The HortIMPACT programme : a new approach for sustainable inclusive business in Kenya

(1)

15-08-2016

MSc thesis Human Geography Sietse Bruinsma – 6081398

Supervisor: Dr. Ir. Y.P.B. van Leynseele Second reader: Dr. N.P.C. Beerepoot

The HortIMPACT

programme

(2)

1

Acknowledgements

First and foremost, I would like to thank my supervisor Dr. Yves van Leynseele for his patience and constructive feedback. I also want to thank Ms. Naomi Hoogervorst for guiding us during our 8-week fieldtrip in Nairobi and for arranging interesting meetings with experts concerning the theme of the project: ‘Feeding the City’. Without the generous help of Ms. Jane Kamau from the HortIMPACT programme my research would not have been possible, so a special thank you to Ms. Kamau.

It has been an exciting and at the same time exhausting period the past couple of months. It has been a constant cycle of learning and adjusting. I think the literature chapter has been completely re-written at least three times.

Lastly, I would like to thank my fellow students Bas, Romana, and Anne-Luz for sharing the great experience of living together in the beautiful city of Nairobi, Kenya.

(3)

2

Abstract

This research was dedicated to map out the approach used for a Dutch agricultural development intervention in Kenya called HortIMPACT. Due to policy changes and budget cuts for foreign development from the Dutch government. Development approaches are now mainly focused on partnering with private businesses to leverage resources. Topic of interest in the research is also how such a partnership is being matched considering that in multi-actor collaboration the interests and needs are likely to conflict.

The research design used for this research is a qualitative case study. In which various qualitative research methods were used such as; interviewing, observations, reviewing official documents, questionnaires etc. for the collection of data.

Key findings were that the approach of HortIMPACT had some unique features. For example, they let the private businesses in the PPP take the lead in arranging the extension. By using a

horizontal collaboration on the farm-level, they believe to achieve responsibility and understanding of the farmers’ context and building trust needed for inclusive business. One particular business case is used to give an in-depth understanding how the different interests between the programme, the private business and the farmers were matched and what indicators are used by the HortIMPACT programme. The matching of interests between the stakeholders are according to the collected data properly matched.

(4)

3

1. Introduction

‘Disgrace: budget for development drops to 0.52% of the GDP’ headed ‘de Volkskrant’ one of the most

influential newspapers in the Netherlands in September 2015 (Vos, 2015). With this drop the Netherlands officially does not meet the UN-norm of spending at least 0.7% of the GDP on

development anymore. Budget cuts in development are received with a lot of criticism in the public debate in the Netherlands. Critics are concerned if development is still inclusive for the less fortunate due to these budget cuts (Schaik, 2011). Nowadays development often is done in conjuncture with the private sector that have commercial interests. These collaborations between donors, NGOs and the private sector are known as Public Private Partnerships (Ion et al, 2014).

The Dutch government published their new agenda for foreign aid, trade & investment in 2013. This new agenda is named ‘A World to Gain’ and is still the official policy of the Netherlands for development today. It includes a focus on food security which is relevant for this research because the purpose of this research is to map out a Dutch agricultural intervention programme recently implemented in Kenya. The intervention that is chosen as a starting point in this research is the Kenya Market-led Horticulture Programme (KMHP) or otherwise known as HortIMPACT. For this research the name HortIMPACT will be used when explaining the case study.

HortIMPACT is a Dutch agricultural intervention programme led by SNV (a Dutch development organization) which is fully funded by the Netherlands embassy in Kenya and is therefore directly influenced by the new agenda for foreign aid, trade & investment of the Netherlands. The main goal of the programme is: ‘to contribute to increased food insecurity and increased incomes through the

development of a dynamic and sustainable horticulture sector in Kenya inclusive of small and medium sized farmers’ (AkvoRSR, 2016).

In the new Dutch agenda for development, Kenya is categorized as a ‘transition’ country. This means that the donor-recipient bilateral relationship will be phased down and in the long-term a fully trade bilateral relationship will be established. This vision about aid for trade is not entirely new. The World Bank in 2005 was one of the first with a publication about providing aid to establish trade as a way for sustainable development.

One of the biggest sectors in Kenya is agriculture. Approximately 80% of Kenya’s workforce is employed in agriculture and the sector contributes about 57% of the national income (Republic of Kenya, 2006). 80% of all the farmers in Kenya is considered small-scale who produce both for markets and for subsistence. Making food available to all Kenyans is also one of the priorities set by the Kenyan government. The challenges within this priority include: low local supply, post-harvest losses and poverty. In the face of rapid urbanization, changing demographics and competing claims on available resources in Kenya it is important to better understand the needs to improve food security (Kamau et al, 2011).

The HortIMPACT programme pursues a Public Private Partnership (PPP) approach to address those issues that are in the interest of a sustainable horticulture sector in Kenya to improve food security. The target groups for activities are: organized small- and medium sized entrepreneurial farmers, international Dutch and Kenyan agricultural companies as well as support service providers (AkvoRSR, 2016). Given the target group of the programme and the recently changed Dutch agenda for development makes an interesting case to investigate how a development programme like

HortIMPACT optimizes its chosen approach within both its contextual and policy environment in mind to achieve sustainable development.

Therefore, the main question in this research is:

How does the HortIMPACT programme optimize its approach for sustainable development considering its contextual and developmental policy environment?

(7)

6 Here, the research tries to map out the approach of the HortIMPACT programme and explain what factors are influencing the chosen approach. Considering that the HortIMPACT programme is targeting small and medium-sized farmers in Kenya and requires private sector involvement, the first section of the literature chapter discusses existing knowledge and experiences from the literature about these types of arrangements (approaches) for sustainable development that are relevant for analyzing the HortIMPACT programme.

Because a programme as HortIMPACT involves farmers, the second section is dedicated to conceptualize the contextual importance of the assets and capabilities of farmers and their

orientations. The livelihoods framework and the concept of livelihood strategies will be explained in this section together with constraint in market access to give insight on the needs of participating farmers in the programme.

An intervention programme such as HortIMPACT requires multi-stakeholder collaboration that each have their own interests and demands before entering in the programme. Therefore, the last section of the literature chapter is aimed to conceptualize the role for an NGO as a brokering organization including the matching process of the different interests and needs that is needed in a development programme structured as HortIMPACT. Before entering these three sections, the literature chapter begins with a ‘setting the scene’ to provide for some basic understanding and facts about practicing agriculture in Africa and possible opportunities in linking smallholders to markets for sustainable development.

(8)

7

2. Literature

2.1 Setting the scene: smallholder farmers in Africa

80% Of all the food that is grown in Africa and Asia is produced by small-scale farmers. The definition of small-scale farmers is; those that produce both for subsistence and markets. Their labour is

generally based on families and they have limited acces to land (Ros-Tonen et al, 2015). Although, the amount of acres as a proxy for smallholder production is useful, it is considered a bad indicator for solely defining smallholders, because land sizes are highly context specific (Murphy, 2012). In Kenya these percentages of small-scale farmers correspond to rest of the continent. Small-scale farmers are often marginalized and face many constraints attaining full agricultural production possibilities and improving quality and quantity of produce. Constraints include limited access to technological advancements, weak institutional capacity and coordination, low output and productivity and inadequate market information and market access (Republic of Kenya, 2006)(Munyua & Stilwell, 2010).

Current engagement of smallholders with markets in Africa as producers is limited. For example only 15% regularly sell in markets, 30% sells only occasionally and purchase food from markets and 50% are considered subsistance farmers (Hazell & Rahman, 2014). Small-scale farmers can benefit from market access both for increased output for sale, access to inputs and services that can improve productivity. It is easy to see why markets could help smallholders to benefit from increased production, by selling to more distant urban or export markets the limited demand of local markets can be overcome. In addition, urban and export markets do not only have larger demand but customers in these markets are often more willing to pay for additional quality and demand more varieties of produce (Wiggins & Keats, 2013).

A persistent concern in the agricultural development debate is that opening up markets that demand for so called ‘cash crops’ might lead smallholders to switch their land use and cutting back on production of staple crops used for their households. However, evidence shows that smallholders rarely specialse in solely commercial crops. When production for premium markets increases it tends to be additional to current farm activities. This suggests that ‘cash cropping’ may lead to greater diversity of production rather then less (Wiggins & Keats, 2013). Besides, cash cropping may boost staple production because of the higher incomes generated by the cash crops which potentially allows smallholders to buy better seeds, fertilisers and pesticides. In essence, more market engagement should lead to higher incomes both for those engaging in markets and others living locally through their social linkages (Wiggings & Keats, 2013).

2.2 Linking smallholders to markets for sustainable development

Until the 1990s the focus for agricultural development in developing countries was on sector

governance models and building institutional capacity, by organizing largely set-up extension services where knowledge and innovations were transferred in a linear way (top-down). The results of these approaches have been mixed. Despite the fact that some were a success, many intervention programmes lacked a connection to markets that are now considered essential for establishing sustainable development. Starting in the 1990s, the concern about social and ecological impacts of small-scale agricultural production increased which has led to interventions shifting focus from sectoral to market-led approaches (Molenaar et al, 2015).

Market-led interventions are founded on the belief that markets can lead to poverty

reduction. According to Sahan & Fischer-Mackey (2011) it is however important to acknowledge that markets are biased systems with their own specific ‘rules of the game’, therefore market-led

interventions should take on a flexible approach and include a nuanced understanding of the context where it is being implemented. The market-led approach also aided the emergence of a new type of

(9)

8 arrangement for sustainable development, namely Public Private Partnerships (PPPs).

PPPs are arrangements between multiple parties involving parastatal bodies, private-sector and government (donor) (in the case of HortIMPACT the Netherlands Embassy of Kenya) which often fund NGOs to implement their interests. PPPs are inherently market-based because of the

involvement of the private sector that have commercial interests for their investments in development interventions. PPPs came up in agricultural development during the withdrawal of public funding starting in the 1980s. Partnering with private-businesses has become a way to both farmers and the state to maintain access to credit, agricultural inputs, extension and marketing channels. PPPs for sustainable development are increasingly encouraged as a vehicle to attain

multiple goals in order to achieve both private sector profits and at the same time support sustainable development (Ros-Tonen et al, 2015).

2.2.1 Public Private Partnerships

Nowadays the question whether governments and NGOs ‘should’ initiate interventions partnering with the private sector shifted to ‘how’ should they structure their engagement with the private sector (Ion et al, 2014). The private sector has become the biggest player in developing countries today accounting for 82% of the total investments in developing countries. Official development assistance is currently only the third largest source of investments in the developing world. These figures indicate the growing interest of the private sector in developing economies as new sources for production or for entering new markets. PPPs require an enabling environment from governments to make it more attractive for the private sector to participate in collaborations for development. To understand the challenges that come along with doing business in developing economies’ contexts, the private sector is attracted to partnering with donors and NGOs that have in-depth knowledge of the local context and long-lasting relationships that can be used to accelerate the implementation and the success of the PPP. NGOs in their respect are attracted to collaborating with the private sector to leverage their investments and helping them achieve development objectives more efficiently and cost-effective. PPPs are implemented in a wide variety of approaches. The scope, reach and structure may vary, but for each the fundamental goal is identical: the private sector and donor-agencies partner to mitigate the risks and share the costs of bringing new innovations and sustainable solutions for livelihood improvements that also have commercial profit (Ion et al, 2014).

2.2.2 Value-chain collaboration and inclusive business

An important strand of market-led development programmes, including PPPs have incorporated a value chain approach. Value chains encompass all activities and services that are required to bring a product or service from conception to eventually markets and their consumers, considering that a product (commodity or service) gains value as it moves along the value chain. NGOs that intervene using the value chain approach can help smallholder producers near the beginning of the chain to capture more value by improving their agricultural practices, produce output and quality (Sahan & Fischer-Mackey, 2011). Agricultural value chain actors include producers, traders, processors, brokers, retailers; a range of business, financial and technical service providers including input dealers; and the final markets where a product is sold whether local, regional, national or global (Kolavalli et al, 2015). While donor agencies and the private sector are increasingly working together to increase the impact and sustainability of agricultural interventions, Ion et al (2014) state that PPPs inherently do not benefit the poor. The private sector tends to focus on the ‘low hanging fruits’ rather than addressing fundamental social and environmental problems (Ros-Tonen et al, 2015). As a reaction to this risk, the inclusion of smallholders for sustainable development has gained interest within the development debate. For example, SNV (one of the NGOs behind the HortIMPACT programme central to this research) sees smallholders as an opportunity instead of a problem. The integration of the

(10)

9 poor as integral to companies’ value chains is what is depicted as ‘inclusive business’ (Murphy, 2012). Inclusive business models (IBMs) as an addition to PPPs are dedicated to expand the access of services, goods and livelihood opportunities to the farmers that are at the base of the so-called pyramid (BOP) in a commercially feasible and scalable manner (Jenkins et al, 2010). Much of the IBM discussion for agriculture is dedicated to include small-scale producers as key partners because they are the largest segment of the BOP that operate in agricultural value chains and are ought to benefit from interventions by providing new knowledge, technologies, innovations and market access to improve their livelihoods (Ion et al, 2014).

Figure 1. This figure provides an example of how PPPs and IBM in agricultural value chains work at every stage of the chain. However, IBMs are specifically focused on including the ‘bottom of the pyramid’ (BOP). Source: Ion et al (2014)

Another strand of development scientists proposes the concept of value chain collaboration (VCC) with the private sector as a way to establish inclusive business. Ros-Tonen et al (2015), in particularly stress the importance of looking ‘beyond the chain’. Looking beyond the chain includes horizontal collaborations with non-chain actors, non-commodity production and environmental issues. Value chain analysis typically focusses on power constellations, agreements and arrangements within value chains and therefore the approach runs the risk of losing sight on contextual factors such as culture, economics, politics etc. (Ros-Tonen et al, 2015). The traditional value chain is a linear system with an upstream flow of producers and suppliers providing

commodities or services for eventually consumption. However, a simple linear value chain is no longer seen fit for describing the complex vertical and horizontal chain and non-chain relationships,

stakeholders and partners that must be supported and acknowledged to establish sustainable value chains (Comere, 2012)(Ros-Tonen et al, 2015).

VCC is defined in the article of Ros-Tonen et al (2015) as a voluntary collaboration between different actors in a value chain and other non-chain actors such as NGOs and public (governmental) organizations. They emphasize merging vertical chain relations with place-based interaction through ‘horizontal collaborations’ where all actors involved should interact with each other at the farm-level so they can achieve a better understanding for the local context of smallholders and the different interests and demands of each stakeholder. An example of horizontal collaboration in a sustainable development approach is to let various actors in a value chain get to the farm level and make them responsible for training sessions and interact with the farmers as will be further explained in the empirical chapters of this research.

Important for successful multi-actor (chain and non-chain) horizontal collaborations is acknowledging that actors are embedded in a complex of horizontal, vertical and business support relationships with organizations and the private sector supporting inputs and services (for example advisory services, donors, NGOs, farmers, and private businesses). Relationships are not only shaped by economic considerations, but also through other concepts like trust, power and reputation that

(11)

10 have impact on the structure of a collaboration. Well established relations may improve the “social capital” of an intervention, by making it feasible to get easier access to information, technical know-how and financial support by encouraging knowledge transfer between collaborating partners that potentially can reduce transaction costs and improve access to markets for farmers (Trienenkens, 2011).

For this research the focus will be on the IBM concept instead of VCC because SNV is an important actor in the HortIMPACT programme and they promote ‘inclusive business’. However, the similarities and main goal driving both approaches are eminent. Both frameworks are concerned with the inclusion of smallholders for sustainable development, are market-led, include PPPs, and have incorporated a value chain approach. Although using the IBM concept for the analysis of the HortIMPACT programme, the importance of establishing horizontal collaborations to achieve smallholder inclusion for intervention programmes put forward by the VCC literature will be supplemented to the IBMs.

2.2.3 Opportunities for using IBMs

Inclusive business models (IBMs) thus support PPPs (in particular the private sector) to turn

underserved populations into sources of supply and consumers of their products. For NGOs, looking for ways to integrate these excluded farmers into PPP development interventions, providing support for inclusive business models is becoming progressively adopted. Opportunities for inclusive business exist across different stages in the value chain, from production to retail for all stakeholders involved (table 1). For instance, at the production stage of the value chain an inclusive business strategy to provide support services, access to financing, information and inputs will lead to private sector benefits because working directly with smallholders allows for more control over the production process so food quality and food safety standard of market demands meet. Smallholders benefit from such a IBM strategy by opening up new markets with higher quality demands which are willing to pay a premium directly beneficiary to farm incomes.

(12)

11

Table 1. Potential opportunities for different stakeholders implementing an inclusive business model for sustainable development at different stages of the value chain. Source: Ion et al (2014)

The role of innovation for sustainable development

A constant factor for sustainable agricultural development is the role of innovations. Whilst the approaches of how, to whom and what kind of innovations should be transferred has been subject to changing narratives through time, the central/integral role for innovations in intervention programmes has remained.

Innovations are also seen as critical elements of inclusive business models (Ion et al, 2014). As development partners in PPPs are leveraging their knowledge, expertise, skills in order to solve challenges for agricultural

development more efficiently. To support the 21ste century challenges such as the rising food demand and food security, innovations are necessary to accelerate improvements in agricultural value chains in developing countries while simultaneously improve the livelihoods of farmers.

The concept that should be clarified is ‘innovation’. In the literature, authors have defined the concept ‘innovation’ in many different ways (Anandajayasekeram & Gebremedhim, 2009). The OECD (2010:1) defines it shortly as ‘the introduction of a new or significantly improved product, process or method’. Or more in-depth:

‘By definition, all innovation must contain a degree of novelty. The Oslo Manual distinguishes three types of novelty: an innovation can be new to the firm, new to the market or new to the world. The first concept covers the diffusion of an existing innovation to a firm – the innovation may have already been implemented by other firms, but it is new to the firm. Innovations are new to the market when the firm is the first to introduce the innovation on its market. An innovation is new to the world when the firm is the first to introduce the innovation for all markets and industries’

(OECD, 2010:1).

Anandajayasekeram & Gebremedhim (2009) state that the most useful definition of innovation is ‘the economically

successful use of invention’ (Bacon& Butler, 1998). Where it is defined as ‘a solution to a problem’. It also implicates that only when an invention is used and disseminated it can be depicted as innovation. Innovations are not limited to only the technical domain but it also includes the organizational, managerial, service delivery and institutional domain.

(13)

12

2.2.4 Approaches for implementing interventions

This section is concerned with discussing approaches for the implementation of interventions that are relevant to this research. First the section will elaborate the focus on entrepreneurial farmers as target population for development programmes. Then the lead farmer model will be explained together with assumptions of a ‘trickle down’ effect for the dissemination of knowledge by addressing these entrepreneurial farmers and use of the lead farmer model.

Recently, renewed appreciation for so-called entrepreneurial farmers has emerged in the

development debate with the awareness that subsistence farming cannot eliminate food insecurity together with the assumption that the commercialization of agriculture is integral to economic growth and development (Jaleta et al, 2009). There is no fixed definition of entrepreneurship (Lans et al, 2013). However, Kahan (2012) identifies two parts to entrepreneurship relevant to agriculture. The managerial skills to run a profitable farm business and the second is a ‘entrepreneurial spirit’. The first can be learned, but the second part; the entrepreneurial spirit cannot (Kahan, 2012).

Farmer-entrepreneurs see their farm as a business and as a means for earning profits. They are passionate about their farm and are willing to take calculated risks to make their profits and business grow. An entrepreneurial farmer is aware of its linkage with the market and knows that new opportunities are found in the market. They have the drive, initiative, and capacity to seize these opportunities (Kahan, 2012). The entrepreneurial farmer is more reliable on external factors than traditional style farmers, because they embrace new types of crops (e.g. hybrid seeds), new techniques, and innovations. The need for more inputs such as fertilizers, pesticides and training in the right use of these new

techniques are examples of external factors. Entrepreneurial farmers are thus, market oriented and willing to adopt new inputs and techniques to generate more income (Kahan, 2012).

This notion has led to a grown interest to use these entrepreneurial farmers in upscaling and out scaling of new technologies and knowledge to many farmers. This strategy is called ‘farmer-to-farmer’ extension and aims at improving economies of scale in knowledge dissemination and innovation diffusion through the social linkages of the entrepreneurial farmers (Amudavi et al, 2009). By training the first few, the farmers are expected to influence other farmers in their network to adopt new practices and techniques to create a ‘trickle down’ effect. That farmers learn from their peers, neighbours and friends is an important assumption that underpins approaches like farmer-to-farmer extension (Foster & Rosenzweig, 1995).

Addressing a ‘lead farmer’ is also a way of farmer-to-farmer dissemination. Lead farmers are more competent than follower farmers in new technology uptake (Alene & Manyong, ,2006). They have a specific role in promoting new techniques within his/her network of farmers. Selection of these farmers are generally based on their level of technical expertise, position in community and level of literacy. Responsibilities for lead farmers are preferably: motivating others to adopt new innovations and knowledge, representing their community and must take the lead in practicing new knowledge (Chabata & de Wolf, nd). The lead farmer thus the representative for his/her community and acts as an intermediary actor.

This lead farmer model approach however also has some risks and critiques. It is dependent for the adoption of new techniques and practices on only a few farmers. Studies have found that although these entrepreneurial farmers and lead farmers gain skills and knowledge through the training sessions of these approaches, the dissemination is sometimes hampered by a reluctance of knowledge sharing (Alene & Manyong, 2006).

The willingness to share information is thus an important supporting factor for the success of farmer to farmer extension and the use of a lead farmer model. Knowledge sharing requires a set of

behaviours that aid the exchange of information. Many researchers have theorized that social capital contributes to knowledge sharing (Chow & Chan, 2008). Social capital exists in the relationships

(14)

13 between individuals. Three main clusters of social capital are identified in the article of Chow & Chan (2008). These are: structural, relational and cognitive. The structural dimension involves the social network relations concerned whose connections defines who and how others can be reached. Relational social capital describes the level of trust in an individual’s network or community. Cognitive is concerned whether individuals feel people in their network have shared goals and visions. All three dimensions are considered important for enabling knowledge sharing (Chow & Chan, 2008).

2.3 Livelihood strategies of smallholders and market access

To create a bridge with the previous chapter the success of the implementation of a PPP and IBM intervention for sustainable development also depend on the composition of its targeted farmer’s livelihood assets and capabilities (combined their livelihood strategies) that either enhance of constrain their involvement to an IBM including horizontal value chain collaboration and the likelihood of adoption and dissemination of new innovations to others through knowledge sharing. The following section will elaborate on the concept of livelihood strategies of farmers and their opportunities and constraints regarding market access.

Considering that IBMs are designed to improve the market access of smallholders, it is valuable to better understand the factors constraining market access for small scale farmers. Examples are; high transaction costs, business attitudes, quality standards leading to a high rate of produce rejection, and weak bargaining power due to small volumes of produce and proximity to the market. It is stressed that when smallholders form producer groups to increase their accumulated volume and bargaining power would have promising prospects to achieve market access (Hazell & Rahman, 2014). Smallholders can deal with markets in numereous arrangements. Wether selling directly to

consumers, through brokers, via contracts etc. Given this context it is important to understand the issue of market access of smallholders in a conceptual manner; small-scale farmers deal with output and input markets through various channels; these markets are often imperfect and unfavourable to smallholders for reasons of scale; proximity issues; higher transaction costs; and limited bargaining power because of resource constraints. In theory smallholders have many channels to access markets but in practice the choices are limited. Smallholders face production and marketing risks which make them vulnerable. Production risks can come from weather, pests, low-yielding seeds etc. Whereas marketing risks involve fluctuating market demand, higher quality standards, and price volatility (Hazell & Rahman, 2014).

These constraints and vulnerabilities for smallholders to interact with markets form the base for a certain strategy towards a market. One factor is whether the farmer is willing and able to accept the potential risks of producing for the market, these choices are context specific and influenced by a farmer’s social network in which he or she is embedded (social capital). Scoones (1998) identifies three different strategies towards markets that also will be considered for the analysis in this research. The first is ‘migration’ which means an active distancing from the markets because the potential risks are deemed to high. The second strategy is ‘diversification’. Difersifying sources of income conducting agricultural and non-agricultural practices, for example growing different types of crops to spread the risks that farmer’s may encounter by entering markets or through other

(environmental) risks. The third strategy is ‘intensification’, this may include the upscaling of production to a particular crop grown for the market and increased integration to value chains to access higher value markets.The interest for this research is to investigate how the participating farmers engage with the HortIMPACT programme to better understand their livelihood strategies and aspirations towards market access.

(15)

14

2.3.1 Livelihoods framework

As was mentioned earlier, smallholder farmers face many possibilities (table 1) but also constraints and potential risks regarding increased market access. To help conceptualize the orientations of smallholders towards markets, livelihood framework will be used. The chosen combination of assets and activities including cultural and social choices influencing their access to these assets are referred to as a farmer’s ‘livelihood strategy’ (Ellis, 1998).

A livelihood comprises the assets (both material and social resources), capabilities and activities that are required for a means of living (Scoones, 1998). The concept of livelihood strategy has become central in the development practice (Brown et al, 2006). The DFID’s sustainable livelihood framework explains the concept of livelihood strategies as:

“the range and combination of activities and choices that people make in order to achieve their livelihood goals. Livelihood strategies include: how people combine their income generating activities; the way in which they use their assets; which assets they choose to invest in; and how they manage to preserve existing assets and income. (DFID, 2001 in FAO website (nd)).

This definition suggests that a livelihood is not merely the access to material assets for well-being but needs to be understood by people’s capacity to make choices and how their institutional context influences their choices. The ability to pursue livelihood strategies is dependent on social and material, tangible and intangible assets that households have at their disposal. Drawing from an economic analogy livelihood resources may be seen as the ‘capital’ base from which livelihoods are constructed (Scoones, 1998). In the livelihoods framework of the DFID five different types of capital are identified to better understand the functioning of a certain livelihood (Morse et al, 2009):

o Natural capital: refers to land resources, roads, water resources, crops grown.

o Physical capital: referring to assets such as farm implements, machines, and methods of capturing water (irrigation), technology.

o Financial capital: access to financing, income, savings, loans.

o Human capital: level of education, skills, experiences, age, gender, occupations etc. o Social capital: refers to social linkages, networks and connections (both vertical and

horizontal), trust, shared goals, willingness to share knowledge, membership of groups/organizations etc.

These five assets in the DFID livelihoods framework form the resource base of a farmer’s livelihood orientation. Each capital is linked and not restricted to its own classification. For example, increased skills and practices (human capital) may lead to growing of higher quality crops which will generate higher income (financial income). With the increase skills and income an individual can invest in better farm implements (physical capital) to illustrate how all capitals are connected in the

framework. However, special emphasis for this research is on social capital. An individual is influenced by its peers and networks in decision making. Social capital is an important enabler but could very well also be a limiting factor for the adoption of new skills and technology or for accessing other capitals (Foster & Rosenzweig, 1995). Considering the research is concerned with an intervention that works with farmers as a way to improve their livelihoods by transferring new techniques, innovations or technologies it is important to investigate how social capital influences the participant farmers’ approach the HortIMPACT programme.

(16)

15

2.4 Matching partners for sustainable development

With the previous two sections of the literature chapter explaining current sustainable agricultural development approaches by multi-actor collaborations, through PPPs, aiming to link farmers to markets in the first part and discussing strategies from these respective farmers towards market access in the second part. This third section is concerned with conceptualizing the role of a broker organization in establishing a PPP for development and how the ‘matching’ of the different needs and interests within multi-stakeholder collaborations (as PPPs and IBMs) occur.

In collaborations the interests and goals of the various (interdependent) actors are likely to be conflicting (Klerkx & Leeuwis, 2008). This means that in practical terms, the question arises how a production base of farmers can organize its demand for technology, knowledge and organizational change; what mechanism will facilitate and foster these linkages; and who can coordinate the collaboration that is needed for innovation and sustainable development (Klerkx et al, 2009). This notion paves the way for brokers that can facilitate linkages between the different stakeholders and matching the demands and needs of each actor involved in a partnership for agricultural development (Wiggings & Keats, 2013). A broker organization is concerned with upholding the interests of each participating actor in the development partnership.

The term ‘broker’ is widely used in social networking literature as an actor which links actors that otherwise would not have been in contact. The absence of a link between actors are depicted as ‘structural holes’ which by a broker can be linked and benefit from access to new information. In this regard brokering is a process of: “an intermediary actor facilitating transactions between other actors

lacking access or trust in one another” (Marsden, 1982:102).

Broker organizations are essentially at the heart of a successful collaboration. A broker should be prepared to be entrepreneurial and experimental in their approach (Tennyson, 2005). A prominent catalyst and intermediary for current development partnerships in IBMs are NGOs that aim to improve linkages as part of their mission to improve livelihoods and reduce poverty. NGOs can make commercial links between the private sector and smallholder farmers by facilitating horizontal collaborations, training, and contacts (Wiggins & Keats, 2013). Considering the HortIMPACT

programme, the brokering role is conceptualized as a (donor)NGO-led intermediary for partnerships for development as PPPs and IBMs.

There are temporal stages in the development of a partnership that are important to consider in order to better understand the need of a brokering organization at each stage of a partnership cycle. In the early stage of partnerships when there are no clear outcomes yet, concrete requirements for collaborating partners are difficult to formulate and therefore ‘transaction costs’ can become high. Transaction costs are in the economic discipline defined as the costs that occur in making an

economic exchange. As Dahlmann (1979:147) explains: ‘in order for an exchange between two parties

to be set up it is necessary that the two search each other out, which is costly in terms of time and resources’. The major reason for failure in collaborations as PPPs are identified as information

asymmetries which can lead to ‘mismatches’ between collaborating partners. These mismatches can lead to frustration, high transaction costs or in the event of termination ‘sunk costs’ (Tennyson, 2005)(Holzmann, 2014).

The brokering organization is presented as a ‘market maker’ that actively creates the market and additionally manages the matching process between demand and supply. It assumes a pro-active role in managing the matching process and is thus an active member of the market with its own right. Brokers do not only engage in one transaction but they are concerned with managing the whole matching process efficiently as a purpose to reduce transaction costs for private businesses and farmers enabling innovation in markets that have potential frictions or asymmetries (Holzman et al, 2014).

(17)

16

Figure 2. Conceptual framework of the role for an intermediary organization as market maker. Source: (Holzmann et al, 2014)

2.4.1 Brokering role: the convener

Stadtler & Probst (2011) identify three main distinctive roles for a broker organization in facilitating PPPs during their life-cycle for development. These are; convener, mediator, and learning catalyst. A convener identifies problems and tends to bring partner together. As a mediator it aligns interests and moderate expectations amongst stakeholders, and the learning catalyst that facilitates learning platforms for evaluation of experiences. For this research the focus is on the convener role. The importance of a convener for bringing potential partners together, guiding the partnership design, and ensure that all parties benefit is acknowledged by scholars (Stadler & Probst, 2011). The role of the convener can better be explained by using the role of social capital theory and a distinction between bonding and bridging social capital. Social capital (as was mentioned in the previous chapter) is defined as a set of resources such as values, norms and trust that are accessed through a network of social relations that can be mobilized to facilitate action. With this context bonding capital stems from strong social ties embedded in close relationships and shared norms based on trust and information sharing that could facilitate cooperation (Coleman, 1990). Bridging capital emphasizes resources through external linkages. By bridging ‘structural holes’ mentioned earlier as a brokering organization, actors can access new contacts and resources. A brokering organization is thus

concerned with matching between different actors and provide bridging ties. As a convener role the broker organization should have a base of trust and are respected for their commitment to

development from collaborating partners with a track-record of successful development projects and have various relations in their target development-area. Figure 3 provides a conceptual scheme of the different steps and activities a brokering organization should include in the life cycle of a partnership for development (Stadler & Probst, 2011).

(18)

17

Figure 3. Functions for the role of a convener as a brokering organization throughout different stages in the partnership. Source: Stadler & Probst (2011).

In the problem-setting phase a broker identifies problems or opportunities and raises awareness for these problems, identifies relevant stakeholders and motivate them to get involved. In the direction-setting phase a broker connects partners to collaborate towards a common goal. The implementation phase makes sure that partners are satisfied with how the process is going by organizing process meetings and if needed make adjustments. In the review phase the brokering organization looks for possibilities to scale up the approach for further development.

2.5 Conceptual scheme

This conceptual scheme is based on the main concepts discussed in the literature chapter deemed important for the research of the HortIMPACT programme. Debates about the inclusion of

smallholders and their livelihoods and strategies are attempted from a ‘inclusive business’

perspective. A development programme as HortIMPACT consists of private sector involvement (PPP) aiming at the inclusion of smallholders to achieve inclusive business. This can be achieved by proper horizontal collaborations as is explained in the literature. How all interests and needs of the different stakeholders in such a collaboration are matched including the livelihood orientations of the

(19)

18

3. Research methodology and methods

Building further on the literature, this chapter outlines the methodology influencing how the research is conducted. It includes what philosophical considerations led to the approach, what methods are used for collecting the data, and how core concepts are conceptualized to enable answering the research questions.

3.1 Research approach

For this research I have used a qualitative approach in order to get an in-depth understanding of the HortIMPACT programme and answering the developed research questions. The research therefore honors an inductive style with acknowledging the importance of rendering the complexity of a situation (Creswell, 2013). The type of research can be typified as being descriptive considering the research questions are designed to map out the HortIMPACT programme. Indicators for descriptive research questions are ‘how’ and ‘what’ also used for the questions that are formulated in this thesis. The philosophical assumption (or world view) I bring to the study is that of interpretivism and constructivism. These assumptions are typically seen as an approach for qualitative research. With the interpretivist assumption I reject that the social world can be measured using the same principles as natural sciences. The social world requires a different logic of research procedure, one that reflects the distinctiveness of human behaviors, mainly focused on understanding this behavior (Bryman, 2012). The constructionist assumption implies that social phenomena and categories are not only produces by social interaction but they are also constantly subject to changes. Therefore, a

researcher always presents a specific version of social reality rather than one that can be considered definitive (Bryman, 2012).

The relevance of the research is that it provides unique insights of how a Dutch development programme in Kenya is executed and what approach is used to establish sustainable development considering its contextual and recently changed policy environment (new Dutch agenda for foreign aid, trade and investment since 2013).

3.2 Research questions

The purpose of this research is to give an insight of how the HortIMPACT programme optimizes its approach by matching different stakeholders considering its context of the implementation and how the approach is influenced by current developmental policy narratives. The main question for this research is therefore:

How does the HortIMPACT programme optimize its approach for sustainable development considering its contextual and developmental policy environment?

To help answering the main research question the following sub-questions are developed: o In what political/developmental policy environment is the HortIMPACT programme

situated?

o What is the approach of the HortIMPACT programme?

o What are the livelihood orientations of the participating farmers and how do they relate to

their constraints in market access?

o How does the HortIMPACT programme execute its role as a brokering organization by

(20)

19

3.3 Operationalization

Based on the literature and the conceptual scheme under chapter 2, an operationalization table is developed which includes three main concepts that are investigated in the research considering the research questions. These concepts are: inclusive business models, livelihood strategies, and matching (see appendix 10.1):

Inclusive Business Models o PPPs o Market access o Horizontal collaborations o Inclusive strategy o Value chain Livelihoods

Livelihoods are conceptualized using the DFID framework. Here the different capitals and their relations to markets are conceptualized:

o Capitals (human, social, physical, financial, natural) o Market relations

Matching

Matching is conceptualized as being: o Market maker

o Intermediary / broker o Convener role

o Lower transaction costs

3.4 Research design & unit of analysis

The research is designed as a case study. A case study entails an in-depth and detailed analysis of a single case. Case study research is concerned with the complexity and particular nature of the case in question (Bryman, 2012). The term ‘case’ is most commonly associated with the study of a location such as an organization. Exponents of a case study design often favour qualitative methods, such as observations and informal conversations because these methods are helpful for the generation of intensive and detailed examination of the case (Bryman, 2012).

Critique on conducting a qualitative case study are centered around the concern about generalizability. How can one particular case be representative for other cases in the same thematic area? However, this critique can be waved because a case study is not meant to be generate typical cases that can be generalized (Bryman, 2012). As the research design being a ‘case study’, the unit of analysis is the HortIMPACT programme’s approach considering its contextual environment (both of participating farmer’s livelihoods orientations, policies, and private sector involvement) and the matching of interests and needs of all different stakeholders in that chosen approach.

3.5 Sampling method

For the research I have made use of what is known as purposive sampling. This type of sampling is concerned with the selection of units which can be organizations, people, documents etc. with direct reference to the research questions. The research questions guide these ‘units’ of interest for the sampling method (Bryman, 2012). First step was the selection of the HortIMPACT programme. Then

(21)

20 the locations which were selected on the base of convenience sampling. These locations were opened to me after by my key informant Jane Kamau. Also the selection of farmer respondents was based on a mixture of purposive and convenience sampling. The purposive sampling method was to select one training location to get a better in-depth understanding of its participants at one location.

Convenience comes into play because this location had the closest proximity to Nairobi and it was the last opportunity due to time constraint to visit another training of the programme.

3.6 Research methods

In order to answer the main and sub-questions of the research a mixture of methods is used. These methods are chosen based on experiences in the field and planned methods of collecting data proposed in the research proposal. First, and major contributor of the collection of data was done by conducting participant observations and having informal conversations during my visits at training sessions of the HortIMPACT programme while joining Jane Kamau (HortIMPACT advisor). Also a transect walk with the lead farmer Micheal in Biberioni was used to gain a more in-depth

understanding of the context and livelihoods of farmers in the area. Questionnaires were used to collect data from participants of the training sessions which included fixed and open ended questions. A semi-structured interview with Klaas de Vries (HortIMPACT advisor), and analyzing existing

interviews online with Stefan Engels (HortIMPACT team leader), Melle Leenstra (food security advisor for the Netherlands embassy) and Kiringai Kamau (advisor to the ministry of agriculture of the

government of Kenya). Lastly, I reviewed official documents of the HortIMPACT programme.

3.6.1 Non-participant observations and informal conversations

Observations were conducted during multiple visits in the Biberioni area and joining Ms. Kamau for two days by participating in two training sessions of the KHS business case. Here the implementation of the training sessions and context of livelihoods were observed. It was a good way of seeing how all actors interact with each other during training sessions and observe local practices of farmers in everyday life identifying possible issues that can be related to constraints mentioned by respondents. In other words; doing the observations helped developing a deeper understanding of the context of where the HortIMPACT programme is being implemented.

I gained access to the training locations by the invitation of Ms. Kamau to join her for a two-day trip covering two training sessions in the Nakuru region of the KHS business case which made her my key informant for the research. After the two-day field trip, I was allowed by the project leader of HortIMPACT (Mr. Engels) to go by myself to one training locations closest to Nairobi, namely

Biberioni. I received the telephone number of the lead farmer of the programme in that area named Michael. He was very helpful and provided useful information about farming in Biberioni. I spent a whole morning with Michael in which he took me on what can be depicted as a ‘transect walk’. This tool can be used for describing and showing the location (in this case the demo-farm of Michael in Biberioni) and distribution of resources, features, landscapes and main land-uses (World Bank, nd). During the walk I had the chance to have an informal conversation with Michael about his life and the community of Biberioni.

Informal conversations also were used during my two-day fieldtrip with Jane Kamau. Reason was that it was not feasible to record all conversations we had during these days. It also did not feel appropriate to record our conversations while being at the training sites. Also during meetings, organized by our handler Naomi in Nairobi the meetings were not recorded. Important quotes or statements made by the respondents were immediately written down in my phone or in the evening using my computer.

(22)

21

3.6.2 Questionnaires

From gained experience during my participant observations at two training sessions of the

HortIMPACT programme I chose to develop a questionnaire to enable collecting data from farmers that participate in the programme. Main reason for choosing this method was because it turned out to be difficult to get the right information using interviews with the farmers. As I learned most people did have had English classes in primary school and/or high school but farmers found it easier to read English instead of speaking English. Thus, the biggest constraint was the language barrier for

collecting valid data. Although a questionnaire does not give the same possibility to gain more insight on motivations and opinions compared to interviewing, the circumstances explained justified the use of this method. Including statements scored on a Likert scale was one of the ways to include

qualitative opinions from the respondents combined with open-ended questions.

The developed extensive questionnaire (78 questions) that could also very well be self-completed covered several themes such as: personal information, farm characteristics, marketing information, household information, knowledge sharing, participation of programme, evaluation of the programme, and livelihood aspirations (see Appendix 10.2). The questionnaire was done at one the training sessions of HortIMPACT in the Biberioni area near Limuru town. This was the closest location from Nairobi taking approximately 1.5 hours’ drive to get there, making it feasible to visit the location several times.

First, a pilot version of the questionnaire was tested (consisted with 48 questions) in which I managed to collect seven respondents with the help of the lead farmer in the Biberioni area. It turned out to be very difficult to find farmers in the scattered landscape of Biberioni and due to the fact that farmers often do not stay at their homes during daytime. Again from this experience the decision was made to collect data at one of the training sessions from HortIMPACT. The expectation was (based on former participation of training sessions in Biberioni) that about 20 farmers would be participating this session. Unfortunately, this training only had about 10 participants due to the weather conditions (it rained that day) and due to the season because farmers were very occupied on their lands as was explained by the lead farmer. These circumstances resulted in again seven completed questionnaires. Therefore, some data has N=14 and other (added) questions from the definitive questionnaire have N=7 (excluding missing answers).

Considering that the questionnaire took an average of 20 to 30 minutes, I arranged five students from the Technical University of Kenya to assist me with conducting the questionnaires before and after the training session. I planned a meeting with the Kenyan students to walk through the different stages of the questionnaire to make sure that everyone understood each question in order to increase inter-interviewer reliability of the data. Another reason for letting Kenyan students conduct the questionnaires was to match the characteristics of the interviewers with the respondent farmers. This should have helped by reducing undesired response sets of the respondents such as acquiescence and social desirability biased answers (Bryman, 2012). Also the ability of these students to explain questions in Swahili turned out to be very helpful to get better answers.

3.6.3 Focus group discussion

One modest focus group discussion was conducted with six women working on the land together in the Biberioni area. It was not planned, but I happened to have a Kenyan translator with me that day. Some topics of interest were addressed in which the women jointly discussed their opinions. Key statements were directly translated to me by my translator which I wrote down in my phone.

(23)

22

3.6.4 Semi-structured interview

A qualitative semi-structured interview was conducted with Klaas de Vries via a Skype call and recording the conversation. The interview facilitated gaining a more in-depth understanding about the approach used in the HortIMPACT programme and its relations with the private sector, farmers and government institutions. Using a semi-structured approach for the interview I compiled a topic list of subjects I wanted to address based on the research questions (see Appendix 10.3). The semi-structured approach of interviewing offered me the possibility to pick up on interesting notions made by Mr. de Vries to uncover other topics allowing for the increased possibility to gain more information (Bryman, 2012).

3.6.5 Reviewing existing interviews online

Using existing television interviews with some key actors involving the HortIMPACT programme also gave better insight about the policy environment and HortIMPACT. One item from the Kenya television broadcast about the HortIMPACT programme consisting of respondents, Melle Leenstra, Kiringai Kamau and Stefan Engels gave insight on how the HortIMPACT programme is positioned in current policy debates in Kenya and the Netherlands embassy. Using this method allowed for collecting information from people that would have been difficult to reach in the short timeframe of my fieldwork.

3.6.6 Official documents

As Bryman (2012) states official documents can be a very important source of information for researchers conducting case studies of organizations using participant observations and qualitative interviews. However, there are risks of credibility and representativeness with using official

documents that come from private sources (Bryman, 2012). The documents used for this research are indeed sometimes summaries of findings, which do not allow the same understanding as raw data. Other documents used are the official agreement (contract) between one of the private companies (Kenya Highland Seeds) and HortIMPACT. Also the official application form of KHS to co-invest together with HortIMPACT was provided by the HortIMPACT programme which allowed for unique knowledge on what terms and agreements the partnership is conceived.

3.7 Data analysis

The research and analysis was done by an iterative process in which the data collection and analysis were constantly subject to comparison and making adjustments. The first step was meetings with people working for the HortIMPACT programme and collecting data about their approach. Through this collected data and provided access by the programme first observations and informal

conversations were used as a starting point to develop the questionnaire used at the third training session.

As the unit of analysis dictates how data was analyzed, this meant the analysis was guided by the concepts developed to investigate the unit of analysis. Interviews conducted by myself of the interviews online were transcribed and from this selected themes that are important to the research questions were used and quoted to give insights about indicators and motivations for choosing the approach of the HortIMPACT programme. Considering the number of respondents with the

questionnaires no quantitative analysis is conducted. The sample is considered too little to make real assumptions and discover correlations. However, the same thematic approach was used as indicators for a modest explanation of the farmer’s motivations and constraints in conjunction with the open ended questions which allowed for distinctive quotes. The official documents were also analyzed using the themes that are directly linked to the research questions.

(24)

23

3.8 Ethical considerations

Ethical issues in research command increased attention today (Creswell, 2013). Ethical considerations that are needed to be anticipated are extensive and reflective throughout different stages in the research process. It is important to gain permission from the respondents and/or organization. Make sure that the purpose of the research is disclosed to the respondents; avoid deceiving respondents and make sure everyone receives the same treatment and avoid siding with the respondents or disclosing only positive results (Creswell, 2013). Ethical considerations in social research can be broken down into four main areas: harm to participants; lack of informed consent; invasion of privacy; and deception (Diener & Crandall, 1978).

Each respondent in the research was involved with attaining informed consent before data was collected. The purpose of the collection was explained as being a research for the University of Amsterdam. This was especially important with the farmers because they might have thought it was for a research for a private business of the HortIMPACT programme. By using key informants as Ms. Kamau and the lead farmer Michael concerns of participants could be communicated with them and when a farmer was not willing to cooperate I never put him or her under pressure.

(25)

24

4. From aid to trade

In what developmental policy environment is the HortIMPACT programme situated?

This section aims at explaining the policy environment in which the HortIMPACT programme is situated and how it may influence the structure of the programme. Hence, HortIMPACT is funded by the Netherlands Embassy in Kenya that has a specific interest and policy agenda for investing in the HortIMPACT programme. The Netherlands Embassy follows the foreign policy agenda and the programme is being implemented in Kenya so these three (policy) government actors will be discussed in this chapter.

4.1 Government of Kenya – Vision 2030

The government of Kenya is included in this section because the policy environment in which the HortIMPACT programme is situated is for a large part dictated by the Kenyan rules and regulations, therefor have an impact on the structure and approach of the programme. The Kenyan government have set out a new multi-term plan that is named; ‘Vision 2030’. This vision is concerned with the development blueprint for Kenya covering the period 2008 to 2030 (2007). The Vision 2030 of the Kenyan government aims to transform Kenya into a middle-income country which can provide a high quality of life for all its citizens. As Dr. Kiringai Kamau (advisor to the ministry of agriculture and fisheries of the Kenyan government) says in an interview: ‘achieving food security is a national goal,

one that the Kenyan government hopes to achieve by shifting from aid-dependent agriculture to agribusiness’1_{. In 2013 the Kenyan government passed a new act which is called the ‘public-private} partnership act’. With this act, a special government unit was installed to function as a resource center for best practice and safeguarding the integrity of PPP processes. Also making

recommendations to the parliament and ensuring that PPP projects meet quality standards such as; value for money, affordability, and transfer of risk (PPPunit, 2016). Also a quote about the adopting the PPP act from their website: ‘the Government of Kenya is confident that through the PPP modality, the private sector can offer a dynamic and efficient way to deliver and manage public infrastructure. These efforts are geared towards achieving Vision 2030, Kenya’s long-term development strategy so that future generations can gain from the benefits of modern services, improved living standards and reduced poverty’ (PPPunit, 2016:”Background”)

However, as Dr. Kiringai Kamau also states that de concept of Public Private Partnership is still new in Kenya. Kamau: ‘we can see that when you create avenues for linkage of foreign organizations

and local organizations that it is opening markets2_{. According to Mr. Engels, the project leader of the} HortIMPACT programme it is very important to create an environment that allows for easy

investments and easy business. He states that the Kenyan government has done a great deal of work the past 10 years to improve the legislative environment for doing business. The government of Kenya about this topic: ‘we are looking at systems and processes, making them as easy as possible so

that no one has any reason to complain when they come to doing business in Kenya’3_{. Thus, can be} inferred from this statement, together with the publication of Vision 2030 that Kenya is actively working on creating an advantageous environment for foreign and domestic businesses but also for initiatives such as the HortIMPACT programme. The programme, however, is a fully Dutch-funded development programme. Therefore, the next section will be involved with current changes in Dutch foreign policy considering that HortIMPACT emerged from this new vision for foreign bilateral relations of the Netherlands.

1_{Interview Kiringai Kamau}

2_{Interview Kiringai Kamau} 3_{Interview Kiringai Kamau}