Partnerships Blending Institutional Logics for Inclusive Global and Regional Food Value Chains in Ghana; with What Smallholder Effect?

We witness a promotion of hybrid partnerships, where actors with different competences and resources collaborate for smallholder inclusive value chain development. To better understand the functioning of these partnerships, we used institutional theory and studied the context of a global and emerging regional food value chains in Ghana, the blending of logics by key actors in Innovation Platforms and Public Private Partnerships, and their effect on value chain relations of smallholder farmers. In the global value chain of cocoa, partnerships adhered to ‘green revolution’ and ‘free-market’ logics, and provided all farmers material support. In the more informally organised regional food sector, local executing partners selectively coupled their logics with those of poor smallholders, who rely on low-input agriculture and solidarity logics to make ends meet. This improved the position and transaction costs of smallholders to participate in the value chain. Hence, it is more likely for partnerships to create smallholder inclusive governance in informally organised regional food value chains, than highly structured global value chains controlled by international buyers. To gain insight in the variety of political effects this triggers in different social–historical shaped farmer communities, households and actors, we recommend complementary local research from a critical institutional perspective.


Introduction
Since the turn of the twenty-first century, a growing number of international development organisations, donors, and scholars promote hybrid partnerships for inclusive development (Pansera and Owen 2018;Heeks et al. 2020). The failure of the science-based technology development and economic growth model to eradicate poverty and inequality in a globalised world, organised around value chains, led to the emergence of the concept of Inclusive Innovation and Development (Gereffi and Lee 2012;OECD 2013;Gupta et al. 2015). There is a special concern for the rural and agrarian population in Sub-Saharan Africa and Southern Asia, as they mostly live below the poverty line of $1.9/day (UN 2019). International development organisations increasingly acknowledge that complex development issues cannot be solved by a single discipline or actor (Austin 2000;Dunning 2006), so when the World Bank funded Forum on Inclusive Development at Beijing 2012 discussed Inclusive Development, they emphasised the role of partnerships (Chataway et al. 2014). As a result, we witness a growth in Multi-stakeholder Innovation Platforms (IPs) and Public-Private-Partnerships (PPPs), where actors with different interpretations of inclusive development collaborate (Pansera and Owen 2018;Heeks et al. 2020). However, there is little knowledge of the contexts, management and impact of such hybrid collaborations (Vurro et al. 2010;Heeks et al. 2020).
This article studies the context and institutional entrepreneurship of actors initiating IPs and PPPs for smallholder inclusive development of agri-food value chains in Ghana. Most companies and development actors work on agri-food value chain upgrading: improvement of quality or quantity of smallholder production, processing or marketing to attain adequate revenue. To attain the upgrading, they focus on the improvement of technologies, service, training and input supply to stimulate productivity, income and capital accumulation of farmers. 'Inclusive value chain' research and development actors typically focus on smallholder farmers as a subset of the poor, characterised by different degrees of marginalisation (Ros-Tonen et al. 2019). Whilst engaging in upgrading activities they display a high concern for smallholder livelihoods and advocacy (Franz et al. 2014). To attend to both concerns donors promote hybrid partnerships, notably PPPs of companies and NGO's, sometimes including research organizations (IFAD 2016). Researchers, on their turn, increasingly acknowledge the limited impact of the top-down science-led technical-economic development and dissemination approach, and embrace transdisciplinary research, initiating multistakeholder deliberation in IPs to create technical, organisational and institutional change (Klerkx et al. 2012). International donors question the impact and sustainability of informal organised IPs (Schut et al. 2016;van Ewijk and Ros-Tonen 2021), and prefer contractual arranged PPPs, to leverage private investment, secure technology and know-how, whilst utilising other actors with social and political capital to ensure social justice (IFAD 2016).
So far, many studies on multi-actor collaboration have focused on the microlevel communication and collaboration, or they assessed the impact of hybrid partnerships, but their findings are inconclusive as impact differed highly per context (Vurro et al. 2010). To formulate policy recommendations, it is critical to understand the conditions and drive of actors that collaborate for inclusive value chain development. The context, structural design of hybrid initiatives, and their management define the development impact (Heeks et al. 2020). In this article, we study the context and the subsequent shaping of hybrid institutional logics by an IP and PPP in a highly structured global cocoa value chain, and more informal relationshipbased regional food value chains such as cassava and soy beans in Ghana, and assess the effect this had on the relationships and transaction costs for poor farmers participating in the respective value chains.

Theoretical Framework
Institutional theory looks at institutions, as taken-for-granted rules, values and beliefs that guide the behaviour and interaction of actors within an organisational field (DiMaggio and Powel 1983;Fuenfschilling and Truffer 2014). It investigates the embeddedness of individual and organisational actors and their management by deep-structural rules of a broader society; how institutions shape and are shaped by actors' rationality and behaviour. Institutional logics concern the content and meaning of formal and informal institutions, affecting and being affected by action of individual or organisational actors (Thornton and Ocasio 1999).
In this article we take an institutional value chain perspective. Since the 1970s the flow of goods, services and capital have become more globally organised in highly competitive markets with multinational enterprises, creating systems of governance linking firms together in a variety of sourcing and contracting (Gereffi et al. 2001). Basically, a value chain is "the full range of activities, which are required to bring products or services from conception, through different phases of production, delivery to final consumers" (Kaplinksky and Morris 2001). In line with Pouw et al. (2019) we take a broader value chain perspective, including research, knowledge exchange and capacity building, agricultural inputs and financial services, government policies, as activities that support and influence value chain functioning. We look at value chains as bundles of institutions, guiding the exchange of good and services by actors, and study the logics, relationships and transaction costs of actors that participate (Luiz et al. 2019).
Critical studies on agricultural commodification have highlighted risks of smallholder value chain integration, as there are numerous accounts of 'adverse incorporation' due to low bargaining power, limited transparency, heavy input cost coupled with high price volatility and delayed payment, indebtedness, reduced freedom of choice in the allocation of labour and land within farmer communities and in households undermining food security and gender equity (German et al. 2020). Structural adjustment reforms forced governments to reduce agricultural services and liberalise markets, making smallholders more vulnerable in value chain contexts (Ibid). In the agrifood domain, we note two emerging phenomena. The rise of global supermarkets, international quality and fair-trade brands give a small number buyers in global value chains the control in highly formal structured trade and production relations with small suppliers (Gereffi and Lee 2012). On the other hand emerging economies in the south provide opportunities for local traders and manufacturers to create regional markets for processed food, in contexts where interaction and knowledge sharing is more based on trust and social relationships (Gereffi and Lee 2012;Luiz et al. 2019). Question is whether and how partnerships for inclusive value chain development are able to incite value chain actors in these different institutional contexts to adequately blend 'free market' logics with concern for social relations, smallholder empowerment, solidarity and social security for the poor, thus mitigating risks for adverse smallholder incorporation.
In development contexts, organisations and partnerships have to deal with multiple logics, which leads to conflict, co-existence or blending of logics (Luiz et al. 2019;Heeks et al. 2020). Powerful actors engage in institutional entrepreneurship, leveraging resources to create new institutional arrangements that they esteem appropriate and legitimate in a context, mobilising partners to align their logics and practices (Garud et al. 2007). To attain smallholder inclusive development, IPs engage in networking and flexible, informal multi-stakeholder dialogue to identify key problems, explore and prioritise solutions for coordinated action (van Paassen et al. 2013). In PPPs, businesses and NGOs collaborate on the basis of commitments that are formalised to a certain extent, utilising complementary resources and competences for market-based approaches of development (Bitzer and Glasbergen 2015). As collaborative arrangement, partnership organisations blend multiple institutional logics. How this blending is shaped; whether and how the partnership is able to act as institutional entrepreneurs creating new inclusive institutional arrangements is not yet well understood. Emerging literature on hybrid organisations highlight patterns such as (Heeks et al. 2020;Mair et al. 2015;Pache and Santos 2013): • selective coupling selectively prioritising the values and practices of one logic more than the values and practices of another logic. • innovation developing new governance practices to support both logics, e.g.
through new type of human resource management. • decoupling symbolic endorsement of one logic whilst operational practices in almost the whole organisation are linked to another logic.
When partnerships aim to create inclusive value chain governance, this means they need to create social structures that connect companies and traders with a formal rulebased market logic with smallholder farmers who mainly adhere to informal relationship-based logics, in such a way that it mitigates market exclusion, information asymmetry and livelihood risks (Luiz et al. 2019). This poses different challenges for the creation and functioning of partnerships intervening in highly formal structured global value chains with powerful (inter-) national actors, than those intervening in more informal relationship-based regional value chains. We want to understand how the institutional logics of partnerships that aim at inclusive agro-food value chain development are shaped during the formulation and enactment in the context of (a) a highly formal structured global value chain, with high stakes of powerful (inter-) national actors, and (b) merely informal structured regional food value chains in Ghana.
To assess the effect of the enacted logics by partnerships for smallholder farmers, institutional analysts study the change in value chain relationships and transaction costs (costs made and benefits received) for participation (Williamson 1985;Luiz et al. 2019).

Research Method
This study is part of NWO funded research project 'partnership arrangements as strategic action for inclusive development' implemented by the Wageningen University and CSIR-Science and Technology Policy Research Institute in Ghana. We used a qualitative multi-case approach (Yin 2009) to study and compare the context, initiation and implementation process of partnerships. Amongst 25 partnerships project functioning or recently finished in agri-food value chains in Ghana, we searched for an IP and PPP operating in a same global value chain, and an IP and PPP in one emerging regional value chain. We found an IP as well as PPP in the global cocoa value chain, but for the regional value chain we had no choice than to focus on different crops: cassava and soy beans, respectively (Table 1). Literature review and analysis of project action plans, progress-and evaluation reports provided us first insights in the value chain context and partnership dynamics, after which in-depth interviews were done in two phases. In the first phase, January 2015 to December 2016, in-depth interviews were executed with relevant donors, programme managers, local NGO managers, and other actors (Table 2), to trace the partnership initiation and implementation process, missions, negotiation and enacted rationalities. All these interviews were transcribed and analysed through iterative coding.
To be able to characterise the blending of institutional logics in partnerships, we formulated typical institutional logics, relevant to study smallholder inclusive value chain governance in the domain of agri-food value chain governance in Ghana (Fuenfschilling and Truffer 2014). We first identified the intrinsic logic of multistakeholder platforms versus Public Private Partnership collaboration. Within conventional upgrading of agri-food value chains, the 'green revolution' and 'free market' logics are dominant, whilst we are interested to see to what extent core logics related to 'smallholder inclusive value chain governance' are enacted by the partnerships. Based on literature, we distinguished the following social and relational dimensions of inclusive agro-food value chain governance (Kilelu et al. 2017;Ros-Tonen et al. 2019;Vellema et al. 2020), which we translated in reciprocity and solidaritycore logics as presented in Table 3.

Relational:
• Smallholders empowerment by strengthening negotiation capacity of smallholders versus companies and other value chains actors. • Addressing constraints to participation and better cost-benefit balance of marginalised (poorest households, women and migrants), e.g. by exclusion and information asymmetry.

Social
• Concern for poverty alleviation/material well-being: right for a relatively stable and decent living and food for all. • Democratic deliberation for knowledge creation, decision-making, monitoring and evaluation, recognising indigenous knowledge. Table 2 Overview of interviews conducted with actors involved in the four partnerships

Category of interviewees Number
Public sector actors-involved in regulating and/or coordinating value chains 2 Private sector actors-involved in large scale buying and processing of cocoa or soybean 7

Farmer-based organisation leaders/members 6
Local NGOs-involved in implementing value chain interventions in rural areas 15 Focal persons-involved in managing projects as lead contact persons 4 Officers of International donors-involved in funding partnership projects 10 • Sensitivity for the diversity amongst smallholder farmers, especially the poor and marginalised with mixed livelihood logics mitigating risks and relying on relations of reciprocity and solidarity.
From February to March 2018 an impact study was executed in the different intervention areas: In consultation with the partnership officers, a representative village was selected from each of the intervention areas. For CORIP, an EMFED village

Green revolution Values Strict adherence to Green Revolution packages of inputs and agricultural practices
Assumptions Science-developed packages of high-yield hybrid seeds, synthetic fertiliser, pesticides and irrigation increases the efficiency of agricultural processes and productivity per hectare

Smallholder empowerment
Values Societal justice and equity requires empowerment of the smallholder to capture a decent share of the value created in a value chain Assumptions Awareness raising, capacity building and organisation enable smallholder farmers to negotiate a fair share of the value created in a value chain (Gender) Equity Values All humans are of equal value and should be free to promote and achieve the plural functioning that they aspire Assumptions Awareness raising of constraining formal and informal institutions and power imbalance; capability development of the marginalised and inclusive action of others could free marginalised humans, enabling them to better articulate and achieve plural functioning that they value Social security/poverty alleviation Values All humans have the right to live a decent living and be food secure Assumptions In a society there is diversity of wealth, but the state and societal actors feel the obligation to ensure a minimum level of food, income and labour conditions for all

Reciprocity and solidarity
Values Within a community, social actors collaborate in reciprocal relationships and express solidarity and care in times of need Assumptions In a community there is social capital: people bond and exchange services and goods for the public and personal benefit, and take care of the resource-poor in case of need was selected as this implementation partner seemed most beneficial for resourcepoor smallholders. In these four selected villages, 24 focus group discussions, 16 indepth interviews and 145 surveys were executed (refer to Table 4) on topics such as resource endowment, aspirations, project participation, process and services gained, impact on knowledge, materials, finance, network and organisation, confidence, level of satisfaction related to aspirations. Focus group discussions and interviews were transcribed and analysed through iterative coding. Surveys were analysed with excel.

The Organisational Field of the Global-and Regional Value Chains in Ghana
In the recent decades, Ghana attained sustained economic growth and is lauded for political pluralism (Resnick 2016). Downside is that politicians display a high interest for urban constituencies and remunerative export sectors but less for the rural food sector (Kolavalli et al. 2012 Savelli et al. 2018). As farmer capacity building for agricultural production and market organisation proved slow and costly, they increasingly focussed on the support of private actors to create viable market structures. The green revolution logics remained important, but free market logics became more prominent at the expense of poverty alleviation and smallholder empowerment logics. In the domestic food sector donors increasingly supported dynamic local entrepreneurs to invest in processing facilities and create reliable market structures, whilst in the export sector they aligned with international companies of their home countries to entice the latter to engage in sustainable sourcing (Kelly et al. 2015;Woodhill 2016). As a result, around 2010, the dominant actors and institutional logics of the cocoa value chain differed significantly from domestic food value chains in Ghana.

The Global Value Chain of Cocoa
Since the mid-1920s Ghana is a main producer of cocoa. At independence, the Ghana government took over the colonial Cocoa Marketing board, COCOBOD, and became the monopoly buyer, whilst also ensuring farmer services and quality control (Amoah 1998). Restructuring programmes however forced COCOBOD in 1992 to liberalise the marketing through the introduction of private Licensed Buying Companies (LBCs). LBCs received a buyers' margin [almost 8% of net free-on-board (f.o.b.)], and purchasing clerks worked on commission basis. COCOBOD remained responsible for quality control, price setting and export marketing, stimulating cocoa production through free pesticide spraying and subsidised seed fertilisers, as the export trade margin and taxes provided essential government revenues (Laven 2010;Essegbey and Ofori-Gyamfi 2012). As LBCs were paid per volume, the quality of the delivered cocoa decreased till 2005, when COCOBOD decided to half the price for bags 'with more than 25% poorly fermented beans' (Vellema et al. 2016). Table 4 Impact interviews with young, female, and male farmers plus processors (mainly females) Focus group discussions (24) In-depth interviews (16) Surveys ( From an institutional perspective, the cocoa sector in Ghana is highly formally structured through the state-owned marketing board. The prevailing logic of COCOBOD is to participate in the global free trade and attain profit through high volume, good quality cocoa production and trade. They aim to stimulate and control sustainable supply through yearly reviews of cocoa prices, fixed LBCand farmer margins. Third priority, partly political motivated but also to ensure sustainable cocoa supply, is to show care for the rural constituency through free seed and subsidised fertiliser campaigns (Houssou et al. 2017).
In 2008 about 800.000 smallholders with farm sizes of 0.4 to 4 ha attained 70 to 100% of their household income from cocoa (Baah and Garforth 2008) but they were not motivated to invest in cocoa production due to inadequate advisory services, relative low cocoa farm prices (76% of net F.o.b. price in 2012), unreliable delivery of (subsidised) agro-inputs, declining soil fertility, and ageing cocoa trees (Adu-Acheampong et al. 2017). This led to low average productivity levels of 263 kg/ha in 2012, whilst some farmers were able to produce 1400 kg/ha (Ingram et al. 2018). In 2014/2015, about 45% of cocoa farming households lived below the World Bank poverty line (1.90 USD per capita), 35% between a decent living (around 2.20 USD per capita in Ghana) and the poverty line, and 20% above the decent living line (Waarts et al. 2019). Despite the unreliable input delivery, most farmers preferred to wait for the subsidised assistance rather than to organise themselves. Weighing opportunities and risks, farmers demonstrate different farm logics. Amongst well-endowed farmers, some embrace the 'free market logic' and readily obtain credit for high productivity cocoa investments, whilst others value their autonomy and social reciprocity logic. The latter refrain from credit relations and invest in diverse crops, using relationships for mutual help to maximise the labour-and cash flow, minimising risks (Jaskiewicz and Laven 2016;Manley and Leynseele 2019). Amongst the resource-constrained, mostly young or female farmers who have little farm investments and highly depend on family labour, some take input credits whilst others opt for social reciprocity and subsistence farming (Ibid).
Multi-National Companies (MNCs) source cocoa from Ghana to process in the Netherlands (33.8%), United Kingdom (12.1%) or Belgium (8.8%) (Essegbey and Ofori-Gyamfi 2012). Embracing the free market logic, they focus on 'sustainable sourcing' of cocoa (Vellema et al. 2016). As they cannot compete on price in Ghana, they try to build direct relations with cocoa farming communities through voluntary certification schemes (Laven and Boomsma 2012) and/or Public-Private Partnerships (PPPs), where lead companies collaborate with donors, NGOs and governments. Certification schemes provide companies price premiums, used to cover the professional training, quality monitoring and farmer bonusses, whilst PPP arrangements help them to mobilise complementary competences and resources to ensure sustainable sourcing. These strategies enable international companies to show commitment to global sustainability standards and keep a 'licence to produce' in consumer markets.

The Regional Food Value Chain
In the last three decades, the government did not invest in the structurally development of the domestic food sector (Resnick 2016). The Ministry of Food and Agriculture (MoFA) lacked pro-active policies for the development of an food-processing industry, the promotion of diversification (Kolavalli et al. 2012), and capacity building of Farmer Based Organisations (FBOs)s at all levels (Senadza and Laryea 2012). When world prices rocketed and food was scarce in 2008, the Ghanaian government launched a national fertiliser subsidy programme. Through time, the political discourse concerning the target of the subsidy programme alternated between 'agricultural development' and 'the creation of a food safety-net for the resource-poor' (Resnick and Mather 2016). In practice, delivery was often too late, and negatively impacted private sector and farmer business initiatives (Ibid). Due to the fluctuating input supply and market demand, and highly variable food prices there was limited farm investment, and investments done mainly consisted of labour-saving mechanisation, acquisition of fertiliser and herbicides rather than sustainable intensification (Houssou et al, 2016). Around the 2010s, commercial food crop production was predominantly done by resource-poor smallholder farmers who mainly sold their produce to local markets or aggregators with whom they had a loan. Cultivation was characterized by low-input and average yields were well below attainable levels (Ibid).
There is some diversity but the main logics of these farmers concern social reciprocity and solidarity for food security. The use of arable land is allocated by the male head of households but all members help each other out (Vercillo 2020). It is the responsibility of household head to secure household food staples, but if they do not manage spouses add from their personal stock (ibid). All members buy inputs for food crop production they intend to sell to pay small daily expenses, decent housing and the children's education. However, the balance is delicate as Vercillo et al. (2020, p. 565) note: 'various smallholders feel adherence to the green revolution is a short-term trade-off to meet subsistence needs at the expense of soil health'.

Blending of Logics
In the cocoa sector, we studied a research-driven IP initiative as well as a donor driven PPP project. The IP was part of a larger Convergence of Sciences-Strengthening Innovation Systems (CoS-SIS) project in West Africa. After research into farmer-led innovations at local level, CoS concluded that smallholder farmers faced small windows of opportunity due to constraining formal and informal institutional arrangements (van Huis et al. 2007). In the second phase CoS-SIS therefore aimed to combine participatory research at farm level with IPs of key actors, able to tackle constraints at institutional level (Hounkounou et al. 2012). Senior CoS researchers persuaded the Dutch Ministry for International Relations & Development to fund this follow-up research, and they contacted the respective National Agricultural Research Institutes (NARIs) in Mali, Ghana and Benin to select the domains of intervention in line with national development priorities. Ghana's cocoa sector was amongst the selected domains (Ibid).
The action research on institutional change for smallholder development started with scoping studies, discussed in multi-stakeholder workshops. These workshops defined the IP priorities and key actors to be involved. At this stage, it became clear that key actors at national level were concerned about young farmers not willing to invest in the cocoa production and acknowledged the inefficiencies of existing institutional arrangements for pesticides-and input supply (Adu-Acheampong et al. 2017). The programme recruited an IP facilitator from the trusted Cocoa Research Institute Ghana (GRIG), linked to COCOBOD to mobilise open-minded IP members amongst farmer representatives, COCOBOD and the Ministry (van Paassen et al. 2013). Together they agreed to study the cocoa price setting: the margin paid to farmers and the margin set to cover the public support programme. Investigations by Cargill of price setting in Cameroon and Ivory Coast enabled the IP to persuade the minister in 2010 to announce a 33% farmer price increase (to GH¢200 per bag of 64 kg), a sum considerably higher than the one proposed by the Producer Price Review Committee (GH¢180 per bag) (Adu-Acheampong et al. 2017, p. 4). Furthermore, a study was done on the technical and organisational efficiency of the pesticide spraying gangs and free fertiliser distribution. This led to delicate discussions, announcements of subsidised input deliveries in national newspaper to enhance transparency and curb corruption, and the provision to gradually reduce the public spraying and input supply to allow for a more readily supply of inputs on a commercial basis (Ibid: 5).
In sum, the leading researchers of CoS-SIS demonstrated institutional entrepreneurship: driven by the logics of poverty alleviation and multi-stakeholder innovation, they mobilised actors to create institutional arrangements, conducive to smallholder development. In the context of the Ghanaian cocoa sector, the farmer representation was weakly organised and the relevant institutional change had to come from the COCOBOD and government authorities, who prioritised the 'green revolution' and 'free market' logics'.

Changing Value Chain Relations and Transaction Costs of Smallholders
When assessing the change in relations and transaction costs of smallholders, created by the CoS-SIS IP, the smallholders provided the following feedback: • Relational-wise, the CoS-SIS leaders aimed to increase the development opportunity of the smallholders, but as farmer organisation in Ghana was weak, the mobilisation of FBOs for advocacy of the resource-poor did not materialise. Smallholders did not notice any change in information provision and transparency of the government input supply, as they read no newspapers nor were able to pay officers and spraying gangs for access to the subsidised support. The effect of the envisaged IP privatisation activities was that the resource-poor could rely less on public support and depended more on own or external resources provided under a 'free market' logic. For the poor this may easily lead to a negative lockin (Laven and Boomsma 2012), hence many rather limit their capital investments in agriculture ('free market' and green revolution' logic), opting for the 'social reciprocity and solidarity' logic. • At the level of Transaction costs for value chain participation, respondents noted they benefitted from the increased farmer price, but the money was spent 'when farm investments were needed'. In Ghana, cocoa is seen as the most profitable crop, and respondents aspired improved cocoa productivity to raise income and accumulate farm resources. However for the poor involvement in commercial credit for fertiliser and pesticides is risky. With lower soil fertility and increased uncertainty of rainfall, many resource-poor households get indebted.

Blending of Logics
As home country of various cocoa MNCs, the Embassy of the Netherlands was interested to allocate substantial development funds to cocoa production. In dialogue with various MNC's, they commissioned a technical-economic assessment of Ghana's cocoa sector in 2012 and learned that the cocoa producers, even the certified producers, could improve the productivity with 25-50% with 'timely and appropriate supply of agro-inputs'. The embassy imagined that Rural Service Centres (RSCs) were the missing institutional arrangement, and driven by the 'Aid for Trade' and PPP logic they contacted NGO Solidaridad to elaborate a concrete project proposal. In the 1980s, Solidaridad promoted Fair Trade products in western consumer markets but later they shifted focus to the development of fair local production structures. In Ghana they were a respected development partner. In line with the PPP logic, the Cocoa Rehabilitation and Intensification Programme (CORIP 2013(CORIP -2017 promised 50% funding and technical support for RSCs created by LBC-MNC ventures. To align with the government, the CORIP involved the Ministry of Finance & Economic Planning and COCOBOD in the consortium. Together with the International Fertiliser Development Centre (IFDC), CRIG provided technical training to the RSC managers. In this way the Dutch Embassy mobilised private companies to create new institutional arrangements for smallholder input supply and training, independent of the public services. The prime logic pursued was 'free market', but the collaboration with Solidaridad demonstrated also commitment to 'social justice'. In 2014, seven LBC-MNCs ventures acquired CORIP funds and created about 20 RSCs with different arrangements. Some ventures constructed local shops with demonstration plots, whilst others opted for mobile service centres. A Ghanaian venture EMFED/Abrabopa employed youngsters to offer smallholder farmers a gamut of farm management services, to be paid back through a part of their harvest. Apart from input supply, many farmers requested assistance for weeding, pruning, fertility and pest management, harvesting, pod breaking and drying. This served the elderly as well as absentee farmers, with small inherited plots. A monitoring system was set up to enhance LBC-MNC knowledge sharing about farmer needs and effective business models. Depending on their interpretation of Corporate Social Responsibility(CSR), LBC-MNC ventures developed different service arrangements, blending 'free market' and'green revolution' with 'social justice' logics, e.g. providing nearby access to inputs and training, integrating inputs for other crops to support food security.

Changing Value Chain Relations and Transaction Costs of Smallholders
As the operationalisation of the RSCs differed per LBC-MNC, the change of value chain relationships and transaction costs of smallholders also differed per arrangement: • Relational All RSCs ensured more timely, nearby access to cocoa inputs, and technical training to lead farmers. Many resource-endowed farmers invested in the relevant inputs, but the resource-poor were hesitant and preferred to wait for the subsidised inputs. Access to RSC for the resource-poor improved when linked to a certification scheme, offering more intensive knowledge exchange, provision of collective spraying, pruning equipment, and price premiums. • Transaction costs for value chain participation Though inputs and information provision was more timely and nearby, transaction costs for participating in the cocoa value chain mainly improved for the resource-endowed. Depending on the engagement of LBC-MNCs in certification schemes and their interpretation of CSR, more resource-poor households benefitted of information and production support. The EMFED farm management services especially served the elderly, ill, females and absentee farmers, who now did not need to arrange labour and received their share of the harvest income.

Blending of Logics
To increase the uptake of the green revolution technologies and attain higher productivity in the food sector of Africa, the Forum for Agricultural Research in Africa (FARA) together with the West and Central African Council for Agricultural Research and Development (CORAF/WECARD) elaborated the 'Dissemination of New Agricultural Technologies in Africa' (DONATA, 2011(DONATA, -2014 project. The uptake of the technologies should be attained through 'multi-stakeholder processes using an innovation platform as key tool for participatory and collective action, facilitating livelihood and/or value chain development'. They negotiated funding from the African Development Bank (AfDB), which promotes 'inclusive growth and the transition to green growth', and contacted the NARIs for implementation. In Ghana, CSIR-CRI was appointed to lead the DONATA project for cassava, and they contacted the District Agricultural Office (DAO) in the Wenchi municipality of Brong Ahafo region to implement this new innovation approach. The DAO adhered to the 'green revolution' logic and appreciated the 'multi-stakeholder innovation' logic but resource availability was such that their advisory practices rather aligned with the 'transfer of technology' logic. However, the DONATA funds enabled them to engage with the 'multi-stakeholder innovation' logics, and they selected 5 farmer groups with commercial cassava production, eager to engage in multi-actor learning. IP membership was flexible, starting with 10 to 30 farmers, including women who also processed cassava into gari. The deliberations started with a joint problem identification and a selective choice of 'green revolution practices' for-farm experiments. Farmers tried highyielding cassava varieties with good grinding quality, new planting and weeding practices to reduce chemical and labour needs. When productivity increased, the focus shifted to transport, processing and marketing issues; hence transporters and traders joined the IP. In dialogue, the local cassava transport was optimised to serve farmer needs and with the support of the DONATA funds cassava processing cooperatives were established.

Changing Value Chain Relations and Transaction Costs of Smallholders
The DAO created a local innovation arrangement based on 'multi-stakeholder innovation' logics for selective use of 'green revolution' practices, actually resembled an 'smallholder empowerment' logic (Osei-Amponsah et al. 2018): • Relational The most active cassava IPs comprised migrants from the north, who did not have land property rights and borrowed land from the autochthone population. Autochthone farmers rather invested in lucrative tree crop farming. Tree ownership implies long term access to land and therewith create rights on ownership; hence migrants were not supposed to plant trees on borrowed land and specialised in seasonal food crops. To attain support for their agricultural and processing activities, migrant community organised themselves in groups to attract project support, like the one provided by the IP project. The IP activities benefited male farmers, but especially the females who acquired an agrobusiness training and started a processing unit. They developed a business network, and negotiated cassava marketing with their spouses, other cassava farmers and transporters. Formally registered as a group, the women managed the cassava processing unit. They strengthened their position in the local cassava market, but were not yet linked to larger distance consumer markets. • Transaction costs for value chain participation The IP enabled the involved migrant community to obtain (a) high productive varieties and input-saving farm practices, (b) well-organised transport arrangements, (c) business training and equipment for cassava processing. Production doubled for all participants and the women acquired lucrative processing and marketing opportunities. Within the processing unit they organised a mutual help fund.

Blending of Logics
2SCALE is a follow-up of the Africa-wide 1000+ project (2006)(2007)(2008)(2009)(2010), led by the NGOs IFDC and ICRA. When the impact of the IFDC Integrated Soil Fertility Management (ISFM) project (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005) was offset by falling product prices, 1000+ adopted a Competitive Agricultural Systems and Enterprises (CASE) approach. IFDC aimed to promote a 'green revolution' plus a 'free market' logic, but ICRA who adhered to a 'multi-stakeholder innovation' and 'smallholder empowerment' logic convinced them to focus on FBO capacity building, strengthening the organisation, processing and marketing capacity. The project blended 'empowerment' with the 'green revolution' and 'free market' logic. At the end of 1000+, IFDC contacted the Dutch Embassy, who was in the process of a policy shift from 'Aid' to 'Aid for Trade' (Ministry of Foreign Affairs of the Netherlands 2013). Prioritising 'free market' logics, the embassy aimed to build local market structures and competitive business management in the food sector. The economic evaluation of 1000+, commissioned by the Embassy, recommended improved project efficiency through scaling. As the Embassy promoted PPPs, IFDC and ICRA wrote a 2SCALE proposal (2012-2017), integrating 'PPP' and 'free market' logics. In Ghana, IFDC arranged 16 clusters with contracts between traders, financial institutions, service providers and secondary FBOs, whilst local NGOs continued their FBO capacity building activities. BoPInc became also involved as partner, executing market studies for Bottom-of-the-Pyramid (BoP) businesses: 'commercially and socially viable business models, which include the people in the BoP as consumers, producers and entrepreneurs in the supply chain.' In northern Ghana, three market clusters focused on soy beans. There was a large processing capacity, but weak market structures constrained farm production. In 2SCALE, secondary level FBOs negotiated contracts with input suppliers, tractors services and soy bean traders, enabling primary level farmer groups to attain group credit and the demanded input-and ploughing services. Local NGOs EPDRA and SEND provided training in nutrition, low-input agriculture, business-and FBO management, whilst coaching primary level farmer groups in the organisation of the services and credit & saving arrangements. There was leniency for the poor, allowing them to use services in a way they deemed fit for their circumstances, as long as they honoured the group loan repayment. When BoPInc. identified a soybean kebab market at nearby schools, the NGOs trained women in kebab production and marketing. Apart from this NGOs had their own agenda, e.g. SEND executed a gender programme, stimulating men to help their spouses with part of the farm and household chorus.
Looking at the evolution of the project activities, the Dutch embassy induced the 1000+ project owners to add PPP contracts for cluster markets arrangements (market logic) to their ongoing work of FBO capacity building ('smallholder empowerment' with selective coupling of 'green revolution' and 'free market' logic). The inclusion of local faith-based and gender-sensitive NGO's for FBO capacity building ensured concern for justice and smallholder reciprocity & solidarity logics.

Changing Value Chain Relations and Transaction Costs of Smallholders
2SCALE induced the following smallholder value chain changes: • Relational The cluster approach created formal relations between regional companies and emerging FBOs, whilst the FBO capacity building aimed to strengthen the FBO functioning and bargaining power. As trust had to be strengthened between the primary and secondary level cooperatives, the NGOs focused on (a) a strong cooperative administration, negotiating favourable prices and delivery contracts with large soy bean traders and processors, and (b) coaching of cooperative members on the clarity and leniency of rules. Resource-poor were allowed to use inputs and services as deemed fit, but soybean quality and loan repayment rules were strict. This served many smallholders, but was too strict for and therewith excluded the poorest. The gender awareness programme first created annoyances, but in due time men began to acknowledge the workload of women and started to help out. • Transaction costs for value chain participation 2SCALE worked with existing faith-based and gender-sensitive NGOs, who aimed to support the smallholder farmers. Hence, the resource-poor (50/50 male and female) got training on lowinput agriculture, increased soy yields, and according to smallholder respondents the increased income helped to build food stocks, expand processing activities or buy livestock. The cluster-FBO approach enabled the farmers (a) to gain timely and easy access to inputs and tractor services (critical for ploughing of compact soils), and (b) attain good prices, sell in bulk, and save money at the cooperative credit union for business investments or housing, or payment of school fees, hospital bills and other necessities. For those who could comply with the strict quality and repayments, transaction costs for relevant agricultural information and services, transport, marketing and cash flow management were considerably reduced.

Analysis and Discussion
In this paper we applied institutional theory to study (a) the context and the blending of logics at the initiation and implementation of partnerships, as well as (b) the change in value chain relationships and transaction costs incurred by smallholders. We compared the management and outcome of IPs and PPPs in a highly structured global cocoa value chain and more informal regional food value chains in Ghana, and came to the analysis as summarised in Table 5. MNCs and governments have high stakes and structuring roles in global agrofood value chains. In our case study, they were the key actors for partnerships, advancing 'green revolution' and 'free market' logics in the cocoa export sector. The IP initiative of the researchers, aimed at empowerment and social welfare for smallholders was selectively coupled, or even decoupled and replaced by COCOBOD members pursuing a 'green revolution' and 'free market' logic. The bilateral donor, initiating the CORIP PPP, already aligned with MNCs at the formulation phase so that there was little tension and change in the implementation. LBC-MNCs could pursue their own combination of free market and social welfare logics. EMFED as new local entrepreneur came with a straightforward innovation: an innovative business model benefitting the poor. The partnerships did not improve the relational position of smallholders. They slightly improved the socio-economic situation of smallholder farmers, especially those of the more resource-endowed (Apart from EMFED, with its innovative approach). As 80% of the cocoa growing households earned less than a decent income of 2.20 USD per capita in 2014/2015 (Waarts et al. 2019), the IP negotiated price increase benefitted many poor, but further pursuance of input liberalisation increases risks. We have no detailed data of the logic enacted in all LBC-MNC intervention areas, but literature highlights the imbalance of power between the MNCs and NGOs in the rule-setting of certification schemes of global demand-driven agri-food value chains (Bitzer 2012;Bitzer and Glasbergen 2015). Most MNCs opt for lenient standards, providing training and inputs for ecological and social sound practices, but with less attention for FBO capacity building (Bitzer 2012;Nelson and Tallontire 2014;Ingram et al. 2018). As certification is mainstreamed, supply regularly outstrips demand, so that price premiums tend to decline (Bitzer 2012). Furthermore, it is not sure whether MNCs keep paying certification costs. Lower prices and higher certification costs may bring farmers in a negative lock-in situation (Laven and Boomsma 2012). The local food sector was devoid of formal policies and market structures, leaving space for low-input agriculture and relation-based transactions supported by informal institutions that underscore intra-community solidarity logics (Venkataraman et al. 2016;Luiz et al. 2019). To create development, partnerships mobilised local NGOs working in agriculture and FBO capacity building, or directly involved farmer groups and smallholder enterprises. Through experience, all these actors were cognisant of farmers' poverty and need for low-input and solidarity logics to cope. Though the prime aim formulated for the IP and PPP intervention were to promote 'green revolution' and 'the creation of formal market structures', both embedded and selectively coupled their development intervention with solidarity logic. Through the IP 'multi-stakeholder deliberation' logics, the DAO engaged in joint sense-making with farmers. Though not intended, the enacted logic aligns with smallholder empowerment. This allows poor farmers to integrate logics in such a way that it mitigated risks of free market participation (Mair et al. 2012;Luiz et al. 2019), e.g. establishing a cassava processing cooperative with a solidarity fund. In line with blending findings of Heeks et al. (2020), we see that in the 2SCALE PPP the 'free market logic prevailed at the higher regional level, whilst at local level NGO's there was space for low-input and solidarity logics within the boundaries set by the 'green revolution' and 'free market rules. As a result, several poor farmers experienced socio-economic improvement, but the most poor were excluded.
Our analysis focussed on value-chain-embedded shaping of logics by partnerships and the effect it had on value-chain-relations and transaction costs by smallholders. This provides a better understanding of the effectiveness of collaborative development approaches, than plain micro-level communication analysis (Vurro et al. 2010;Heeks et al. 2020). It shows the challenge to create partnerships that prioritise and enact logics, linked to critical dimensions of smallholder inclusive value chain management as conceptualised by Ros-Tonen et al. (2019) in the present era where bilateral donors increasingly embrace 'free market' logics stimulating private capital accumulation and competition. All studied partnerships paid tribute to one or more logics of smallholder inclusive value chain governance in the formulation phase, but in the global value chain of cocoa it did not always sustain in the enactment. The partnerships that worked on the creation of regional food value chain displayed more concern for issues such as smallholder empowerment; eye for diversity amongst smallholders and inclusion of the 'reciprocity and solidarity' logic of the marginalised; high concern for a decent living and equity for all. Our results show the effects this triggers on the relations and transaction costs of smallholder households, regarding the participation in the value chain.
The research focussed on the challenges of two different types of partnerships to create more smallholder inclusive value chain logics and relationships in different value chain contexts. This provides relevant insight for policy makers and donors, considering the initiation of partnership in a highly structured value chain, centrally steered by distant value chain actors with no farmer organisation, or an emerging, informally organised regional value chain with some initial farmer organisation. However, we cannot be conclusive of the effect of the partnership interventions for value chain integration had within varying local communities and households. To be conclusive on the issue of inclusion or exploitation of specific households or individuals at local level, our study needs to be complemented by in-depth anthropological research on the ripple effect of such interventions in the variety of socio-historical shaped local communities and households that populate the intervention zone (Bakker and Gill 2019). Critical institutional research zooms in on historical shaped politics of power within a plurality of village communities, cooperatives as well as farmer households (Hall et al. 2014, Mangnus andSchoonhoven-Spreijer 2020). It also adds individual farm labourer perspectives, distinguishing types of work done and relative benefits received from all types of tasks executed for mixed crop farming and social reproduction (Gore and LeBaron 2019). Only with such complementary research, the multi-scalar complexity of institutions entwined in everyday life, the historical shaped dynamics of value chain-, local community and household politics, can be revealed to draw final conclusions on actual improvement or continued exploitation of marginalised actors in rural areas (Ibid).

Conclusion
Objective of this study was to understand the conditions, management and impact of partnerships, that aim to create smallholder inclusive value chain development. We used institutional theory to assess the context, and blending of actor logics at IPs and PPPs and effect on value chain relationships and transaction costs of smallholders in Ghana. IPs have a different principles than PPPs, but both are collaborative arrangements between actors that play a role a respective value chain. The actual composition and pursued intervention logics of the different partnerships highly differed per value chain context. In the global cocoa value chain, partnerships mainly adhered to 'green revolution' and 'free-market' logics, and provided farmers material support to increase productivity. Regional food value chains lacked formal policies and market structures. Here partnerships selectively coupled their logics with those of poor farmers, who rely on low-input agriculture and solidarity logics to make ends meet. The PPP integrated poor farmer logics within the limits set by formal credit contracts. The IPs deliberation logics enables poor smallholders to organise themselves (empowerment), gain material benefits whilst also creating a solidarity fund. Hence partnerships are likely to be more successful in creating inclusive value chain governance in informally organised regional food value chains than highly structured global food value chains. However, though various smallholder respondents expressed their satisfaction with the material gains, enhanced social respect and negotiation position in the regional food value chain, the study cannot be conclusive on the positive effects of partnerships activities, triggered in the plurality of historical shaped village communities and households linked to the food value chains. We therefore recommend complementary multi-scalar critical research from farmer labourer, household and village community perspective.