanaian cocoa farmers† behavior fromtwo perspectives. First, we consider farmers in the position of comparing the characteristics andperceived benefits of conventional agricultural production for cocoa with the characteristics andperceived benefits of growing cocoa usingsustainable agricultural practices. We assume thatfarmers choose the production approach (conventional orsustainable agricultural practices) thatmaximizes their expected utility, conditional on their decision (Kassieet al.2013). Second, weconsider farmers as having some volitional control over theirsustainable agricultural practicesadoption decision. Some previous work suggests that a farmer†s decision to adoptsustainableagricultural practicesis influenced by what the farmer thinks of his/her capacity to adopt,theinnovation costs and benefits,and consideration of societalnorms (Martínez-García, Dorward, &Rehman 2013; Borges et al. 2014; Jorgensen & Martin 2015).These perspectives of farmers† decision-making form the basis of previous studies† integration ofmultiple theoretical bases insustainable agricultural practicesadoption studies. Similar to thecase of cocoa certification, the integration of Rogers† theory of DOIand Triandis† TIB wassuccessfully applied by Telet al.(2014). It turned out that the theory of DOI was useful inunderstanding farmers† perceptions of the attributes of cocoa certification while use of the TIBaccommodated the roles of social factors, emotions, and behavior (a function of intention,habitual responses, and situational constraints) (Teyet al.2014).In line with the theory of DOIand TIB, six core sets of factors have been identified by empirical research as influential infarmers† decision-making regarding program participation. These factors include: 1) socio-economic factors, 2) agro-ecological factors, 3) institutional factors, 4) informational factors, 5)psychological factors, and 6) the nature and special qualities (attributes) ofthe program inquestion(Padel 2001; Smithers & Furman 2003; Teyet al.2014).Previous studies have measured program participation as a binary response variable and as afunction of other varaibles that represent elements of the six core sets offactors (Neill & Lee2001;Pietola & Lansink 2001; Sahmet al. 2013; Teyet al.2014). —Participation– has typicallybeen measured by respondents† actual participation/non-participation or by respondents† statedintentions/readiness/willingness to participate or not(Smithers & Furman 2003; Van Herzele etal. 2013; Aidoo & Fromm 2015; Meijeret al. 2015). The elements in the sets of socio-economic,agro-ecological, institutional, and informational factors in these studies are hypothesized toeither impede or facilitate participation.The set of program attribute factors constitute a form ofsubjective evaluation, which is likely to yield participation when positive. Habitualpsychological factors indicate established participation status while intentional psychologicalfactors indicate the strength of the willingness to participate or continue participation (Teyet al.2014).In previous studies, socio-economic factors are usually considered at the household andindividual (household head or main farmer) levels. Household characteristics consideredtypically include households† family labor availability (e.g. household size, number of adults,number of children) and households† dependence on the farming activity(s) concerned with theprogram (e.g. percentageof household income from activity, stability of income sources, non-farm income). Individual characteristics usually considered are principally those that indicatehuman capital (e.g. health, experience/age, education) and psychological factors (e.g. riskattitudes, perception of farming, profit orientation, environmental and social attitudes). Agro-ecological/farm characteristics considered usually are those related to the productivity orefficiency of the farm production process such as farm size, land tenure, source of farm labor,input use, and soil characteristics. Informational factors usually considered include contact withextension agents, use of information media, learning from others, and learning by doing (Feder,Just & Zilberman 1985; Neill & Lee 2001; Moser & Barrett 2006; Läpple & Rensburg 2011;Namome 2013).As previous research demonstrates, these factors can be incorporated in a subjective utilitymodel, where farmers are likely to participate when the subjective utility of participating isgreater than that of non-participation.If a farmer produces an output (Yc), under cocoacertification with (Xc) conditions, he derives a utility (Uc), which is a function of the conditionsand outcomes of the certified cocoa production process: i.e.Uc=f(Xc, Yc). On the other hand,another farmer who produces an output (Ync) under conventional cocoa production would haveher conditions of production (Xnc) that yield a utility (Unc) as a function of her output andconditions: i.e.Unc=f(Xnc,Ync)(Moser & Barrett 2006; Herzfeld & Jongeneel 2012; Dabbert,Lippert, & Zorn 2014). Because producing certified cocoa means incurring some costs that aredifferent from the costs of producing cocoa conventionally, farmers would typically considerwhat they have to invest in light of their expected outcome (Yc) as a result of their investment.Whether or not a farmer decides to participate in a program depends on whether or not theexpected outputs make sense in light of the cost of their inputs. Previous studies compare thedifferences in costs and outputs of alternative programs to understand adopters and non-adoptersof alternative programs(Gedikoglu & McCann 2012).Previous studies onsustainable agricultural practicesprograms have characterized participationin such programs as a static activity. This has however been criticized and some flaws have beenidentified. For example, some initial adopters might disadopt while some initial non-adoptersmight adopt eventually. Results from static conceptual models analysis can therefore bemisleading, as they would yield biased coefficients if the participation process were incomplete.Also, some variables that are treated as independent variables in such static models do changeover time. It is however impossible to control for such changes and how they affect participationover time (Feder & Umali 1993; Neill & Lee 2001; Moser & Barrett 2006).These issues notwithstanding, the use of static models toanalysecross-sectional data in studyingparticipation and adoption continues to dominate the literature because collecting panel data hasproven to be expensive, especially in developing country contexts. Also, governments and otherpolicy makers do not usuallyfindex postconclusions very useful. Furthermore, analysis ofcross-sectional data has been useful in providing meaningful insights into the participationprocess. Static analysis of cross-sectional data helps to identify barriers to participation as wellasgroups that find it difficult to participate. This then facilitates the design of appropriateinterventions to ensure adequate participation to yield optimum results for programs (Moser &Barrett 2006). Another important issue regarding the study of the adoption ofsustainableagricultural practiceshas to do with the fact that they usually come in the form of a package ofdifferent technical and institutional innovations. It is therefore important to consider the inter-relationships between the different components and their adoption decisions, especially whenthey can be adopted individually and their adoption decisions are not made exogenously.Otherwise, the influences of various factors on the adoption decisions may be under or over-estimated (Kassieet al.2013).MethodsThis section presents the methods employed in designing the study, selecting communities andhouseholds for the study, collecting andanalysingdata, and estimating parameters.Study Design and SettingThe study was designed to cover farmers in Ghana†s cocoa growing communities where cocoacertification has been implemented under FLO, SAN-RA, or UTZ standard. In Ghana, cocoacertification is implemented either through alicensed buying companyor a farmer cooperative.Therefore, the study was designed so that in half of the communities in the study certificationimplementation is throughlicensed buying companiesandin the other halfcertification isimplemented through farmer cooperatives. The communities in our study were purposelyselected to be roughly the same size, with approximately 200 to 300 households, in the 2010population and housing census, except for one communityfor whichinformation on the numberof households was not available. Participants in the study included both certified and non-certified cocoa farmers.Data collection was done in two phases; first were in-depth interviews and focus groupdiscussions, and second was a household survey. At the time of field data collection,8licensedbuying companies, 15 farmer cooperatives and 2 farmer-cooperatives-cum-licensed buyingcompanieswere implementingthe FLO, SAN-RA, and UTZ standards in Ghana. Six of thelicensed buying companiesand the 15 farmer cooperatives wereimplementing cocoacertification in a total of 39 cocoa districts in 5 out of the 6 cocoa growing regions. Thecommunities used for the study are located in 5 cocoa districts across 4 regions.Appendix Apresents the number of districts, the number of communities in selected district, and the numberof households in selected community for each certification standard and implementation channel.Appendix Bis a map of Ghana showing the location of the study communities.Sample,Sampling Techniques & Procedures, and SampleWeightsThe study was conducted in 6 communities, 2 each for FLO, UTZ and SAN-RA.For each of thestandards, 1 of the communities had certification implemented through alicensed buyingcompanyand 1 using farmer cooperative. In each community, afocus group discussionwasconducted and a total of 56 cocoa farmers participated in thesefocus group discussions. Again,in each community, 30 certified and 30 non-certified farmers were sampled, with the exceptionof one community where only 22 certified farmerswere enumerated and all of them included inthe sample. This resulted in a total of 352 certified and non-certified farmers selected for thehousehold survey. A total of 312 households (150 and 162 certified and non-certifiedrespectively) were interviewed, representing an 88.6% response rate.Communities for the study were selected on purpose to ensure that each had only onecertification standard implemented through a particular channelandapproximately200 to 300households.Focus group discussionparticipants were also recruited on purpose based on theiravailability and willingness, gender, social status (age, leadership, migrant/native etc), andcertification status. Stratified random sampling was used to select certified and non-certifiedhouseholds for the survey. In-depth interviews with officials working with cocoa certificationhelped to identify cocoa districts in which the various standards have been implemented. Twodistricts where then selected for each standard, 1 with implementation throughlicensed buyingcompanyand the other through farmer cooperative. Then 1 community was selected from eachselected district. In each community, all households were enumerated and cocoa farminghouseholds were grouped into certified and non-certified.Households were then selected for thesurvey from the strata of certified and non-certified farmers.Certified and non-certified households were not equally represented in the sample population.Selecting equal numbersfromeach category of householdsfor the sample therefore resulted inunequal probabilities for certified and non-certified households.Sampleweights were thereforecalculatedwith detailed information from the enumeration of householdsand applied in theestimation of the probit regression model.Details of the calculation of sample weights arepresented in Appendix C. the total number of cocoa farming households enumerated was knowfor each community as well as certified and non-certified households.Errorsin the classificationof households duringenumeration were tracked and corrected for. The corrected populationswere used to obtain population proportions for certified and non-certified households. Theproportions of certified and non-certified households in the sample were also calculation withinformation from the survey data.Weights were then obtained as a ratio of populationproportions to sample proportions.Data CollectionAn open-ended discussion guide that included questions on the barriers to participation in cocoacertification was used to conduct thefocus group discussions.Focus group discussionswereconducted in Twi, audio recorded and transcribed in English. The household survey wasconducted with the use of a structured questionnaire through computer assisted personalinterviews (CAPI) programed with the survey and census processing system (CSPro). Theinstrument was designed to collect information on certification status, reasons for not joiningcertification as well as farm, farmer and household characteristics. A four-day training andpre-testprogram was organized to equip interviewers with the necessary skills, knowledge andinstructions to help complete the survey as well as to refine the survey instrument.Variables and MeasurementsSampleCharacteristicsHousehold income is the total amount of money a householdreceived from all sources in the last12 months leading up to the surveymeasured in Ghana Cedis.Household size is the total numberof people in a farmer†s household. Children in schoolindicatethe percentage of householdmembers 5 to 17 years old that are currently enrolled in school.Child educationis theaveragenumber of years of formal schooling completed per child.Child absenteeismis the number ofdays (out of 10) a childmissed school in the last 2 weeks.Number of farms is the number ofseparate pieces of land a household has under cultivation; total farm size is the total amount ofland (measured in acres) that a household has under cultivation; and cocoa farm share is thepercentage of the total farm holdings the is planted with cocoa.Cocoa output is the amount ofcocoa beans a household harvested in the last 12 months leading up to the survey. Potentialoutput lost is the additional amount of cocoa beans a household could harvest if not for loses dueto drought and flood, pests and diseases, and bush fire.Regression VariablesAppendix Dpresents description and hypotheses of regression variables.Farmers†participationin cocoa certification wasmeasured with a binary response of 1and 0 for certified and non-certified farmers respectively.Certification standard was measured by whether a farmer†scertified group is certified by FLO, UTZ or SAN-RA, and dummies were created for eachstandard. Implementation channel was measured by whether a farmer†s certified group isorganized through a licensed buying companyor farmer cooperative, and dummies were createdforeach channel.Genderis measured as dichotomy with0forfemaleand1 for male. Migrant indicates whether ornot a farmer was born outside his/her community of residence andhas1for migrantsand0fornon-migrants.Age is a farmer†s age measured in completed years and age squared is a squaredterm of a farmer†s age.Education indicates the number of years of formal schooling completedby a farmer.Leader indicates whether or not a farmer holds a leadership position in his/hercommunity; household member leader indicates if a farmer has at least a household member whoholds a leadershipposition; and acquaintance leader indicates whether or not a farmer has at leastan acquaintance that holds a leadership position. For all three variables, 1 is for the affirmativeand 0 is otherwise.Number of adultsis the people in afarmer†s householdthat are 18 years old and aboveandnumberofchildrenindicatesthosethat are from 5 to 17 years old. Income from cocoa indicatesthe percentage of a farmer†s household†s income in the 12 monthsleading up to the surveythatwas obtained from cocoa farming. Cocoa farm size is the totalamountof land that a farmer†shousehold had undercocoacultivation at the time of the survey and it is measured in acres.Sharecropped measures the percentage ofthetotal cocoa farmland that is cultivated undersharecropping arrangements.Hybrid cocoa variety measuresthe percentage of the number ofafarmer†shousehold†scocoa farms established,at least in part,with hybrid cocoa variety prior tothe introduction of cocoa certification. Lining and pegging measures the percentage of thenumber ofa farmer†shousehold†scocoa farms established,at least in part,using lining andpegging prior to the introduction of cocoa certification.Data Analysis and Estimation ModelDescriptivestatisticswere used to analyze and present data and results on sample characteristics,the barriers to farmer participation, as well as variables in the regression model.Abinaryprobitregression model was used to estimate the effects of selected individual, household, and farmcharacteristics on whether or not a farmeris a member of a certified farmer group. The classicalsubjective utility maximization model motivates the regression model. Let†s consider a latentvariable Yi*, which represents the utility behindafarmer†sdecision to join a certified farmergroup.Yi*is a linear function ofprogram characteristics as well asthe farmer†sindividual,household and farmcharacteristics:+, where Xiis a vector ofprogram, individual,household and farm characteristics,is a vector of regression coefficients andis a normallydistributed error term.is not observed, but the outcome( which is the farmer†smembershipof a certified farmer group) is determined by the value ofas follows:if> 0 andif< 0.The probability thattakes either 0 or 1 is expressed as follows:Whereis the cumulative distribution function (CDF) for the standard normal distribution.This implies that:. The parametersandalways appear togetherand cannot be reported separately.  Only the ratiosappear in the regression results. Theseratios are not the marginal effects. However, post-estimation prediction of the marginal effectswas obtainedwithSTATA.ResultsThis section presents the findings of the study inthreesubsections. The first subsectionpresentsselected household and farmcharacteristics with descriptive statistics and t-test of means.Thesecondis on the determinants of participationandhasdescriptive statistics of the regression*=bb=1=0Pr(=1)=Pr(*>0)=Pr(b+e>0)=Pr(e>-b)=1-(es<-bs)=1-F(-bs)=F(bs)FPr(=0)=1-F(bs)bsbsvariablesandthe binaryprobitregression estimates.Thethirdsubsection presents barriers tofarmer participation with descriptive statisticsandinsightsfrom thefocus group discussions.Characteristics of Survey SampleTable 2.1below presentsselected household and farm characteristicsforthecertified and non-certified subsamples.The first part of the table presents means and standard deviations of somevariables (2ndand3rdcolumnsfor certified and non-certified householdsrespectively) andt-testsofthe differences in the means (4thcolumn).The second part of the table presents percentages ofhouseholds in various categories of the remaining variables for the entire sample (2ndcolumn),certified subsample (3rdcolumn), and non-certified subsample (4thcolumn).In addition to thecharacteristics presented in table 2.1, Appendix E shows details of some housing characteristicsof the survey sample.Across the entire sample, household annual income in the last 12 months isaveragelyGH µ456and has no significant difference between certified and non-certified households.The entiresurvey sample has a mean household size of 4.41 andit can be seen from the table that thedifference between the means for certified (4.69)and non-certified (4.14)is statistically differentfrom zero.An average of about 91% of children in households in the entire sample are in schooland there is no significant difference between the means of these percentages for certified andnon-certified households.Achild in a household in the entire samplehas on average completedabout 6 years of formal schooling and againthisdoes not matter whether the child is in acertified or non-certified household.Across the entire sample, a child in a household missed onaverage about half a day of school out of 10 schooldays in the last two weeks.Most childrenmissed school either because school was on vacation(50% to 59%), or they were disabledorill(19% to25%), or they are not interested in school(13% to 22%).It might be interesting to notethatlack of interest in school as a reason for missing school days is 9% more in certifiedhouseholds.Alsoworthy of noteis that work on household farm only accounts for 2% of reasonsfor absenteeism across households in the entire sample, 3% in the certified subsample and nonein the non-certified subsample.On average, a household cultivates about 2.3 pieces of farmlandand there is a significantdifference between the mean number of farmscultivated by certified(2.46)and non-certifiedhouseholds(2.14).Thetotal size of cultivated farmland averages at about 10.5 acres for theentire sample and also has a significant difference between certified and non-certifiedhouseholds.Averagely, about 90%ofahousehold†scultivated farmlandisplantedwith cocoaand this isverysimilarfor each category of households.Ahousehold on average harvested about1084kg of cocoabeans in the last 12 months and this also has no significant difference betweencertified and non-certified households.An average of about 406kg (about 437 and 376 for certified and non-certified householdsrespectivelywith no significant difference)of potential cocoa output was lost to pest anddiseases, drought and flood, and bush fire in the last 12 months.Regarding the status of cocoaoutput over the last 5 years,37% of farmers(41% and 35% for the certified and non-certifiedsubsamples respectively)believe that their households† cocoa output has been about the same.These percentages are similar for farmers whosay their output has been increasingand in bothcasestheyare higher for certified farmers.However, 33% of non-certified farmers believe theiroutput has been decreasing as compared to 17% of certified farmers.Table2.1: Selected Household and Farm CharacteristicsVariableMean (SD)B-A(Std. Err.)CertifiedFarmers(A)Non-certifiedFarmers(B)Household CharacteristicsHousehold income (GH µ)7977 (6618)8900 (12224)924 (1126)Household size4.7 (2.3)4.1 (2.3)-0.6(0.3)**Children in school (%)90.2 (23.7)91.9 (21.1)1.8(3.1)Child education (years)6.0 (2.6)5.5 (2.6)-0.4(0.4)Child absenteeism (days/10)0.4 (1.1)0.5 (1.3)0.1(0.2)Farm CharacteristicsNumber of farms2.5 (1.2)2.1 (1.1)-0.3(0.1)**Total farm size (acres)11.6(9.2)9.6 (10.0)-2.0(1.1)*Cocoa farm share(%)90.4 (16.8)88.1 (19.9)-2.3(2.1)Cocoa output (kg)1154 (1148)1021 (1184)-133 (132)Potential output lost437.2(746.2)376.4(724.2)-60.8(85.2)N150162312PercentageEntire SampleCertifiedNon-certifiedReason for Absenteeism(last 2 weeks)VacationDisability/IllnessNot interested in schoolCannot afford schoolFamily does not allow schoolingWork on household farmBad weather552217222250192233305925130003Cocoa Output Status (last 5 years)About thesameIncreasingDecreasing373726414217353233N312150162Note:*** p<0.01, ** p<0.05, * p<0.1Determinants of Farmer ParticipationDescriptive StatisticsDescriptive statistics of the independent variables in thebinaryprobitregressionmodel arepresented in Appendix F.In accordance with the studydesignthe sample is almost equallydistributed among the three different certification standards as well as thetwochannels ofimplementation. The entire sample as well as the certified and non-certified subsamples eachconsists of approximately 78% male farmers.Less than a third of the farmers in the sample (26%-entire sample,22%-certified subsample, and 30%-non-certified subsample) were bornoutside of their current community of residence. Approximately 27% of the farmers in the entiresample hold leadership positions in their communities, and this is 37% and 19% for the certifiedand non-certified subsamples respectively. About 8% offarmers haveat least a member of theirhousehold being a community leaderandapproximately 44%have an acquaintance that is acommunity leader. The mean ageis about 47 years and the average farmer has completed about8 years of formal schooling.On average there are about 2.5 adultsand 1.5children in a farmer†s household. A household onaverage earned about 65%ofitsincome in the 12 monthsleading up to the surveyfrom cocoa.The average size of land a household has under cocoa production is 9.34acres; about98% of thenumber of cocoa farmsa householdcultivates areplanted at least in part with hybrid cocoavarieties; averagely,99.5% of the number of cocoa farms a household cultivates is planted atleast in partusing lining and pegging; and ahouseholdcultivatesabout 31%ofitstotal cocoafarm holdings under sharecropping arrangements.Regression EstimatesTable 2.2below presents results of the binaryprobitregression of the determinants ofparticipation.The first column ofthe table contains the explanatory variables, columns 2 and 3have the regression coefficients and their respective robust standard errors, and the 4thand 5thcolumns have the average marginal effects and their respective robust standard errors.Thepresentation of the findings focuses on the average marginal effects in column 4.The number ofobservations, Wald Chi-square, and the Pseudo R2are presented at the bottom of the Table.Certification Standard and Implementation ChannelAs can be seenfrom table 2.2, the certification standard implemented in a farmer†s communitydoes not affect the probability of the farmer being certified. The results show that farmers incommunities where certified farmer groups are organized through farmer cooperatives havelower probabilities of being certified. On average, the probability of a farmerin a communitywith cooperative-organized certified farmer group to be certified is about 0.08 less than a farmerin a community where the certified farmer group is organized through alicensed buyingcompany. This is found to be significant at the 10% significance level.IndividualCharacteristicsThe regression estimates indicate that a farmer†s gender, migration status, and years of educationdo not affecttheprobabilityofthe farmerbeing certified. Also, whether or not a farmer has ahousehold member or a close acquaintancethatis a community leader doesnotaffectthefarmer†s probability of beingcertified. There is however a significant and nonlinear relationshipbetween the age of a farmer andthe probability of the farmerbeingcertified.On average, everyadditional year to a farmer†sageincreases the probability of the farmer being certifiedby 0.026,(but at a decreasing rate) up to age 58 (inflection point), after which it starts to decrease.Thiseffect is significant at the 1% significance level.Again, a farmer that holds a leadership positionin his/her community has a significantly (at 10% significance level) higher probability of beingcertified than a farmer who is not a leader.On average, holding a leadership position increases afarmer†s probabilityof being certified by 0.1.Householdand FarmCharacteristicsThe number of adults in a farmer†s household has a significant effect (at the 10% significancelevel) onthe farmer†s probability of beingcertified. Every additional adult householdmemberincreasesa farmer†s probability of being certified by0.035.The number of children in a farmer†shousehold as well as the percentage of his household†s income from cocoafarmingdoes notsignificantly affectthe probability of the farmerbeing certified.Also, none of the farmcharacteristics has a significant effect on a farmer†s probability of being certified.Barriers to Farmer ParticipationTable2.3below presents the reasons why some farmers are not participating in cocoacertification. The reasons are presented for non-certified farmers who are willing to join theprogram as well as those who would not want to join the program at all. It can be seen from thetable that there are four reasons mentioned by more than 10% of each groupof non-certifiedfarmers.First is the lack of adequate information about certification:38% ofnon-certifiedfarmerswhoarewilling to be certifiedand 40%of non-certified farmers whoare unwilling to becertifiedattributed their non-participation tothe fact thatthey do not know enough aboutcertification.Secondis that some farmers have no particular reason for not being certified (19%and 28% of willing and unwilling farmers respectively).Thirteen percent of non-certified butwilling farmers and19% of unwilling farmers say they are not certified or would not want to becertified because they cannot afford to be certified or cannot meet certification requirements.Table2.2:Binary Logistic Estimates for Determinants of ParticipationVariablesCoeff.RobustStd. Err.Average MarginalEffectsRobustStd. Err.Certificationstandard(Base: FLO)SAN-RA0.0300.1960.0080.057UTZ0.1030.1780.0300.052Implementationchannel(Base: LBC)Cooperative-0.276*0.152-0.081*0.044FarmerCharacteristicsMale-0.0950.196-0.0280.057Migrant-0.2650.181-0.0770.053Age0.089***0.0310.026***0.009Age2-0.001**0.0003-0.0002**0.001Education(years)0.0170.01840.0050.005Leader0.343*0.1880.100*0.054Household memberleader0.1310.2760.0380.081Acquaintance leader-0.0890.158-0.0260.046Household CharacteristicsNumber of adults0.120*0.0630.035*0.018Number of children0.0700.0550.0210.016Cocoa Income(%)0.0050.0030.0010.001FarmCharacteristicsCocoa Farm Size0.0050.0100.0020.003Sharecropped0.00040.0020.00010.001Hybrid Cocoa Variety-0.0010.006-0.00030.002Lining & Pegging0.0070.0150.0020.004N = 312                 Wald chi2(18) = 40.13***PseudoR2= 0.1006Note:*** p<0.01, ** p<0.05, * p<0.1; LBC-licensed buying companyElevenper centof non-certified but willing farmers and 12% of unwilling farmers say they arenot certified or would not want to be certified because they already sell their cocoa beans to otherlicensed buying companies.Onefocus group discussionparticipant puts it this way;—Somefarmers have loyalty with certain purchasing clerks and do not want to trade with any othercompanies.‡This is because certified farmers have to sell their beans to designated licensedbuying companies in order to receive price premium.Anotherbarrier is the issue ofloss of trust in interventions:mentioned by 4% and 35% of willingandunwilling farmers respectively.Based oninsights from the focus group discussions, thisissue relates closely with that of lack of adequate information.Farmers do not know enoughabout certification and who is promoting it in order to trust the program and this is why theyhave not joined or are unwilling to join.In one of thefocus group discussionsa farmer said,—Others were alsoskepticalof the companies behind the farmers groups as result of unpleasantprevious experiences they had‡.Some farmers are not participating or unwilling to participatebecause they do not know enough about certification in order to be convinced by its expectedbenefits.Other specific barriers mentioned include: delays in payment for beans sold as certified;lack of trust in leaders and members of farmer organization; dislike for purchasing clerk whobuys certified beans; and certified farmer organizations being politicalor ethnocentric.Table2.3:Barriers to Farmer Participation in Cocoa CertificationBarriers to ParticipationFarmersWilling to beCertified (%)FarmersUnwilling to beCertified (%)Lack of adequate information about certification3840No reason1928Cannot afford/meet certification/membership requirements1319Sell to another buyer1112Loss of trust in interventions435New in cocoa farming11-Organization no longer accepting members2-New in community74Cocoa farm notlarge enough18Don't own cocoa farm18Unwilling to practice certification requirementsN-90968DiscussionTheresults presented above are discussed in this section under similar headings as in thepresentation of the results: determinantsof participation (certification standard andimplementation channel, individual characteristics, and household and farm characteristics) andbarriers to participation.Determinants of ParticipationCertification Standard and Implementation ChannelIn general, the aim of certifying cocoa producers is to promote the adoption ofsustainableagricultural practicesin order to ensure sustainable cocoa production. However, the FLO, SAN-RA and UTZ standards have strong affiliations with particular thematicareas in terms ofrequirements and expected benefits(KPMG 2012; andMahrizal et al. 2012).As mentionedearlier under studies onsustainable agricultural practicesadoption,the attributes of an innovationplay key roles in its adoption. Because of this, the differences in the thematic areas of particularinterest to the different certification standards were expected to show up in the likelihood offarmers† participation. Theresult of the regression analysis however does not support thisexpectation. This finding may be explained by an assertion made in literatureoncocoacertification. According toGockowskiet al. 2013 andMahrizalet al. 2012,what sets thedifferent certification standards apart in practice does not really lie in the characteristics of theproduction process but rather in the certification thereof.In other words, the differentcertification standards in practice promote the adoption of the same set of recommendedpractices and do not focus on particular ones in isolation.Again, according to literature onsustainable agricultural practicesadoption, the channel throughwhich an innovation is communicated is important in determining the adoption of the innovation.This is what gives meaning to the finding that implementation channel significantly affects theprobability of certification.Inthe case of this study, a reasonable explanation to this finding liesin the role oflicensed buying companiesin the promotion of certification.Inboth approaches toorganizing farmers for certification, there is a specificlicensed buying companyinvolved interms of purchase of certified beans. Meanwhile, in the cases wherefarmers are organizedthrough alicensed buying company, the name of thelicensed buying companyand for thatmatter its business objectives are very prominent. This translates intomaking immediate materialbenefits of certification being relatively more important.It is therefore not surprising that incommunities where certification is still relatively new (like in the study communities),organizing farmers through alicensed buying companyhas a positive effect on the probability offarmers being certified.Individual CharacteristicsThe finding relating to theprobabilityof being certified increasing with age may be explained bythe assertion that older farmers usually have observed the production process well enough toknow the optimal input mix (Neill & Lee 2001). In the case of cocoa certification, especially inthe survey communities, this is important because the program is still young. Therefore it takesexperience for a farmer to have quickly gone through the process ofanalysingthe cost andbenefits of certification and deciding to participate in the program.The finding relating to the likelihood of being certified decreasing with age after age 58 may beexplained by the assertion that experience yields conservativeness. Therefore, as farmers growolder and become more experienced, they become reluctant to try new practices. One issue thatfeatured prominently in thefocus group discussions•conservativeness-also supports thisfinding and assertion. In thefocus group discussions, farmers said that those who are notparticipating in the program are farmers whodo not want to leave the farming practices that theyare used to. One farmer said, —Some farmers see the conditions of the groups regarding bestfarming practices as cumbersome and want to stick to their old or conservative method ofharvesting their cocoa.‡The findingthat being a community leader positively affects the probability of being certifiedisimportant from two perspectives. First is that, some of the expected benefits of cocoacertification go beyond the farm family, particularly community infrastructure development,improvement in public health and safety, and improved management of natural resources. Someof these issues are reasonably likely to be of concern to community leaders prior to theintroduction of cocoa certification. It is therefore reasonable to expect a community leader to bemore interested in contributing to the aims and objectives of the program. The second reasonwhy the finding is important is thatas custom demands, a common approach used in enteringcommunities with cocoa certification is first introducing the program to community leaders. Thismeans that community leaders would have firsthand information on the requirements andbenefits of the program. This firsthand information might therefore be playing a role in thedecision-making of farmers who are community leaders.A third point that is worthdiscussingwith regards to leaders being more likely to be certified hasto do with elite capture.This is because of the reasonable assumption that a community leader islikely to be of a higher social class. However,data onhousehold and housingcharacteristicspresented under —Characteristics of Survey Sample– above and in Appendix E do not support thisassumption.For instance,t-tests of means showedno significant differences between certifiedand non-certified households with regardshousehold income, childreneducation, and rooms peradult member.Furthermore, household size for instance has a significant difference but this israther against the assumption of elite capture: certified households have more members and thisis known to be uncharacteristic of elites.Though the results also show that certified householdscultivate significantly more farms and have larger farm holdings, this is not easily attributable tothey being elites.Household and Farm CharacteristicsFrom Hayami & Ruttan in Eicher & Staatz 1998, we know that agricultural technologies aredesignedto use more or less of an input(s). According to Feder, Just & Zilberman 1985 as wellas Neill & Lee 2001, a technology that uses morelaborand less other inputs would be easilyadopted by householdsthathave moremembers who can work on the farm.  The finding that thenumber of adult household members has a significant difference in a farmer†s certification statusmay be explained by an earlier description ofsustainable agricultural practicesas being generallylabor-intensive(Teyet al.2014). Also, in the next part below, where the barriers to participationare discussed, time for attending meetings of certified farmer organizations is mentioned as oneof the major reasons why some farmers are not certified or willing to be certified. This gives animportant insight to why farmers in household with more adult members are more likely to becertified. Reasonably, more adult household members(and also significantly more householdmembersas indicated earlier)increase the time endowment of the household and thereforereduce the opportunity cost of each member†s time. This makes it easier for members to devotetime to relatively less important uses such as attending farmer organization meetings.Barriers to ParticipationIn Chapter 1, it was mentioned that farmers† knowledge of certification is low, and this assertionhas beensupported hereas lack of adequate information has been found to be an importantbarrier to participation.Recallthat one of thereasonsattributed to community leaders havinghigher probabilities of being certified is the access to information about certification. Based onthese related findings, it is important to mention thatindeed information matters in farmerparticipation in cocoa certification.Onthe other hand,and in a related result,lost of trust inintervention programs is an important reason why some farmers are not willing to joincertification.This couldalso be seen as reason why they are scepticaland would want to knowmuch more about the program before making commitments.With regard toaffordability of certification/membership requirements,specific survey andfocusgroup discussionresponsesindicate thatitmainly has to do with the cost of membershipregistration and also time for attending meetings. Certified farmer groups charge an average ofGH µ2as one-time membership registration fees,and most of them meet twice every month.Non-certified but willing farmers are mostly constrained by their inability to afford theregistration fees. While unwilling farmers say they findprogramparticipationtime consuming.Considering the cost of one-time member registration fees, itwould bereasonable to assume thatfarmers who are not able to participate in certification because of this are most likely from verypoor households.Thisassumptionis however not supported by the data on grounds similar tothose discussed above in relation to elite capture.The time constraintother handsupports theearlier discussion on the implications of labor endowments for participation in cocoacertification.In bothlicensed buying companyand farmer cooperative certification implementation, specificlicensed buying companiesare engaged in the purchase of certified cocoa beans. This means thatin acommunitycertified farmers must sell their beans to a particularlicensed buying companyinorder to receive price premium. However, thereismore than onelicensed buying companyinmost of these communities. Prior to the introduction of certification, farmers had particularlicensed buying company(s) they sold their beans to and this choice is typically informed by veryimportant reasons. Thisis whatmakes it difficult for some farmers to join certified farmerorganizations if the organization has arrangements with alicensed buying companyother thantheir existingone. µ3.96CONCLUSIONS AND IMPLICATIONS FOR POLICY AND RESEARCHThe aim of thisstudywas to contribute to the understanding of contextual conditions forimplementing certification programs that promote sustainable agricultural practices.Specifically,the study sought answers to questions regarding theways and extents to whichfarmers areengaged inthe activities and processes of cocoa certification;stakeholders† assessments of cocoacertification;andfactors that determine participation in cocoa certification.Basedon theresultsdiscussed in the two chapters above, the study draws the following conclusionsand suggeststheir implications for policy and future research.One conclusionis the fact that farmer knowledge of cocoa certification is inadequate and thisbesides undermining the expected democratic nature of farmers† participationisalsoa majorreason why some farmers are not certified.It also shows up in the finding that communityleaders are more likely to be certified.This calls for intensification in the advocacy forcertification: all the objectives of the program and the associated costs and benefits tostakeholders need to be made more explicit in the campaignsto get cocoa farmers certified.Another one is thatthe results raise little or no concernsabout how farmersget involvedincertified farmer groupsand identify with group and certification requirements and objectives.This means that if the flow of information about certification is better managed, there could besome guarantee thatonce farmers join certification, theywould adopt practices promoted by theprogram.Also,farmers are not engaged in inspections for compliance and decisions on price premiumdistributionas expected. These two are however very critical pillars of the programand need tobe considered carefully. Research on innovative ways to increase the efficiency of inspections aswell as streamlining therole of licensed buying companieswould be helpful in these regards.Itisalso a conclusionthat all stakeholdersfavourably identify with the requirements, objectivesand prospects of cocoa certification and this speaks positively to the sustainability of theprogram.Conditions that compel farmers to sell certified beans as conventional are howeverworth immediateattention as this continues to be a major challenge of the program.This issuewas actually found to beanothermajor reason why some farmers are not participating in theprogram.Again, learning about how to streamline the role of licensed buying companieswouldbe helpful in this regard.Thestudyalsoconcludes thatthe channel or approach used to organize farmers for certificationhasimportant implications on farmers† participation.Itis important forresearch to explain howorganizing farmers through licensed buying companies increasethe probability of theirparticipation. Such knowledge wouldenable better utilization ofthe opportunitiesembedded inthat approach.Anotherconclusion of thestudyis thatexperience isacatalyst for as well as ahindrance to participation. Further research is however needed to show how experience plays outin these regards in order to inform policy on how to bring in the less and overly experienced.The final conclusion of the study is that household labor endowment is important in determiningfarmer participation in cocoa certification.This may have significant implications from theperspective of structural transformation and theassociated movement of labor within andbetween sectors of the economy.This calls forresearch to shed light onlabor dynamicswithinthe cocoa sector of Ghana.APPENDICESAPPENDIX A:SAMPLING FRAMEWORKTableA.1:Sampling FrameworkCertification Standard/Implementation ChannelDistrictsCommunitiesin SelectedDistrictHouseholds inSelectedCommunity(Census Report)Households inSelectedCommunity(Enumerated)UTZ/LBC2221247204UTZ/Farmer Cooperative28Unknown201FLO/LBC129371341FLO/Farmer Cooperative734282273SAN-RA/LBC422252547SAN-RA/Farmer Cooperative334265210Totals391481776Note: LBC-licensed buying companyAPPENDIX B:LOCATION OF STUDY COMMUNITIESFigureB.1: Location of Study CommunitiesAPPENDIX C:DETAILS OF SAMPLE WEIGHT CALCULATIONTableC.1: Details of Sample Weight CalculationHousehold TypeEnumeratedPopulationCorrectedPopulationPopulationProportionSurveySampleSampleProportionWeightCertifiedCommunity 0103Community 0107Community 0203Community 0207Community 0303Community 030744226237588043.694121.8848661.6827836.8810357.538780.318530.30343120.27018350.2681860.23948720.32692440.17089052621272626240.50.420.51923080.4905660.48148150.47058820.60690.64320.51650.48820.67900.3631Non-CertifiedCommunity 0103Community 0107Community 0203Community 0207Community 0303Community 030710059168117118390100.305959.11514168.3172117.119118.4613389.68150.69656880.72981660.7318140.76051280.67307560.82910952629252728270.50.580.48076920.5094340.51851850.52941181.39311.25831.52221.49291.29811.5661APPENDIX D:DESCRIPTION AND HYPOTHESES OF INDEPENDENT VARIABLESTable D.1: Description and Hypotheses ofRegressionVariablesVariableDescriptionExpectedsignCertification StatusCertified1 if farmer is a member ofa certified farmer organization; 0 otherwiseCertificationstandardFLOSAN-RAUTZCertification standard implemented in community1 ifFLOis implemented in community; 0 otherwise1 ifSAN-RAis implemented in community; 0 otherwise1 if UTZ isimplemented in community; 0 otherwise???Implementation ChannelLBCCooperativeChannel used to organize farmers for certification in community1 ifLBCis used; 0 if farmer cooperative is used1 if farmer cooperative is used; 0 if LBC is used??Farmer CharacteristicsMaleMigrantLeaderHousehold member leaderAcquaintance leaderAgeAge2Education1 if farmer is male; 0 if farmer is female1 if farmer was born outside of community; 0 otherwise1 if farmer holds a leadership position incommunity; 0 otherwise1 if farmer†s household member holds a leadership position in community; 0 otherwise1 if farmer†s close acquaintance holds a leadership position in community; 0 otherwiseAge of farmer in completed yearsFarmer†s age squaredNumberof years of schooling completed by farmer-Or+?+1+1+1-Or+-Or++1Household CharacteristicsNumber of AdultsNumber of ChildrenCocoa incomeNumber of household members 18 years and olderNumber of household members 5 to 17 years oldPercentage of household income from cocoa in the 12 monthsleading up to the survey-Or+-1-Or+Farm CharacteristicsCocoa farm sizeHybrid cocoa varietyLining & PeggingSharecroppedTotal land area (in acres) that household has under cocoacultivationPercentage of the number of household cocoa farms established at least in part withhybrid cocoa variety prior to the introduction of cocoa certificationPercentage of the number of household cocoa farms established at least in part usinglining and pegging prior to the introduction of cocoa certificationPercentage of total cocoa farm land sharecropped by household-Or++1+1-1Note: LBC-licensed buying companyAPPENDIX E:HOUSING CHARACTERISTICSTable E.1: Housing CharacteristicsVariableFrequency (Percentage)Entire SampleCertifiedFarmersNon-certifiedFarmersSource of drinking water (rainy season)Bore-hole/tube wellRain waterRiver/streamPipe-borne outside houseSachet waterPipe-borne inside houseWell168 (54)103 (33)23 (7)12 (4)5 (2)1(0.3)088 (59)44 (29)12 (8)4 (3)2 (1)0080 (49)59 (36)11 (7)8 (5)3 (2)1(0.6)0Source of drinking water (dry season)Bore-hole/tube wellRiver/streamPipe-borne outsidehouseSachet waterWellRain waterPipe-borne inside house234 (75)45 (14)14 (4)13 (4)5 (2)1 (0.3)0115 (77)20 (13)5 (3)7 (5)3 (2)00119 (73)25 (15)9 (6)6 (4)2 (1)1 (0.6)0Source of energy for lightingNational gridDrycell/regular batterySolar-powered batteries288(92)21(7)3(1)141(94)6(4)3(2)147(91)15(9)0Source of energy for cookingFirewoodCharcoalGasDungElectricity283 (91)21 (7)6 (2)1 (0.3)1 (0.3)139 (93)10 (7)01 (0.7)0144(89)11 (7)6 (4)01 (0.6)Dwelling roof materialCorrugated iron sheetsPalm leave/raffia thatchCement/concreteAsbestos/slate305 (98)3 (1)2 (0.6)2 (0.6)147 (98)1 (0.7)1 (0.7)1 (0.7)158 (98)2 (1)1 (0.6)1 (0.6)Dwelling outer wallmaterialCement/concreteEarth/mud/mud bricksWoodBurnt bricks269 (86)41 (13)1 (0.3)1 (0.3)132 (88)17 (12)01 (0.7)137 (85)24 (15)1 (0.6)0Mean (Standard deviation)Rooms per adult HH member1.18(0.68)1.16(0.63)1.2(0.72)N312150162APPENDIXF:DESCRIPTIVE STATISTICS OFREGRESSIONVARIABLESTable F.1:Descriptive Statistics of Regression VariablesVariableFrequency (Percentage)Entire SampleCertifiedFarmersNon-certifiedFarmersCertificationstandardFLOSAN-RAUTZ105(34)105(34)102(33)53(36)50(32)47(32)52(32)55(34)55(34)ImplementationchannelLBCCooperative158(51)154(49)79(53)71(47)75(49)79(51)Farmer CharacteristicsMaleMigrantLeaderHousehold member leaderAcquaintance leader243(78)81(26)85(27)26(8)138(44)117(78)33(22)55(37)16(11)66(44)126(78)48(30)30(19)10(6)72(44)Mean (Standard deviation)AgeAge246.66(13.85)2368.30(1388.62)49.56(12.60)2613.8(1281.64)43.96(14.45)2139.57(1448.17)7.90(4.39)Education (years)8.15(4.25)8.43(4.08)Household CharacteristicsNumber of Adults2.50(1.37)2.78(1.45)2.25(1.25)Number of Children1.51(1.47)1.6(1.55)1.42(1.38)Cocoa Income (%)65.38(26.17)67.98(25.12)62.97(26.95)Farm CharacteristicsCocoa Farm Size (acres)9.34(8.72)10.52(9.05)8.24(8.18)Hybrid Cocoa Variety (%)98(12.39)98.67(9.13)97.38(14.64)Lining Pegging (%)99.50(6.12)99.78(2.65)99.23(8.09)Sharecropped (%)30.57(42.71)28.28(40.23)32.69(45.07)N312150162APPENDIXG:QUALITATIVE INTERVIEW AND DISCUSSION GUIDENature and Extent of Farmer Participation in CocoaCertification in GhanaQualitative Interview/Discussion GuideIn-depth InterviewsCOCOBOD Officials1.How does COCOBOD manage cocoa certification?2.What is the current state of cocoa certification in Ghana?3.Which division or affiliate of COCOBOD would haveinformation on the currentcoverage of cocoa certification in Ghana?4.What is the best means and approach to access such information?5.Which division of COCOBOD would have statistics on communities?6.What is the best means and approach to access such information?Standard Bodies/Certification Officials andLicensed Buying Companies1.Which farmer organizations do you work with?2.Where are they located?3.What are the sizes of their membership?4.For how long have you being working with these organizations and when did youcertify them?5.How does an organization get to be certified by your standard?6.What is your mode of operation?7.How do you choose the organizations that you work with?8.Which external auditors andlicensed buying companiesdo you work with?9.How much do you pay as price premium?10.How do you decide how much to pay as price premium?11.How is the price premium distributed?12.Have you had any difficulties with price premiums? If yes, what are some of suchdifficulties?13.How does your standard go aboutverification and compliance?14.Have you had difficulties with verification and compliance? If yes, what are someof such difficulties?15.How does your standard deal with the marking of certified cocoa beans?16.Have you had difficulties with marketing of certified cocoa beans? If yes, what aresome of such difficulties?Focus Group DiscussionsOrganization and Membership1.How many farmer organizations are in this community?2.Are any of the farmer organization(s) certified?3.What are the names of these farmerorganizations and what are they commonlyreferred to?4.When was/were the organization(s) formed?5.How was/were the organization(s) formed?6.What was/were the main motive(s) for forming the organization(s)?7.What other reason(s) form the basis for the existenceof the organization(s)?8.Who qualifies to be a member of the organization(s)?9.Why do farmers join this organization?10.Are you organized at the national, regional, district or local level(s)?11.How is the leadership of the organization(s) structured at each level(s)?12.How many farmers are currently registered with the organization(s) at each level?13.What kinds of information do you have on your members that you can share withme? (E.g. farm sizes, average yields, gender, age, location etc.)14.How often are meetingsheld by the organization(s) at each level?15.How are decisions typically made at each level of the organization(s)?Affiliations16.Which certification standard bodies,licensed buying companies, and externalauditors do the organization(s) work with? Anyspecial reasons for working withwhom they work with?17.For how many years have the organization(s) been working with the variousentities and for how many years has this organization been certified?Management18.How do farmers and for that matter the organization really get certified?19.How does your organization manage price premium (amount, distribution, usesetc.), who makes such decisions and how?20.How does your organization manage verification and compliance (ICS, externalauditing), who makes such decisions and how?21.How do members of this organization sell their certified cocoa beans?SWOT Analysis22.What do you think are the strengths of cocoa certified?23.What do you think are the weaknesses of cocoa certification?24.What opportunities do you see in cocoa certification?25.What are your assessments of cocoa certification so far, in terms of the expectedbenefits?APPENDIXH:HOUSEHOLD QUESTIONNAIREUNIQUE IDNATURE AND EXTENT OF FARMERPARTICIPATION IN COCOACERTIFICATION IN GHANAHOUSEHOLD QUESTIONNAIREJuly-August 2015DEPARTMENT OF COMMUNITY SUSTAINABILITYMICHIGAN STATE UNIVERSITYCommunity NumberHousehold NumberInterviewer NumberCompleted Questionnaire Checked andApproved:Check if yes:Date Approved:DD / MM / YYYYUNIQUE IDCommunity Name(Circle one)1.Adarkwa2.Anwiam3.Asempaneye4.Kwaboanta5.Mafia6.BesibemaCommunityNumberHousehold Number(Copy from listing and selection sheets)Interviewer Name(Write Name)1.Addo Kingsley2.Essandoh Francis3.Owusu Mensah Vida4.Quansah Kenneth5.Ebenezer Offei AnsahInterviewer NumberWas an interpreter used for this interview?(Circle one)1.Yes2.NoInterview DateDD/MM/YYYYInterview Start TimeHH:MMINFORMATION SHEETUNIQUE IDParticipant Name:_________________________________________Interviewer Name: _________________________________________Date: ________________________Time______________________My name is [Researcher†s name], a graduate student at Michigan State University. Areyou at least 18 year old? [If not, thank and terminate interview].You are being asked to voluntarily participate in a research study of cocoa certification inGhana. The findings of this study will help form the basis for policy recommendationsregarding the design of cocoa certification programs.If you agree to participate, I will ask you some questions about cocoa certificationprograms and about your participation in such programs. I will also ask some questionson issues regarding your household and farm, farmer organization, and community. Theinterview will takeapproximately thirty to forty minutes of your time.You should know that your identity and responses to questions will be kept confidentialand your privacy will be protected to the maximum extent allowable by law.All reportsand publications resultingfrom this interview will be written and shared usingpseudonyms and code numbers. Only the researchers will have access to your responsesand the data will be stored on a secure, password-protected computer and in offices at theMichigan State University with no identifying information linking them to you.Your participation is voluntary and you may choose not to participate at all, refuse toanswer certain questions, or stop the interview at any time without any consequences.Itis also important for youto know that there are no right or wrong answers.If you have any questions or concerns regarding your participation in this study, you maycontact the researchers whose contact information is on the sheet I am handing you [handInformation Sheet torespondent]. If you feel your rights have been violated or you aredissatisfied with any aspect of the study, please contact Michigan State University†sHuman Research Protection Program using the contact information on the InformationSheet.Do you haveany questions?YesNo[if yes, answer questions and proceed]You indicate your voluntary agreement to participate in this study by beginning theinterview with me. May I begin?YesNo[If no, thank and end]CONSENT FORMUNIQUE ID1.How knowledgeable are you of cocoa certification?(Circle one)(1)Very knowledgeable(2)Somewhat knowledgeable(3)Not knowledgeable at all(98)Don†t know(99)Refused to answer2.What are the main issues of interest to cocoa certification?DO NOT PROMPT(Circle all that apply)(a)Conserving/protecting natural resources(b)Improving farmers output and income(c)Eliminating child labor(d)Improving working conditions of farm workers(e)Community development(f)Other (specify)_________________________________________(98)Don†t know(99)Refused to answer3.Have you ever been a member of any certified cocoa farmer organization(s)?(1)Yes(2)No(SKIP TO Q17)(98)Don†t know(99)Refused to answer4.Are you currently a member of any certified cocoa farmer organization(s)?(1)Yes(SKIP TO Q6)(2)No(98)Don†t know(99)Refused to answer5.Why did your leave the certified cocoa farmer organization?(Circle all thatapply)(a)Unwilling to practicecertification requirements(b)Inability to meet certification requirements(c)Another reason (specify)_________________________________(98)Don†t know(99)Refused to answer6.What is the name of the certified cocoa farmer organization you are/were amember of?(Write name)†††††††††††††††††††††††(98)Don†t know(99)Refused to answerSECTION 1: ORGANIZATION ANDCERTIFICATIONUNIQUE ID7.What certification standard is this organization certified/working with?(1)Fairtrade(2)UTZ(3)Rainforest Alliance(98)Don†t know(99)Refused to answer8.For how many years have you been/were you a member of this organization?(Write number, 98 = Don…t know, 99 = Refused)9.How are leaders typically chosen for this organization?(Circle only one)(1)Appointed by external agent(2)Appointed by village leaders(3)Vote of all members(4)Self-nominated(5)Other (specify)(98)Don†t know(99)Refused to answer10.Do/did you hold any leadership position in this organization?(Circle one)(1)Yes(2)No(98)Don†t know(99)Refused to answer11.Roughly how many times does/did the organization hold meetings in a typicalyear?(Write number, 98 = Don…t know, 99 = Refused to answer)12.In the last 12 months of you membership, how many of these meetings have/didyou attend(ed)?(Write number, 98 = Don…t know, 99 = Refused to answer)13.When was the last time your farm was inspected by internal inspectors?(Circleonly one)(1)More than three years ago(2)Within the last three years but more than a year ago(3)Within the last year(4)Never(98)Don†t know(99)Refused to answer14.When was the last time your farm was inspected by external auditors?(Circle only one)(1)More than three years ago(2)Within the last three years but more than a year ago(3)Within the last year(4)Never(98)Don†t know(99)Refused to answerUNIQUE IDISSUE15.To what extentdo/did you agree ordisagree with decisionsof the organizationregarding[ISSUE]?(Use Codes Below)16.To what extent doyou agree or disagree thatmembers of the organizationare/were engaged indecision-making regarding[ISSUE]?(Use CodesBelow)[SKIP TO Q 20]Amount paid as price premiumDistribution of price premium amongvarious usesTime ofpaying price premiumSelling of certified cocoa beansMeetings (frequency, days, times,venue, duration etc.)Membership fees and other paymentsInternal inspectionExternal auditingRequirements for cocoa certificationCodes: Strongly Agree = 1      Somewhat Agree = 2      Neutral = 3     Somewhat Disagree = 4Strongly Disagree = 5       97 = N/A              98 = Don•t know       99 = Refused to answer17.Would you like to join a certified cocoa farmer organization?(Circle only one)(1)Yes(2)No(SKIP TO Q 19)(98)Don†t know(99)Refused to answer18.Why have you not joined a certified cocoa farmer organization?(Circle all that apply)(a)Cannotafford membership registration fee(b)Organization no longer accepting members(c)Another reason (specify)___________________________(98)Don†t know(99)Refused to answer[SKIP TO Q 20]19.Why would you never join a certified cocoa farmer organization?(Circle all that apply)(a)Loss of trust in interventions(b)Unwilling to practice certification requirements(c)Inability to meet certification requirements(d)Another reason (specify)___________________________(98)Don†t know(99)Refused to answerUNIQUE IDREQUIREMENTS20.To what extent do you agree or disagree with[REQUIREMENT]as arequirement for cocoa certification?(Use codes below)Cocoa farm establishment and rehabilitationFarm managementand maintenanceSoil management and fertilizationIntegrated pest management and crop protectionHarvest and post-harvest practicesSafe and healthy farm practicesWorkers† rights, including child labor andinformal workersWaste managementEnvironment and natural resource protectionOrganization for implementationBENEFITS21.To what extent to you agree ordisagree that you expected cocoacertification to bring about[BENEFIT]?(Use codes below)22.To what extent do youagree or disagree that cocoacertification has brought about[BENEFIT]?(Use codes below)Improvement in farm managementImprovement in awareness of environmentalprotection and farm environmental conditionsImprovement incocoa outputAccess to price premium and increase incomeImprovement in awareness of labor rights andconditions of workers and childrenAccess to credit/financial assistanceImprovement in knowledge of safety andhealthy farm practicesAccess to farm inputsCommunity infrastructure developmentAccess to market/buyer requested certificationAccess to extension servicesUNIQUE IDCodes:            Strongly Agree = 1                  Somewhat Agree = 2              Neutral = 3Somewhat Disagree = 4Strongly Disagree = 5                97 = N/A                             98 = Don•t know                  99 = Refused to answerUNIQUE IDPART A: HOUSEHOLDFARMING23.How many separate farm(s) does your household cultivate?(Write number, 98 = Don…t know, 99 = Refused to answer)QuestionFarm1Farm2Farm3Farm4Farm5Farm Number24.Do you own, rent or sharecrop this farm?(Write Code)1.Own 2.Rent 3.Sharecrop 98 Don…t know 99 Refused to answer25.What is the size of this farm?NumberUnit Codes1.Acres 2.Poles 3.Ropes 4.Hectares 98 Don…t know99 Refused to answer                                                                Unit26.What is the size of the part of this farm planted withcocoa?Number(SKIP TO NEXT FARM IF 0)Unit (Use Codes in 25 Above)27.How old is a typical cocoa tree on this farm?Write Number of Years, 98= Don…t know, 99 = Refused to answer28.Are any of the cocoa trees on this farm of the hybridvariety?1.Yes         2.No            98 Don…t know          99 Refused to answer29.Are the cocoa trees on this farm planted in rows, usinglining and pegging?1.Yes        2.No98 Don…t know          99 Refused to answer30.How far is this farm from a water body?NumberUnit Codes1.Meters 2.Miles 3.Kilometers 97 N/A  98 Don…t know99 Refused to answer                                                                Unit31.How far is this farm from a forest reserve/protected area?NumberUnit (Use Codes in 30 Above)SECTION 2: HOUSEHOLD WORKUNIQUE ID32.In the last 12 months, how much cocoa did you harvest from all of your farms?(Write number of bags, 98 = Don…t know, 99 = Refused to answer)33.What amount of this total cocoa harvest was produced as certified cocoa?(Writenumber of bags, 98 = Don…t know, 99 = Refused to answer)(SKIP TO Q37 IF 0)34.What amount of this certified cocoa did you sell as certified?(Write number of bags, 98 = Don…tknow, 99 = Refused to answer)(SKIP TO Q36 IF EQUAL TO Q 33)35.Why did you sell some of your certified cocoa beans as conventional?(Circle all that apply)(a)Certified buyer rejected beans(b)Certified buyer was not available(c)Certifiedbuyer did not have money(d)Another reason (specify)__________________________________(98)Don†t know(99)Refused to answer36.In the last 12 months, how much did you receive as price premium on yourcertified cocoa?(Write Ghana Cedis for either perbag or in total, 98 = Don…tknow, 99 = Refused to answer)Per bag:ORTotal:37.In the last 12 months, did you lose some of your cocoa due to any of the followingproblems?(Circle all that apply)(a)Pests and diseases(b)Drought and flood(c)Bush fire(d)None(SKIP TO Q39 IF †NONE•)(98)Don†t know(99)Refused to answer38.How much cocoa would you harvest if not for these loses?(Write number of bags, 98 = Don…t know, 99 = Refused to answer)39.Over the last 5 years, has the output of your cocoa farm(s) been increasing, aboutthe same or decreasing?(Circle one)(1)Increasing(2)About the same(3)Decreasing(98)Don†t know(99)Refused to answerUNIQUE ID40.In the last 12 months, roughlywhat proportion of the following farm activitieswas done by each of the following source of labor on your farm(s)?(Writefraction e.g. 3/4, 97 = N/A, 98 = Don…t know, 99 = Refused to answer)PART B: HOUSEHOLD INCOME41.In the last 12 months, roughly how many Ghana Cedis did your household receivefor(SOURCE)?(Write Ghana Cedis, 98 = Don…t know, 99 = Refused to answer)SourceGhana Cedis(GH µCocoaOther cropsPaid employmentNon-farmenterprise/businessLivestock and animal productsHunting and gatheringPalm wineRent from houses you ownRent from equipment/animals you ownTrading of non-agricultural goods (e.g. crafts, clothes etc.)TourismFishingRemittancesAnother source (specify)‡‡‡‡‡‡‡..ActivityFamilylaborExchangelaborHiredlaborCaretakerClearing new fields for plantingPlanting cropsWeedingPruningFertilizing and mulchingSoil and water managementPests and diseases controlNursery operationsHarvestingPost harvest activitiesUNIQUE IDPART A: DEMOGRAPHICS42.Name(Write onlycommonname)43.ID44.Gender1 = Male2 = Female98 = Don…t know99 = Refused toanswer45.Age (years)98 =Don…t know99 = Refused toanswer46.What is(NAME…S)relationshipto thehouseholdhead?47.How manyyears ofschoolinghas(NAME)completed?(SKIP TO49 IF 0)98 = Don…tknow99 =Refused48.Is(NAME)currentlyenrolledinschool?Yes•1(SKIP49)No•249.What isthe mainreasonwhy(NAME)isnotenrolled inschool?(Usecodesbelow)50.Howmanydays ofschooldid(NAME)miss inthe lasttwoweeks51.What isthe mainreasonwhy(NAME)missedschool?(Usecodesbelow)SECTION 3: HOUSEHOLD AND INDIVIDUALCHARACTERISTICSUNIQUE IDQ 46: Relationship Codes1‡‡‡‡‡‡‡‡‡‡..Household head2‡‡‡‡‡‡‡‡‡‡‡‡‡‡Spouse3‡‡‡‡‡‡‡‡‡‡‡‡‡‡...Child4‡‡‡‡‡‡‡‡‡‡‡‡‡Step child5‡‡‡‡‡‡‡‡‡‡‡Niece/Nephew6‡‡‡‡‡‡‡‡‡‡‡‡..Grandchild7‡‡‡‡‡‡‡‡‡‡‡‡‡....Sibling8‡‡‡‡‡‡‡‡‡‡‡‡‡‡.Parent9‡‡‡‡‡‡‡‡‡‡‡‡‡‡.In-law10‡‡‡‡‡‡‡‡‡‡...Other relative11‡‡‡‡‡‡‡‡‡‡‡‡House help12‡‡‡‡‡‡‡‡‡‡‡.Non-relativeQ 49 & 51: School enrolment and absentcodes1‡‡‡‡‡‡‡‡‡‡.Disability/Illness2‡‡‡‡‡‡...No School/School too far3‡‡‡‡‡‡‡‡..Cannot afford school4‡‡‡‡Family does not allow schooling5‡‡‡‡‡‡‡..Not interested in school6‡‡‡.Education not considered valuable7‡‡‡‡‡‡‡‡‡‡...School not safe8‡‡‡‡‡‡‡‡‡‡‡.Learning a job9‡‡‡‡‡‡‡Work on household farm10‡‡‡‡‡‡‡..Other household work11‡‡‡..Work for pay outside household12‡‡‡...Unpaid work outside household13‡‡‡‡‡‡‡‡‡.Household chores14‡‡‡‡‡‡‡‡‡.Competed School15‡‡‡‡‡‡‡‡‡‡‡‡...Vacation16‡‡‡‡‡‡‡‡‡‡..Teacher absent17‡‡‡‡‡‡‡‡‡‡‡..Badweather19‡‡‡‡‡‡‡..Other reason (specify)97‡‡‡‡‡‡‡‡‡‡..Not applicable98‡‡‡‡‡‡‡‡‡‡‡.Don†t Know99‡‡‡‡‡‡‡‡....Refused to answerPART B: INFRASTRUCTURERainy SeasonDry Season52.What is your household†s main source ofdrinking water?(Use codes below)53.How many minutes does it take to go getwater from this source and come back?98 = Don…t know, 99 = Refused to answer54.In the last12 months, how many days, ifany, was water not available?(Write Number ofDays, 98 = Don…t know, 99 = Refused to answer)55.Roughly how much do you pay for waterfrom this source monthly?(Write Ghana Cedis, 98= Don…t know, 99 = Refused to answer)Water Source Codes1. Pipe-borne inside house    2. Pipe-borne outside house3. Tanker service     4. River/stream      5. Bore-hole/tube well        6. Well7. Dug out/pond     8. Rain water 9. Another source (Specify)________________98. Don…t know                 99. Refused to answerUNIQUE ID56.What is your household†s primary source of energy for lighting?(Circle only one)(1)National grid(2)Generator(3)Car battery(4)Dry cell/regular batteries(5)Wind-powered batteries(6)Solar-powered batteries(7)Kerosene(8)Another source(Specify)______________________________(98)Don†t know(99)Refused to answer57.What is your household†s primary source of energy for cooking?(Circle only one)(1)Charcoal(2)Firewood(3)Dung(4)Gas(5)Electricity(6)Kerosene(7)Solar-powered stove(8)Another source(Specify)_______________________________(98)Don†t know(99)Refused to answer58.What material is the roof of your dwelling made of?(Circle only one)(1)Palm leaves/raffiathatch(2)Wood(3)Corrugated iron sheets(4)Cement/concrete(5)Asbestos/slate(6)Roofing tiles(7)Mud bricks/earth(8)Bamboo(9)Another material(Specify)______________________________(98)Don†t know(99)Refused to answer59.What material is the outer wall of your dwelling made of?(Circle only one)(1)Earth/mud/mud bricks(2)Wood(3)Stone(4)Cement/concrete(5)Burnt bricks(6)Ceramic/tiles(7)Another material(Specify)________________________________(98)Don†t know(99)Refused to answer60.How many separate rooms (including bathrooms and kitchens) are in your dwelling?(Write Number, 98 = Don…t know, 99 = Refused to answer)PART C: INDIVIDUAL INFORMATION61.ID OF PERSON INTERVIEWED62.Wherewere you born?(1)This village/town(SKIP TO Q 64)(2)Another village/town in this district(3)Another district in this region(4)Another region(5)Outside Ghana(98)Don†t know(99)Refused to answer63.For how many years have you lived in this community?(Write number, 98 = Don…t know, 99 = Refused to answer)64.For how many years have you been cultivating cocoa?(Write number, 98 = Don…t know, 99 = Refused answer)65.Do you hold any leadership position in this community?(Circle)(1)Yes(2)No(98)Don†t know(99)Refused to answer66.Does any member of your household hold any leadership position in this community?(Circle)(1)Yes(2)No(98)Don†t know(99)Refused to answer67.Does any close acquaintance of yours hold any leadership position in this community?(Circle)(1)Yes(2)No(98)Don†t know(99)Refused answer68.How much control do you believe individuals have over whether they succeed or failin life?(Read first 3 options aloud and circle only one)(1)A large amount of control(2)Some control(3)Very little control(98)Don†t know(99)Refused to answer69.Who should take responsibility for the development and success of this community?(Read first 4options aloud and circle only one)(1)Government(2)Villagers(3)Both equally(4)Another external agents (e.g. NGOs)(98)Don†t know(99)Refused to answer70.Would you recommend that your children go into cocoa farming?(Circle one)(1)Yes(2)No(SKIP TO Q 72)(98)Don†t know(99)Refused to answer71.Why would you recommend that your children go into cocoa farming?(1)Stable source of income(2)Important family property(3)No better option(4)Important for national economy(5)Another reason(specify)______________________________(98)Don†t know(99)Refused to answer(END INTERVIEW)72.Why would you not recommend that your children go into cocoa farming?(1)Low income(2)Low status(3)Work too hard(4)Better opportunities in other fields(5)Not enough land(6)Another reason(Specify)____________________(98)Don†t know(99)Refused to answer(END INTERVIEW)Interview End TimeHH:MMTHANK YOU VERY MUCH FOR THE INFORMATION YOU HAVE PROVIDEDTO US. WE APPRECIATE YOU TAKING THE TIME TOPARTICIPATE IN THESTUDY.BIBLIOGRAPHYBIBLIOGRAPHYAidoo, Robert, and Ingrid Fromm. —Willingness to Adopt Certifications and SustainableProduction Methods among Small-Scale Cocoa Farmers in the Ashanti Region ofGhana.–Journal ofSustainable Development8, no. 1 (February 2015): 33•43.Antle, John M., and Bocar Diagana. —Creating Incentives for the Adoption of SustainableAgricultural Practices in Developing Countries: The Role of Soil Carbon Sequestration.–American Journal ofAgricultural Economics85, no.5 (December 1, 2003): 1178•84.Borges, J. A. R., Oude Lansink, A. G. J. M., Marques Ribeiro, C., & Lutke, V. (2014).Understanding farmers† intention to adopt improved natural grassland using the theory ofplanned behavior.Livestock Science,169, 163•174.Caviglia•Harris, JillL. —Sustainable Agricultural Practices in Rondƒnia, Brazil: Do LocalFarmer Organizations Affect Adoption Rates?–Economic Development and CulturalChange52,no. 1 (October 1, 2003): 23•49.Dabbert,S., Lippert, C., & Zorn, A. (2014). Introduction to the special section on organiccertification systems: Policy issues and research topics.Food Policy,49, Part 2, 425•428.Darkwah, S., & Verter, N. (2014). An empirical anaylsis of Cocoa bean production in Ghana.European Scientific Journal,10(16).Eicher, Carl K., and John M. Staatz.International Agricultural Development. JHU Press,1998.Ewusi, K (2013). Policies and Options for Ghana†s Economic Development. 3rdedition.Legon: Institute of Statistical Social and Economic Research, University of Ghana.Feder, Gershon, Richard E. Just, and David Zilberman. —Adoption of AgriculturalInnovations in Developing Countries: A Survey.–Economic Development and CulturalChange33, no. 2 (1985): 255•98.Gedikoglu, H., & McCann, L. M. J. (2012). Adoption of win-win, environment-oriented, andprofit-oriented practices among livestock farmers.Journal of Soil and WaterConservation,67(3), 218•227.Gershon Feder, Dina L. Umali. —The Adoption of Agricultural Innovations: A Review.–Technological Forecasting and Social Change43, no. 3•4 (1993): 215•39.Ghana Statistical Service, 2013 —provisional_gdp_2013.pdf.–Gockowski, J., Afari-Sefa, V., Sarpong, D. B., Osei-Asare, Y. B., & Agyeman, N. F. (2013).Improving the productivity and income of Ghanaian cocoa farmers while maintainingenvironmental services: what role for certification?International Journal of AgriculturalSustainability,11(4), 331•346.González, A. A., & Nigh, R. (2005). Smallholder participation and certification of organicfarm products in Mexico.Journal of Rural Studies,21(4), 449•460.Herzfeld, T., &Jongeneel, R. (2012). Why do farmers behave as they do? Understandingcompliance with rural, agricultural, and food attribute standards.Land Use Policy,29(1),250•260.Johnston, Bruce F., and John W. Mellor.—The Role of Agriculture in EconomicDevelopment.–The American Economic Review51, no. 4 (1961): 566•93.Jorgensen, B. S., & Martin, J. F. (2015). Understanding farmer intentions to connect to amodernised delivery system in an Australian irrigation district: a reasoned actionapproach.Journal of Environmental Planning and Management,58(3), 513•536.Kassie, Menale, Moti Jaleta, Bekele Shiferaw, Frank Mmbando, and Mulugetta Mekuria.—Adoption of Interrelated Sustainable Agricultural Practices in Smallholder Systems:Evidence from Rural Tanzania.–Technological Forecasting and Social Change, Future-Oriented Technology Analysis,80, no. 3 (March 2013): 525•40.Kotile, Dido G., and Robert A. Martin. —Sustainable Agricultural Practices for WeedManagement:Implications to Agricultural Extension Education.–Journal of SustainableAgriculture16,no. 2 (August 7, 2000): 31•51.KPMG, —Study on the Costs, Advantages and Disadvantages of Cocoa Certification October-2012.pdf.–Kyei, L., Foli, G. and Ankoh, J.—Analysis of factors affecting the technical efficiency ofcocoa farmers in the Offinso district-Ashanti region, Ghana–American Journal of Socialand Management Sciences2011, 2(2): 208-216: doi:10.5251/ajsms.2011.2.2.208.216Läderach, P., Martinez-valle, A., Schroth, G., & Castro, N. (2013). Predicting the futureclimatic suitability for cocoa farming of the world†s leading producer countries, Ghanaand Côte d'Ivoire.Climatic Change,119(3-4), 841•854.Läpple, D. (2010). Adoption and Abandonment of Organic Farming: An EmpiricalInvestigation of the Irish Drystock Sector.Journal of Agricultural Economics,61(3),697•714.L⁄pple, Doris, and Tom Van Rensburg. —Adoption of Organic Farming: Are ThereDifferences between Early and Late Adoption?–Ecological Economics, Special Section:Ecological Economics and Environmental History, 70,no. 7 (May 15, 2011): 1406•14.Mahrizal, L., Lanier Nalley, Bruce L. Dixon, and Jennie Popp. Necessary Price Premiums toIncentivize GhanaianOrganic Cocoa Production: A Phased, Orchard ManagementApproach. HORTSCIENCE 47(11):1617•1624. 2012.Manda, Julius, Arega D. Alene, Cornelis Gardebroek, Menale Kassie, and Gelson Tembo.—Adoption and Impacts of Sustainable Agricultural Practices on MaizeYields andIncomes: Evidence from Rural Zambia.–Journal of Agricultural Economics67, no. 1(February 1, 2016): 130•53.Martínez-García, C. G., Dorward, P., & Rehman, T. (2013). Factors influencing adoption ofimproved grassland management by small-scaledairy farmers in central Mexico and theimplications for future research on smallholder adoption in developing countries.Livestock Science,152(2•3), 228•238.Meijer, S. S., Catacutan, D., Ajayi, O. C., Sileshi, G. W., &Nieuwenhuis, M. (2015). Therole of knowledge, attitudes and perceptions in the uptake of agricultural and agroforestryinnovations among smallholder farmers in sub-Saharan Africa.International Journal ofAgricultural Sustainability,13(1), 40•54.Melisa Paschall and Don Seville, 2012. Certified Cocoa: scaling up farmer participation inWest Africa. International Institute for Environment and Development/Sustainable FoodLab 2012.Menozzi, Davide, Fioravanzi, Martina, and Donati, Michele. —Farmer†s Motivation to AdoptSustainableAgricultural Practices,– 2015.Moser, Christine M., and Christopher B. Barrett. —The Complex Dynamics of SmallholderTechnology Adoption: The Case of SRI in Madagascar.–Agricultural Economics35, no.3 (November 1, 2006):373•88.Namome, Catherine. —An Economic Analysis of Certified Organic Smallholders in LimpopoProvince, South Africa.– M.Sc., University ofPretoria (South Africa), 2013.Neill, Sean P., and David R. Lee. —Explaining the Adoption and Disadoption of SustainableAgriculture: The Case of Cover Crops in Northern Honduras.–Economic Developmentand Cultural Change49, no. 4 (2001): 793•820.Ntiamoah, A., & Afrane, G. (2008). Environmental impacts of cocoa production andprocessing in Ghana: life cycle assessment approach.Journal of Cleaner Production,16(16), 1735•1740.Osei-Akoto, Isaac, Ampaabeng, Irene and Fenny, Ama Pokuaa. —Customary Land TenurePractices, Investment and Agricultural Production in Ghana– Technical Publication No.76 (2008) Institute ofStatistical Social and Economic Research (ISSER), University ofGhana, LegonPadel, S. (2001). Conversion to Organic Farming: A Typical Example of the Diffusion of anInnovation?Sociologia Ruralis,41(1), 40•61.Pietola, K. S., & Lansink, A. O. (2001).Farmer response to policies promoting organicfarming technologies in Finland.European Review of Agricultural Economics,28(1), 1•15.Quarmine, W., Haagsma, R., Sakyi-Dawson, O., Asante, F., van Huis, A., & Obeng-Ofori, D.(2012). Incentives for cocoa bean production in Ghana: Does quality matter?NJAS-Wageningen Journal of Life Sciences,60•63, 7•14.Sahm, H., Sanders, J., Nieberg, H., Behrens, G., Kuhnert, H., Strohm, R., & Hamm, U.(2013). Reversion from organic to conventional agriculture: A review.RenewableAgriculture and Food Systems,28(03), 263•275.Saltiel, John, James W. Bauder, and Sandy Palakovich. —Adoption of SustainableAgricultural Practices: Diffusion, Farm Structure, and Profitability.–Rural Sociology,59(2), 333-349.Smithers, J., & Furman, M. (2003). Environmental farm planning in Ontario: exploringparticipation and the endurance of change.Land Use Policy,20(4), 343•356.Staatz, John M (1994). —The strategic role of food and agricultural systems in fighting hungerthrough fostering sustainable economic growth.–ResearchGate. Accessed December 3,2015.Teklewold, Hailemariam, Menale Kassie, and Bekele Shiferaw. —Adoption of MultipleSustainable Agricultural Practices in Rural Ethiopia.–Journal of Agricultural Economics64, no. 3 (September 1, 2013): 597•623. doi:10.1111/1477-9552.12011.Tey, Yeong Sheng, Elton Li, Johan Bruwer, Amin Mahir Abdullah, Jay Cummins, AliasRadam, Mohd Mansor Ismail, and Suryani Darham. —A Structured Assessment on thePerceived Attributes of Sustainable Agricultural Practices: A Study for the MalaysianVegetable Production Sector.–Asian Journal of Technology Innovation21, no. 1 (June 1,2013): 120•35. doi:10.1080/19761597.2013.810952.Tey, Y. S., Li, E., Bruwer, J., Abdullah, A. M., Brindal, M., Radam, A., ‡ Darham, S.(2014). The relative importance of factors influencing the adoption of sustainableagricultural practices: a factor approach for Malaysian vegetable farmers.SustainabilityScience,9(1), 17•29.Tropical Commodity Coalition (2010).Cocoa Barometer 2010Tropical Commodity Coalition(2012).Cocoa Barometer 2012Tropical Commodity Coalition(2015).Cocoa Barometer 2015Van Herzele, A., Gobin, A., Van Gossum, P., Acosta, L., Waas, T., Dendoncker, N., &deFrahan, B. H. (2013). Effort for money? Farmers† rationale for participation in agri-environment measures with different implementation complexity.Journal ofEnvironmental Management,131, 110•120.http://doi.org/10.1016/j.jenvman.2013.09.030Waldman, K.B., & Kerr, J. M. (2014). Limitations of Certification and Supply ChainStandards for Environmental Protection in Commodity Crop Production.Annual Reviewof Resource Economics,6(1), 429-449.