Governance challenges of cocoa partnership projects in Indonesia: seeking synergy in multi-stakeholder arrangements for sustainable agriculture

This paper investigates multi-stakeholder arrangements initiated by businesses and NGOs from the North that aim to enhance a more sustainable agricultural production at specific localities in Southern countries. We aim to better understand the search for concerted action in multi-actor arrangements. Therefore, this paper presents a diagnostic framework with three strategic challenges the partnership projects are facing: linking global economic objectives to local needs, values and interests; bridging public and private interests and responsibilities; and seeking trade-offs between social, environmental and economic values. Starting from the partnerships’ Theory of Change, this diagnostic framework is applied to comparative case studies of partnership projects in the cocoa sector in Indonesia, which are part of a Northern-based public–private partnership to improve farmers’ prospective. It is concluded that the economic reality faced by the farmers differs from that of the Northern actors; collaboration with governments is difficult because of different organizational cultures; and the partnership projects underestimate the strength of vested social relations the smallholders are part of. Overall, the initiators of the partnerships seem to work with a too restricted economic interpretation of the local reality.


Introduction
Initiatives to improve the sustainable production of agricultural commodities in developing countries are mainly taken by Northern-based businesses, nongovernmental organizations (NGOs) and many forms of partnership between them. One of the ambitions is to positively influence the livelihood of smallholder farmers as the main producers of these commodities. Through various forms of voluntary private governance, occasionally with involvement of Northern governments, these partnerships have become important initiators of sustainable change in the production of coffee (Auld 2010;Kolk 2012;Manning et al. 2012), palm oil (Schouten and Glasbergen 2011;Brandi et al. 2013;Oosterveer 2014;Hospes 2014;Wijaya and Glasbergen 2016), tea (Glasbergen 2013;IDH 2013), cocoa (Bitzer et al. 2012;KPMG 2013;Vaast and Somarriba 2014), aquaculture (Hatanaka 2010;Douma and van Wijk 2012;Van der Geest and Unno 2012;Bush et al. 2013) and cotton (i.e., Bitzer and Glasbergen 2010).
According to Glasbergen, such partnerships provide a managerial response to the general ethical ideal of societal progress. As self-organizing alliances, they are set up to solve sustainability problems on the basis of a shared commitment of the actors involved (Glasbergen 2011). In essence, they aim to realize concerted action of actors that often have different and even conflicting values and orientations (Austin and Seitanidi 2012;Perez-Aleman and Sandilands 2008). Stakeholders who are used to, or inclined to, act independently must now be stimulated to take part in concerted actions (Ansel and Gash 2008). This asks for some level of alignment or synergy.
Processes of alignment and synergy in multi-stakeholder arrangements have particularly been studied at the level of global institution building in the field of private standard setting and certifying partnerships (Bartley 2011(Bartley , 2014Gulbrandsen 2012;Steering Committee 2012). However, less attention has been paid to private multi-stakeholder initiatives from businesses and NGOs from the North that aim to induce a more sustainable agricultural production in interaction with local public actors in Southern developing countries.
Some of these local initiatives have been initiated as local experiments to test the feasibility and limitations of private standards and certification schemes. Examples include the development and application of the Fair Trade, UTZ and 4C standards (Manning et al. 2012;Manning and Von Hagen 2010). Other local initiatives have been developed as a governance form independent from the certification schemes. This category of multistakeholder arrangements aims to bring together a broad array of public and private stakeholders in temporary organizational arrangements to work directly with farmers at the sites where economic, social and environmental reform is intended to take place. Rather than standard setting or certification, networking is the key lever toward social change here. Notwithstanding the temporary character of these projects, their intention is to induce long-lasting changes among a selection of smallholders that can be scaled up to an entire region. Since these are voluntary private initiatives, local actors might regard them as external incentives for sustainable changes.
This paper takes a governance perspective to improve our understanding of the search for synergy in such partnerships, particularly in the cocoa sector in Indonesia. The governance perspective directs our attention to the process of oriented social change, aiming at the mobilization of problem-solving capacities in multi-actor arrangements. From this governance perspective, we also assume that the practical meaning of the partnership projects has to emerge from an interactive process of dialog, reflection and learning, while recognizing the (problematic) power dimension in their interactions (Pajaro et al. 2010;Meadowcroft 2007;Palmujoki 2006).
In our study, concerted action refers to the ability to induce more sustainable farming practices and eventually sustainable system changes in the Indonesian cocoa sector (Glasbergen and Schouten 2015). Based on the literature on partnerships, we postulate three governance challenges inherent to these arrangements. Each of these challenges depicts a strategic dilemma faced by partnership projects that demand the actors involved to align their interpretations of sustainable cocoa farming. This paper uses these governance challenges as a ''diagnostic framework'' to investigate how the challenges are handled in practice and how this generates a common view on the development of new farming practices. Our ultimate objective is to identify potential encouraging and deterring factors regarding the development ambitions of the partnership projects. We derive these ambitions from a reconstruction of the partnership projects' Theories of Change (ToC). The concept of Theory of Change is particularly used in international development programs. It refers to assumptions, underpinning practices and pathways toward change, and it generally articulates why a given intervention may lead to a desired result (Stein and Valters 2012;Mayne and Johnson 2015). In our case, the ToC postulates a specific sustainable development outcome as a result of a specific Northern intervention in the farming practices in the South. This paper is organized as follows: First, we sketch the problem of cocoa farming in Indonesia. Second, we introduce the case studies followed by the conceptualization of the governance challenges. Next, we explain our research field and methods. The subsequent sections, then, present our research findings. The paper's conclusions reflect on these findings, on the ToC, and on the partnership projects as a form of governance to induce long-lasting sustainable change.
2 On the problems of cocoa production Cocoa production is generally regarded as problematic from different sustainable development perspectives (Suryatin and Beatrice 2013;KPMG 2013;Moriarty et al. 2014). Economically, the problems are centered around the low quality of the cocoa beans, which is generally caused by improper management at the farm level, limited knowledge on good agricultural practices, and over 25-year-old plants (Suryatin and Beatrice 2013;Moriarty et al. 2014;Fahmid 2013). Consequently, cocoa plants are prone to pests and diseases and less productive (Susanto 1994;Shapiro and Rosenquist 2004;KPMG 2013).
Cocoa is one of the main export-oriented commodities of Indonesia. In addition to its value for export and trade, cocoa provides employment opportunities in rural communities and encourages regional economic growth (Antara and Efffendy 2009;Fahmid 2013). Indonesia is the third largest cocoa producer in the world with a total production of approximately 410 thousand tons of cocoa beans in 2014. However, it is still far behind Côte d'Ivoire and Ghana, the top two cocoa producers with a production of 1449 and 835 thousand tons in 2014, respectively (ICCO 2015).
In Indonesia, both the land area and the production of dry cocoa beans decreased over the last 5 years, particularly in South and West Sulawesi. The most significant decline occurred in 2011 when land area devoted to cocoa production in South Sulawesi shrunk by 8.7 % from 273.91 ha in 2010 to 244.47 ha in 2011 (Ministry of Agriculture 2014). At the time, the most severe decline in cocoa prices led to a shift from cocoa to other crops. Prices dropped by 30 % due to an abundant production in Africa that was not paralleled by a similar increase in demand due to the economic crisis in Europe (Bappebti 2012). Because of these low prices farmers lost their interest in cocoa farming and switched to other crops or abandoned their land (Fahmid 2013;Li 2002;Neilson 2007).
Recent international forums have emphasized the need to intensify cocoa cultivation (Vaast and Somarriba 2014). However, it is unlikely that cocoa farmers are able to change their production practices on their own. They are often weakly organized and almost all of them are low income and small-scale farmers with less than 2 ha of land (Fahmid 2013;Saleh 2012). In general, they lack investment capital for technical innovation, they are exposed to price volatility, and do not have much insight into price setting mechanisms (Yanuardy 2014;Moriarty et al. 2014;Vaast and Somarriba 2014). In addition, they have limited knowledge on quality assurance systems and sustainability standards as demanded by the global market (Abdulsamad et al. 2015;Saleh 2012), and the information they do have mostly originates from successful farmers in their neighborhood (Antara and Efffendy 2009). The challenge is therefore to increase the Indonesian cocoa production to meet a growing global demand in such a way that it also improves the livelihood of the (smallscale) farmers (Vaast and Somarriba 2014).
3 The sustainable cocoa production program (SCPP): a case study One of the biggest Northern-based programs addressing problems in the Indonesian cocoa production is the Sustainable Cocoa Production Program (SCPP), initiated and funded by public and private actors. This partnership has been established in 19 districts in 6 provinces in Sulawesi and Sumatra since 2012. Public actors that are involved include the Swiss State Secretariat for Economic Affairs (SECO) and many Indonesian government agencies, while the private actors are the (government-funded) Dutch Sustainable Trade Initiative (IDH), and the businesses Armajaro, ADM Cocoa, Ecom, Mars and Nestlé. All these actors collaborate to provide farmers with training and education in agricultural practices and nutrition, and issues related to environment, community, finance and businesses (Swissconcact 2013).
The projects under scrutiny here are part of this SCPP partnership, implemented by Swisscontact; an international development NGO founded by the Swiss private sector. With over 40 years of experience in Indonesia, Swisscontact mainly works on the topics of tourism and cocoa. Its objective is to contribute to sustainable economic development at the regional level. In this paper, we focus on three regions in Sulawesi, which is the most important cocoa producing area in Indonesia.

Governance challenges: a diagnostic framework
Following Van Dijk and Trienekens (2012) and Vellema et al. (2013), we position the interactions among actors in the partnerships at the crossroads of the ''vertical'' value chain and the ''horizontal'' network. The vertical dimension refers to the value chain the farmers are participating in which relates to supply and demand issues and to the market the farmers are part of. In value chain partnerships the vertical dimension is chiefly represented by Northern-based actors. The horizontal dimension refers to the local socioeconomic conditions and the network of noneconomic actors and non-material conditions in which (local) actors participate, including their relationships with governmental bodies (Oosterveer 2014).
From the farmers' perspective, Northern-based actors, including NGOs backed by multinationals, represent a largely unknown part of the chain. Therefore, they conceived the projects initiated by Swisscontact, at least initially, as external initiatives (from the other end of the value chain). At the bottom of the value chain, the farmers need to make sense of these projects, decide on how to respond to, and handle them. Rather than by global market, the farmers' choices will be supported or constrained by their local socialeconomic and environmental conditions, preferences and social relationships. In brief, at the crossroads of these vertical and horizontal dimensions, global sustainability considerations are likely to meet different interpretations of reality that may not necessarily fit to one another.
Positioning these partnerships at the nexus of vertical and horizontal axes reveals some vital governance challenges to be addressed through interaction between the actors involved. In governance literature these challenges are often debated as potential strategic dilemmas in the search for synergies that would allow the management of long-term systemic changes (see also Boons et al. 2012;Bush et al. 2014;Tallontire et al. 2011).
The first dimension of the challenges referred to in literature is how to link global economic objectives to local needs, values and interests. The partnership projects that we study here introduce a managerial, market-based approach to development, initiated by external actors. As has often been recognized, such initiatives are not power neutral, but unfold within a context of existing resource and power asymmetries. Schouten and Bitzer (2015, p. 175) address this as the ''context contingency'' issue. In the literature, the recognition of this issue induced a (sometimes normative) debate about power imbalances in relation to the change promoted by the initiators of partnership projects. The main question to be addressed is who benefits and who loses as a result of the suggested change (for an overview see Bitzer and Glasbergen 2015). Thus, this issue refers to the problematic, even dilemmatic balancing between top-down interventions, a characteristic of partnerships initiated by Northern-based actors, and the room left for local interpretation and adaptation, deemed necessary to create local ownership and thus the effective implementation of the initiative.
The second dimension refers to discussions about the relationships between private and public responsibilities in partnerships projects. Partnership literature generally recognizes that the trajectory of private governance cannot be addressed without simultaneously considering the trajectories of public governance (Mayer and Gereffi 2010). Gulbrandsen (2012), for example, suggests that private standards and public policies can reinforce each other, particularly in cases where the state depends on private activities to realize its public objectives. Glasbergen (2013) shows that private sustainability standards need political backing and alignment in governments to become legitimate. The FAO (2014) also refers to potential synergies that can be effectuated through hybrid forms of governance in which voluntary standards interact with public regulations and institutions. Therefore, we assume that to effectively work in Indonesia, the partnerships need to adapt to the local regulatory conditions in order to align private initiative with public policies. Government support is necessary to open avenues for sustainable change. However, such support needs to be realized in a fragmented power structure where different levels of government with different stakes in the process of change are involved. Notwithstanding their different stakes, these multi-level governmental bodies need to align their practices among them. The question is how governments can get involved in the partnership projects in a way that allows them to develop capacities to sustain changes at the farmer's level and to preserve the advantages in the long run.
The third dimension refers to governance as a collaborative learning process in which the aim is to develop a shared understanding of reality and the activities that are needed (Ansel and Gash 2008). The partnership projects researched here face the challenge of reconciling divergent interests, representing the economic imperative of trade in cocoa and noneconomic social and environmental values to which this imperative relates. This reconciliation process includes the framing of sustainability in cocoa farming, which is an inherently normative process and rooted in real world problems with different sets of values and moral judgments (Robinson 2004). Trade-offs may easily occur when dimensions of a concept are competing. This is of course inherent to the concept of ''sustainable development'' that aims to juxtapose social, environmental and economic objectives (Matzdorf and Müller 2010). Klapwijk et al. (2014) define conditions that make trade-offs ubiquitous, such as managing land with multiple ambitions and goals in mind, constrained resources, and conflicts between the goals of the different stakeholders. Therefore, a central dilemma in sustainable change processes is what choices to make in practice, i.e., what value dimensions of sustainable cocoa production need to be prioritized, probably at the detriment of the others and why.

Research methods and research field
This paper applies an interpretative research methodology to a comparative case study design. Interpretative researchers take the existence of multiple interpretations of reality as a given and regard meaning-making as key to the scientific endeavor. The aim is to interpret events, that is, to understand how actors make sense of an issue and why they do so. According to this approach, understanding and interpretation originate from interactions between researcher and the researched (Schwartz-Shea and Yanow 2012; Yanow and Schwartz-Shea 2014).
Our starting point is a reconstruction of the Theory of Change underlying the partnership projects. This reconstruction is based on an analysis of partnership documents and interviews with the program director and field coordinators of the partnerships.
Next, we use the aforementioned strategic dilemmas as different lenses to analyze the practices of the partnership projects. Our interactions with the research field were conducted from December 2014 to April 2015 and included semi-structured interviews, document analysis, a focus group, and observations. First and regarding the semi-structured interviews, respondents were chosen through purposive sampling. We interviewed the program director and the field coordinators of Swisscontact to obtain information about the background and the development of the partnership projects. Subsequently, we interviewed informants from (local) government agencies and companies who have been engaged in the implementation of the program. In total we interviewed 43 people from various organization: 26 government officials and staff from the national, provincial and district governments; 7 staff members from Swisscontact; 3 representatives from donor organizations and companies; and 7 other actors in the Indonesian cocoa sector and projects, such as the Cocoa Sustainability Partnerships (CSP), a sustainability standard body and a cocoa community forum in Sulawesi. Interviews covered the perception of the involved actors on the projects as an external initiative; their perception of implementation problems; the engagement with various actors, including local governments; the role of involved actors; and the anticipated future of the projects. In sum, the interviews mainly informed us about how the governance of the projects worked out. The results of these (recorded and categorized) interviews are the main data of this research. Their analysis allows us to comprehend the different views of and interaction between the actors.
Second, we analyzed public documents about the cooperation between Swisscontact and government agencies. Annual reports and work plans issued by Swisscontact were important to understand the progress of the projects in each district and the role of implementing actors. In addition, we analyzed two kinds of documents from governmental bodies: The Strategic Plan and the Report of Accountability Performance of the Government Agency (LAKIP) of Forestry and Estate Crops. The analysis of these public documents enables us to comprehend how the actors deal with the strategic dilemmas and what they consider to be the core of problems in the cocoa sector.
For example, to comprehend the perspective of the governments on the cocoa development, we analyzed the Strategic Plans from 2013 to 2018 of the Forestry and Estate Crops Agency of the Districts of Parigi Moutong and Mamuju, the District of North Luwu for the years 2010-2015, the Report of Accountability Performance of the Forestry and Estate Crops Government Agency (LAKIP) of the District of Parigi Moutong, Mamuju, and North Luwu in the years 2013 and 2014. These documents helped us to understand the engagement and contribution of the government in and to the projects. We also analyzed statistical documents on cocoa land use and productivity that reveal trends in cocoa production in Sulawesi. This helped us to indicate the achievement of the projects so far.
Third, we conducted a focus group discussion (FGD) with a ''Tallusikambi'' farmer group in the Mamuju district, West Sulawesi. This farmer group consists of alumni of the SCPP Field School. Their chairman is a key farmer, appointed by Swisscontact to disseminate information from SCPP. The FGD took place on the 10th of March 2015 and was attended by 21 farmers. During the FGD we discussed benefits and achievements of the SCPP for farmers and the role of governments in the project. Farmers also shared their concerns and expectations about the prospective determination of the project by Swisscontact. The information from the discussion was essential to better comprehend the position of farmers in the program and their perception of the field school.
Lastly, we conducted observations, for example, we observed activities and meetings like the Training of Trainers (ToT) for key farmers in the North Luwu district, South Sulawesi, and interactions between facilitators, government officials and participants. This information complemented our understanding of the farmers' perspective.
We conducted our comparative case study in Sulawesi, which is the most important cocoa producing area in Indonesia. Sulawesi covers approximately 70 % of the national cocoa production (Ministry of Agriculture 2014). Our research took place in three districts in three provinces. Although the project areas in each province spread over more than one district, we restricted our analysis to one district per province (see Table 1).
The site selection was based on the diversity of the participating actors and the progress of the projects. Mamuju was chosen because it became a pilot project location. Here, the Indonesian Coffee and Cocoa Research Institute (ICCRI) coordinated a project to develop experimental farms. The Parigi Moutong project has the least numbers of partners due to its vulnerability for social conflicts, which may affect the implementation of the project. In North Luwu the project collaborates with Rainforest Alliance (RA) to support sustainability certification of cocoa.

The theory of change underlying the partnership projects
In this section we analyze SCPP's strategies through reconstructing their underlying Theory of Change (Anderson 2005;Vogel 2012). A ToC is not a scientific theory, yet the concept refers to an amalgamation of visions, ideas, conceptions, wishes and hypotheses that (policy) practitioners have about the causal processes and mechanisms through which change occurs as a result of (policy) programs. We use the ToC as a heuristic tool to identify and discern different elements of Swisscontact's program and practices. ToCs have ontological, normative and strategic features, i.e., about what reality is, why reality may be problematic, and how problems should be solved, respectively. As will become clear below, these three aspects are intermingled. Our reconstruction reveals the following characteristics: • First, within the projects, the reality of cocoa production is largely framed as an economic problem. Low productivity and poor quality are considered the main problems that should be solved by improving the knowledge and skills of farmers. A core target of the program is to achieve productivity levels of at least 1.000 kg/ha per year, which is two times higher than the current level of production. • Second, it is assumed that increasing profitability and enlarging the scale of cocoarelated businesses will enhance household incomes, living quality, the creation of new jobs and subsequently the development of the regional economy. The program also aims to encourage cocoa smallholders to apply sustainability standards and certifications, which is believed to enhance the long-term prospects of the Indonesian cocoa sector. • Third, the SCPP sees the engagement of private actors, who need reliable cocoa supplies for the survival of their industry, as essential for the effectiveness of the program. Private actor engagement could link farmers directly to the market. • Fourth, it is assumed that the farmers' knowledge base and capacities regarding good agricultural practices, pest management, post-harvest handling, and business or managerial matters, are essential to realize SCPP's objectives. Training includes private sector officers and governmental representatives, in particular the so-called extension officers (see below) who are directly involved in assisting local farmers in their production processes. • Last, collaboration with (local) governments is regarded essential to enhance the implementation of the program. This relates to the previous point, as the involvement of extension officers requires an official cooperation with governmental agencies. Based on the agreements, the governments are also responsible for the continuation of the program, in which they are supposed to adopt and integrate into their policies after determination of the project (Swisscontact 2012(Swisscontact , 2013.

The first challenge: balancing the global and the local
The initiators of the partnership projects perceive low productivity and low quality as the main problems of the cocoa sector. Through the projects, they intend to ensure the supply of good quality cocoa beans. In addition, these global actors regard farmer capacity essential to improve quality and productivity. However, at the producer level farmers perceive a different economic reality.

Lack of capital is seen as the most urgent problem
The degree to which farm land is maintained by the farmers is influenced by the farmer's capital. Although the partnership projects provide knowledge to improve farmer's capacity to better till the land, they do not solve what the farmers see as their main problem: the lack of capital to invest in, and improve, the farm. As long as this problem is not solved, it is unlikely that local conditions in the cocoa sector will be sustainably improved, as expressed by an official from the provincial Estate Agency: Many cacao projects coming to this region neglect the issue of economic strengthening. We expected that the management they brought, would lead to improved welfare. What I have seen so far are only trainings. But these trainings cannot solve the real problem. Indeed, farmers do have better knowledge, but their financial ability is still limited. Therefore, they are still unable to improve their farm (23 February 2015).
Limited capital to invest in the cocoa farm is also mentioned by farmers in the focus group discussion: The Swiss NGO gave us knowledge about better farming and marketing, which we consider comprehensive knowledge. However, not all farmers directly apply the lessons learnt into their daily farming practices because they lack capital. We expect to develop a kind of cooperative to help farmers in collecting capital to invest in their farms, but this is also not easy as we need more capital to establish a cooperative (10 March 2016).

Lack of short-term economic profit
For farmers, achieving a higher income is important. By targeting a productivity of at least 1000 kg/ha per year, the projects offer the prospects of a higher income, which is supposed to attract farmers to participate in the activities. For farmers, however, this prospect does not resonate a short-term economic opportunity and implies the careful maintenance of their cocoa plants for quite some time, before they can observe an increase in productivity. Although they are attracted to the prospect of an increasing income, the deferred profit averts them from participating in the projects. For farmers, selling their crops to middlemen is still a better option to quickly earn cash money than selling it to the trader with certain requirements.

The farmers are seeking the economically most attractive commodity
Farmers do not have advanced knowledge to fully understand the causes of low productivity and quality. They also lack a sense of leadership and entrepreneurship needed to maintain farmer organizations. Many farmers easily lose their interest in their farm and instead of improving their cocoa farming they prefer to shift to other cash crops when these crops promise higher profits, as indicated in some interviews: In Mamuju, West Sulawesi, cocoa-land areas decrease in size, because many farmers shift their land to oil palm and patchouli plantations. Many farmers also give up because they cannot fight against the pests and diseases attacking the cocoa plants (Staff of the District Forestry and Estate Agency, 2 March 2015).
Cocoa in Parigi, Central Sulawesi is a specialty of this district; however, in the last years due to old plants, the production decreased. From the farmers' point of view, a higher price is needed to maintain their cocoa farms. Because they do not have enough capital to improve their farm, it is more interesting to shift their land to other commodities (the NGO's field coordinator, 13 March 2015).
Approximately 70 % of the cocoa export comes from South Sulawesi, particularly North Luwu; however, the cocoa area here is decreasing in size. A large portion of nonproductive land has been transformed into rice paddy fields or other commodities after being confronted with pests and diseases. The tendency of South Sulawesi farmers to easily shift their land results from the higher prices for other commodities in the market, such as palm oil and patchouli. When cocoa prices are good, everyone plants cocoa, but when the price goes down, farmers easily shift to other commodities (Staff of the District Forestry and Estate Agency, 31 March 2015). Table 2 shows the decline in cocoa land area and production in all three regions, with the exception for productivity (kg/ha) in Mamuju and Luwu where we observe an increase when comparing levels in 2010 with levels in 2014.

Only a limited group of farmers reached
The partnership projects focus on an empowerment approach through the training of trainers (ToT) and farmer field schools (FFS). ToT is targeted at key farmers (village lead farmers of farmer groups) and extension officers. During the training, they learn about good agricultural practices, nursery management, post-harvest processing, top grafting techniques, and are given an introduction to sustainable cocoa certification and good business practices. Moreover, as they are responsible for spreading this knowledge to other farmers, they are also taught in the area of facilitation and presentation skills (Swisscontact 2012, p. 11). The activity of ToT is followed by FFS. In the FFS, farmers obtain the technical and practical skills needed to improve the farm productivity. Usually, the FFS takes place once a week for every farmer group with 25-30 farmers participating. Our research shows that not all assisted farmers in the projects can apply the newly acquired knowledge to their daily practices. This is partly due to the already mentioned capital requirements, but also to a lack of motivation. In addition, since the projects are limited to selected farmer groups, they only reach a small part of the farmers in Sulawesi, which are not the most vulnerable ones. Swisscontact began the selection process with identifying potentially motivated farmers in designated areas through village officials, and the extension offices in the sub-districts. Requirements determined by Swisscontact were a minimum land area of 0.5 ha with 300 trees, the availability of a farm for a demonstration plot, and a clear organizational structure of farmer groups. This selection procedure, however, is questioned by local actors, particularly the district Forestry and Estate Agency, as expressed in the interview with one of them: We want the NGO to assist and develop farmers from scratch, not to assist farmers whose farms are already in a good condition. I feel that what they do in this project is looking for the better farms, establishing farmer groups, then making up and reporting it into a good publication (31 March 2015).

No changes in power structures
As mentioned above, the projects forecasted an economic profit for participating farmers. The projects encourage farmers to establish and improve their organizational capacity to improve their position in the value chain. Strengthening farmer groups, however, is not easy due to some internal and external constrains. Our research indicates that the effectiveness of the farmer groups depends on the motivation of the leader and the members. However, even with a very motivated leader, as in one of our cases, it is often difficult to motivate the members to better maintain their cocoa farm. One reason for this lack of motivation is the fragmented farm land, with cocoa plots that are often far away from the village where the farmers live, which makes it difficult to practice the labor intensive farming. An external constrain relates to governmental policies. For instance, although in Governance challenges of cocoa partnership projects in… theory subsidized fertilizer is available, in practice the cocoa farmers have limited access to it as the regulations are not very transparent and the majority of the government budget is attributed to rice paddies. We see that the projects do not necessarily create a stronger bargaining position for the farmers in the value chain. Farmers in the partnership projects are still in a rather powerless position, as they hardly influence on the cocoa prices predominantly set by the buyers. Hence, farmers easily take a realistic opportunity to transform their cocoa plantations into other crops such as oil palm, corn and patchouli, especially when fellow farmers have already done so successfully (Antara and Efffendy 2009;Saleh 2012).

The second challenge: balancing public and private
The partnership projects are part of the largest public-private cocoa partnership in Indonesia, bringing together multinationals, a research institute, an NGO and Indonesian governmental agencies at various administrative levels. In Indonesia, every project funded and implemented by foreign organizations must be formally acknowledged by the national government. However, our research reveals many challenges related to the involvement of the government.

Fragmented responsibilities
A legal basis of the partnership projects was set out in the cooperation agreements between Swisscontact and district governments in 2012, which marked the start of the implementation. Swisscontact cooperated with two governmental bodies in every district: the Forestry and Estate Agency (Disbunhut) and the Agency of Agricultural Extension. Both governmental bodies work on cocoa on a farmer level and can therefore support the projects. The district Disbunhut played a role as a facilitator and provided a database, technical staff and funds for an assistance program, while the Agency of Agricultural Extension played a role in assigning the extension officers to participate in the ToT of the projects.
Shortly after the formal start of the projects, the projects came to the next level of coordination following the Memorandum of Understanding (MoU) between Swisscontact and the Ministry of Home Affairs (MoHA) under the umbrella of the Community Economic Development Program, signed on June 14, 2013. The MoU was needed since SCPP operated in areas that were spread over many districts in two large islands, which made SCPP formally become a ''national project.'' The MoHA attempted to establish a coordination mechanism by bringing in responsibilities for provincial governments. This coordination responsibility at the provincial level was supposed to be handled by the Regional Development and Planning Agency (Bappeda), which applies also at the district level. 1 In the project's scenario, the MoHA played a role as a regulator and a facilitator. The regulatory role results from Indonesian regulation, in which the presence of Swisscontact, as an international NGO in Indonesia, must be in accordance with the Regulation of the Ministry of Home Affairs No. 15, 2009, regarding guidelines for cooperation between the Ministry of Home Affairs and non-governmental organizations. Under the Ministry of Home Affairs, Swisscontact coordinates with the Directorate General of Regional Development as they have similar interests in regional economic development. The facilitation role refers to connecting Swisscontact and the projects to local governments. In the table below, we summarize the roles of public and private actors in the three projects under scrutiny (Table 3).
With many public agencies and various private actors involved, a transparent allocation and distribution of responsibilities are crucial. As one can draw from the table, each agency has specific roles and makes specific contributions into the projects based on its Governance challenges of cocoa partnership projects in… competencies, capabilities and resources. However, the fragmentation of the latter is very likely to create tensions between the public and the private actors or within the public sector, e.g., between government bodies at different levels.

Different organizational cultures
Our research particularly reveals such a tension between the provincial governments and the NGO. At the provincial level and according to the MoU, two agencies are supposed to support the projects: the Regional Development and Planning Agency (Bappeda) and the Estate Crops Agency. These agencies have the same name and function as their counterparts at the district level. In the three districts, the provincial Estate Agency provided work space for the NGO's staff members. In practice, however, the role of the provincial agencies was rather unclear; although they were expected to facilitate and monitor the projects, there was no intensive communication between the NGO and these provincial governments. This was mainly due to the fact that provincial governments were involved a year later than the district governments. Therefore, the Regional Development and Planning Agency at the provincial and district levels responded quite late in fulfilling its role in the projects. This was recognized by one of the officials of this agency: We knew that we (Bappeda) were involved when the national team visited us in October 2014. During that meeting, we were shocked to know that we should play a role based on the Memorandum of Understanding. These documents were not in our hands before, that is why we did not do anything relating to the project. They (The NGO and the district Disbunhut) were supposed to tell us in the beginning about the project. We thought that this project was between the district Disbunhut and the NGO only (2 March 2015).
In addition, all government agencies requested regular reports on the projects' activities and progress from Swisscontact. They also asked for coordination meetings in order to align the projects with government programs. Swisscontact, however, focused on the projects' achievement in terms of profit and sustainable cocoa, as mentioned by the field coordinator of the NGO during our interview: Coordination with the governments is still weak although we have attempted to communicate and coordinate as much as possible. Our organizational culture is quite different; the government is too bureaucratic. While we focus on the work in the field, the government officials are difficult to meet since it is difficult to match with their schedule (18 February 2015). This issue is highly representative for public-private partnerships (Brinkerhoff and Brinkerhoff 2011;Shapiro and Rosenquist 2004). Both parties face difficulties in obtaining and sharing information about progress since they have different opinions on the what's and why's of reporting due to their diverging interests. While government agencies ask for information to fulfill administrative and bureaucratic obligations, the NGO focuses on concrete project achievements. These differences in cultures of public and private organizations easily result in miscommunication.

Unclear relationships between government agencies
The lack of coordination also applies to the public sphere itself, both between different governmental agencies, such as the Disbunhut and the Bappeda, and within each agency between the head and staff. As we observed in the second project region, the frequent change of the Head of the Disbunhut had a negative effect on the information provision. Staff members and Agency's new director were hardly informed about the projects. This personnel rotation happens frequently, as it is part of the decentralization policies, with frequent changes in political leadership and interests in the region as a result. As a consequence, the NGO staff had to explain the partnership projects to the government staff over and over again, which negatively influenced the progress made.
The Ministry of Home Affairs (MoHA), through the Memorandum of Understanding with Swisscontact, has a monitoring and evaluation system for the projects. The MoHA expected the coordination in each territory to be handled by the Bappeda. Better internal communication among government agencies would make the coordination more effective albeit to relieve Swisscontact's burden of reporting to all governmental agencies. The NGO expected this latter function to be handled by the Bappeda as stipulated in the MoU.
We built an intensive communication strategy through regular progress reports and meetings with the district governments because our implementation areas are in those districts. We expected that the district governments would disseminate the progress reports to other related governmental agencies under their authority, however, this did not happen (19 January 2015).

Conflicting private and public policies
Another tension regards conflicting policies between public and private actors. In project number three, this became manifest in the different purchasing policies for cocoa beans. The implementing partner in this district prefers to purchase wet beans from farmers. This preference, however, contradicts with the Regulation of the Ministry of Agriculture Number 67, year 2014, regarding quality and marketing requirements that necessitates farmers to ferment their beans before selling them, as the fermentation is supposed to create added value and a better price for the farmers. The district Disbunhut faced difficulties to get this policy internalized as farmers prefer selling wet beans. Farmers believe that they gain more profit with wet beans compared to fermented beans, as selling wet beans saves energy and time. Based on our interviews with the field coordinator of the NGO and the official of the district Disbunhut (March 31, 2015), we calculate that farmers' profit from wet beans is at least 15 % higher than from dry beans. Conflicting regulations like the one referred to above, still persist and may even intensify in the future as the government cannot effectively interfere with business issues. In the end, farmers risk to be mostly affected by this divergence and confusion.

Partnership objectives weakly integrated in government policies
The involvement of Indonesian government agencies should ensure that the partnership projects align with the government's program regarding the development of the cocoa sector. This alignment comprehends a joint problem definition, avoiding overlapping work, and ensuring long-term effects of the projects. Based on our document analysis, we found that the projects are not (yet) integrated in the annual program of the governments. The governmental reports do not mention the projects at all. Moreover, although cocoa can be considered to be a specialty in the three regions, there are not many governmental activities focusing on the development of this sector. This is caused by the fact that the agency's attention and budget for agriculture needs to be shared with other commodities, such as food crops and livestock, which are generally prioritized over estate crops such as cocoa in national, provincial and district policies.
The first project region is an exception here as the performance report written by the Forestry and Estate Crops Agency in 2014 mentions two programs focusing on cocoa. The first program focuses on improving farmer welfare and provides extension, training, marketing, and the development of demonstration farms. The second program focuses on improvements in agricultural production and provides agricultural facilitation, such as fertilizer, farmer field schools for cocoa farmers to improve human resources and to provide good quality seedlings. These programs are in accordance with the strategic plan of West Sulawesi Province, as explained by an official of the Provincial Bappeda: The main contributor to economic welfare in our province is the plantation sector, i.e. cocoa and palm oil. As a new province, the first priority (2006)(2007)(2008)(2009)(2010)(2011) is to lower the number of people in poverty, in particular cocoa farmers, because in our region cocoa farmers are dominant. By improving cocoa farmers' welfare, we expect that poverty will decrease (3 March 2015).
Furthermore, our analysis of the public documents of the three regions shows that the governments apply a technical approach to address improvements in the cocoa sector, e.g., programs providing training, field schools and seedlings. As a consequence, and given the governments' limited budget, they only reach a small number of cocoa farmers. Therefore, one might question whether the governmental policies are sufficiently effective to solve the aforementioned problems in the cocoa sector. Engagement with private actors through the cocoa projects could potential fill the mentioned gaps in the governmental policies, but this has not been realized yet.
The partnership projects should not be evaluated on their short-term achievements only, but also in terms of their contribution to the institutionalization of long-term changes. The involvement of local governments should ensure this institutionalization. As the projects are temporary, local governments have the responsibility to continue the spirit of the projects based on the MoU after determination of the projects. This is in accordance with the instruction of the MoHA, stating that local governments should have an exit strategya strategy to adopt the projects into their regional policies when the projects have expired. However, only one local governmental agency has prepared this strategy-the district Disbunhut of Mamuju, West Sulawesi (the first region). In its program for 2015, it utilizes farmers' skills gained from attending the field school meetings of the project. In 2015, the district Disbunhut changed its method from distributing seedlings toward establishing nurseries and providing equipment, so that farmers can actually apply the learnt skills in the field of top grafting the seedlings. In doing so, the government encourages farmers to take part in the process and it is expected that the farmers will feel more responsible in maintaining their farm. This seems a first sign of readiness to continue the cocoa sustainability project in the region after the SCPP project expires.

Low involvement of extension officers
In their attempt to connect the global initiative to the local level, the projects involved agricultural extension officers; staff members who are responsible for assisting farmer groups in villages related to agricultural practices. Their role can therefore be conceived as boundary workers as they transfer and translate knowledge from Swisscontact to the farmers (McMillan 2011;Tisenkopfs et al. 2015). Their involvement seems an adequate strategy from the NGO to help them in training farmers in the field schools. It is also a strategy to ensure the sustainability of the projects, given their position as governmental officials who are supposed to continue to assist farmers (Swisscontact 2014, p. 15).
However, although extension officers and local governments will continue to work with farmers in the field, there was some doubt among our respondents whether the extension officers would keep focusing on cocoa farming after expiration of the projects, because their responsibility (and therefore maybe priority) covers other crops that are regarded more important, such as rice paddies. During the FGD, cocoa farmers expressed their disappointment about the inactive role of government extension officers in their villages, particularly in providing farmers with information about governmental agricultural programs. Consequently, the cocoa farmers sensed that they get less attention from the government than farmers with food crops. In an attempt to fill the gap resulting from the inactive role of agricultural extension officers at the producer level, many NGOs and private companies nowadays assign their agronomists and field officers to train cocoa farmers about good agricultural practices. This private extension, which is embedded in many partnership programs, could help farmers to improve their cocoa farm.

The third challenge: balancing values
The third dilemma is to find a balance between economic, environmental and social aspects of a change toward sustainability. As mentioned before, the Theory of Change underlying the projects frames the cocoa problem as predominantly economic problem. This problem definition shapes the activities of the projects that focus on increasing production and improving quality.

The subservience of environmental issues
Although the ToC mainly refers to the economic aspects of cocoa production, environmental issues are embedded in at least three main activities, namely training the farmers and field schools, the rehabilitation of cocoa plants and the certification schemes (Project report, 2012).
First, in the trainings and field schools, farmers participate in outdoor activities to learn about good agricultural practices. The training includes post-harvest processing to meet quality standards. One of the topics in the training and field schools is the sustainable use of fertilizer. For example, as stated by farmers in the FGD, one of the benefits of the NGOdriven training was that they learnt that certain fertilizers do not benefit cocoa plants and might even endanger the environment. Second, the rehabilitation of cocoa plants: most cocoa plants in Sulawesi are old plants, about 20-25 years old, which make them vulnerable to disease that decreases productivity. Therefore, farmers are taught grafting techniques and encouraged to establish nurseries.
Third, the projects also promote sustainability certification of cocoa. During the trainings and field schools, farmers are introduced to sustainability certification schemes, i.e., UTZ-certified, Rainforest Alliance and Fair Trade. Some principles mentioned in the certification schemes, for example good agricultural practices and the use of fertilizers, are taught to the farmers. Through certification, environmental aspects of cocoa farming can get more prominence. However, from our FGD and observations, we infer that environmental issues can only get meaning for farmers if they are framed as economic prospects.
Farmers prioritize the economic aspects of farming and their social relationships (see below) over environmental issues.
This prioritization issue is also visible in palm oil and coffee certification schemes. According to Hidayat et al. (2015), palm oil smallholders perceive certification as an economic tool to pursue a better livelihood rather than a tool to ensure a more sustainable agriculture. Other than economic values are less appreciated by smallholders, unless they obviously result in economic benefits. Moreover, since there are many coffee certification schemes in Indonesia, farmers prefer the ones that offer a tangible price premium (Ibnu et al. 2015). Economic prospects are therefore the single most important factor for farmers in their decision to join a specific certification scheme. Regarding Indonesian cocoa, there is no clear evidence that certification schemes create economic benefits for farmers, as many traders pay and hold the certificate, without providing transparency on how the premium price is distributed to farmers (Moriarty et al. 2014;Abdulsamad et al. 2015). Therefore, cocoa farmers hesitate to participate in certification schemes unless they are ensured about the economic benefits that they will receive.
The projects are committed to facilitate farmers in the process toward certification. The NGO perceives certification as a levier to improve the competitiveness of the Indonesian cocoa sector in the global market. Certification schemes function as a guarantee for consumers that production takes place in a more sustainable way. Farmers could benefit by receiving a premium price once their farm is certified and this may improve the long-term prospect of cocoa production. Moreover, as reported by the NGO, the certification process also has created job opportunities, for example, in farmer organizations and the internal control system (ICS). Among the three locations mentioned in this paper, only the third project in North Luwu has already been certified by Rainforest Alliance (RA). Here, Rainforest Alliance (RA), together with Swisscontact and PT Mars, is supporting cocoa farmers in maintaining their cocoa farms in accordance with sustainability standards. In this district, the implementer of the internal control system (ICS) for RA is CV Marewa 45, which is a business unit under the ''Masagena'' cooperative. Since 2013, this cooperative receives support from VECO, PT Mars and Wasiat (a local NGO). An earlier project than SCPP has thus obtained a Rainforest Alliance certificate. This certificate is accredited to 1074 farmers with a total land area of 1306 hectares of cocoa plantation (VECO Indonesia 2014, p. 11). In North Luwu, the holder of the RA certificate is the farmers' cooperative. Already from their second year of RA certification, farmers could take advantages from the certificate, such as access to better marketing channels, the receipt of a price premium and a formal contract that allows selling the certified beans directly to PT Mars. Being informed about the advantages of certification by their fellow farmers, many farmers are now interested in joining the cooperative to become certified, even though it is rather difficult to apply the standards of RA certification (interview with CV Marewa, April 3, 2015). Through the SCPP projects, the representatives of CV Marewa 45 and RA actively encourage farmers to comply to the principles of good agricultural practices.
Farmers' enthusiasm in responding to certification can also be seen in the first project location in Mamuju, West Sulawesi. In this district, farmers are given information about sustainability certification by Swisscontact. Despite the 90 requirements, farmers are encouraged by the economic advantages that they will get later. By the end of 2015, farmers were preparing to be examined by an external auditor from UTZ certification with the idea that at least 1500 farmers can get certified (FGD, March 10, 2015;Swisscontat 2015).
However, in the last decade many cocoa farmers have shown less interest in cocoa cultivation and preferred transforming their cocoa farms into oil palm plantations. The projects play a role in convincing farmers to continue their cocoa plantations. Through the projects, farmers became aware that cocoa is more environmentally friendly than oil palm. Oil palm demands more water and change to other commodities is much more difficult in the case of oil palm since the soil will generally be too dry and depleted to be used for other crops.
Moreover, the government officer from the district Disbunhut provided additional information about the pros and cons of oil palm compared to cocoa for farmers: The conversion from cocoa to oil palm has become one of the threats in West Sulawesi. We try to defend our cocoa plantations by promoting programs on cocoa. We also consider environmental issues. We realize that farmers do not have proper information about oil palm. They think that oil palm is more profitable, but this is not always true. Farmers only follow others with the idea that oil palm does not require intensive care like cocoa. However, many farmers start to regret their decision, because cultivating oil palm is costly and it is difficult to sell the fruits individually (9 March 2015).

The value of persistent social relationships in economic transactions
One of the projects' strategies is the creation of a shorter supply chain. By bringing trader companies in the projects, farmers can directly sell their beans to these trader companies instead of middlemen or local collectors, who are by-passed. This provides farmers with an opportunity to gain better and more competitive prices. The representative of a buyer company in the first region explained the advantages of selling cocoa beans directly to traders as follows: We create advantages for assisted farmers who sell their beans here. First, we provide a free sack with a logo of the project brand. Second, farmers receive a premium fee every three months (approximately 2 % of the average price per kilo). Third, reward points are given to farmer groups based on their selling, which can be accumulated on a yearly basis and exchanged for farming tools. Fourth, farmers will get a higher price for fermented beans compared to unfermented beans. In addition, together with the NGO we provide farmers with equipment to improve the quality of the cocoa beans and marketing (March 4, 2015).
And yet, although assisted farmers are assumed to choose for the shorter chain, some of them still prefer selling to middlemen as these transactions are far more than economic ones. A strong social relation exists between farmers and middlemen, either through kinship or through a debt legacy. Moreover, middlemen are usually open to provide cash money whenever the farmers need it. These debts can be paid back in the future with their crops. As a consequence, the farmers in the projects face a dilemma between economic rationality and social considerations: The opportunity to increase their profit might go at the detriment of their social relations with their neighborhood middlemen.

Gender concerns difficult to implement
Similar to other typical development projects, and related to social values, the projects promote the inclusion of women in cocoa farming. The projects have targeted 20 % female participation in agricultural practices (Swissconcact 2013, p. 14). This gender concern is integrated in the activities of the farming field schools, and women receive material about family nutrition and finance. Generally, women's role in cocoa production is rarely recognized, which is also the case in Sulawesi. Although women are highly involved in preand post-harvest work, they hardly participate in farmer groups or other public organizations. As we found in the ToT activity in the third project location, only one female representative of a farmer group attended the event among thirty male representatives. Most farmer groups are already established with a male leader. We found one exception in the FGD in the first project location where the majority of members of the ''Tallusikambi'' farmer group were women. The objective of women inclusion in farmer groups is that women play an important role in encouraging their husbands in maintaining their farms and in supporting the household economy. From the observations in the two locations above, we learned that there is a work division between women and man in cocoa farming. Mostly, men take part in public activities, such as meetings and trainings, while women are more active in farming practices in the field.

Conclusions
This research investigated the strategic issues and dilemmas of (sustainable) cocoa production in Indonesia, focusing on three districts in three provinces in Sulawesi Island, covering the 2012-2015 period. While we should be careful with generalizations, the partnership projects we studied are similar to other projects initiated by Northern-based businesses and NGOs in the Indonesian cocoa sector.
The overall assessment of the projects self-evidently depends on the scope and criteria one starts with. From a narrow interpretation, one might easily conclude that the projects are doing rather well, as they reached a lot of farmers, organized training and achieved that the participating farmers are better prepared to improve their production. However, this picture changes when we take a broader scope on the problems of the cocoa sector and the opportunities for long-lasting changes. The low productivity and poor quality of cocoa beans are generally regarded to be essential issues of cocoa farming in Indonesia. The Theory of Change underlying the partnership projects recognizes this problem definition and connects to it through its emphasis on training, organizational capacity building, and building sustaining relationships with governments.
Our first observation is that our conceptualization of the three governance challenges is valuable to understand the complex problems the projects face in practice. The way in which these strategic dilemmas are handled largely determine whether a global partnership between Northern NGOs and businesses will result in a well-functioning partnership with relevant actors at the local level and therefore will induce the envisioned system changes at the production level in the South. Our research revealed tensions in the search for synergy with regard to each of three governance challenges.
On the global versus local dimension, we observe tensions between framings of the problem by the Northern initiators of the projects and the farmers. While the former stress training, the latter emphasize the need for capital and short-term profit. The prioritization of economic values also plays a role in farmers' attitude toward sustainability certification.
On the public versus private dimension, we observe divergent organizational cultures, miscommunication, coordination failures and problems with allocation of responsibilities. Similar coordination problems are visible among Indonesian government bodies as the various administrative layers display varying political priorities and political interests in the cocoa sector.
Regarding the reconciliation of diverging values and interests, the partnership projects tend to underestimate or even ignore the traditional relationships at the local level. While the shorter chain strategy envisages an elimination of the role of middlemen and local collectors, farmers need them to get instant money. Moreover, their relationships with middlemen are more than only economic in nature; they are an essential part of the community's social interactions.
Concerning the institutionalization of a system change our research results into a rather pessimistic conclusion. First, because the reach of the projects is limited and only represents a small selection of cocoa farmers, marginal farmers outside the project areas, who encompass the majority of cocoa farmers in Sulawesi, are not reached. Second, intuitively and from our empirical data, one can anticipate that farmers will still easily shift to other crops in the next future if the return on investment from other commodities is higher than from cocoa. Third, we recognized a lack of ownership of the public actors. It is very hard to anticipate them to translate the projects' objectives in government policies. An illustration of this issue regards the future role of extension officers as they will certainly continue to work with farmers, also when the projects expire. As the extension officers' responsibilities do not solely cover cocoa, but also other agricultural commodities, their attention can be expected to shift easily to other crops, which is currently the case.
Moreover, land area and yields in most Sulawesi regions tend to decline along with the implementation of the projects. It is hard to see how the projects could successfully intervene in this process. One can conclude that the rather managerial approach by the Northern actors, quintessentially based on the assumption that the economic problem of low productivity and low quality can be solved through training, is not sufficient to effectively handle the issue at stake. Therefore, it neglects too much that the persistent problem of land use change relates to other problems underlying the tensions in the governance dilemmas, such as the short-term economic interests of the farmers, their weak bargaining position in the economic process, the lack of commitment of government agencies, and the institutionalized social relationships of the farmers.
The analyses based on our diagnostic framework seem to be policy relevant as they may help to localize and better understand the persistent problems regarding externally initiated partnership projects in collaboration with Southern governments and local producers.
An important issue for further research is to look for opportunities for improving the role of extension officers as intermediaries between the global projects, the national agricultural policies and the farmers, particularly regarding sustainability issues. Our research indicates a change from traditional public extension work to private extension. This brings up the question on how to assess and evaluate this change in terms of its implications for the traditional public role in extension work and the process toward a more sustainable production of agricultural commodities.
The continuation of the researched projects and alternatives for such types of partnership projects refer to other important issues for further research. Mid-2015 Swisscontact signed a contract with MCA-I (Millennium Challenge Account for Indonesia) for additional funding for another 3 years (2015-2018) to reach more farmers and give them better access to finance. Such a project funding seems to be rather easily available. However, we also observe the existence of rather new, bottom-up Indonesian partnership initiatives, particularly between producers, research institutes, local governments and traders that aim to improve the economic and social welfare of Indonesian farmers (Mawardi et al. 2006;Soemarmo et al. 2009;Virgiano 2012). It seems paradoxical that these initiatives to better connect farmers to the global value chain have to develop without large-scale funding. We suggest that a comparison between the governance challenges of the two types of partnership projects can further improve our knowledge about the pathways toward a more sustainable cocoa farming.