The effectiveness of contract farming for raising income of smallholder farmers in low‐ and middle‐income countries: a systematic review

Abstract

1. Background Motivation for the review Contract farming is a commercial relationship between a firm and a group of farmers. It is a business model in which production is bought in advance by a firm in exchange for certain services and other benefits. Although principally a commercial initiative, contract farming is considered to be a way to overcome the challenges that small farmers face when linking to remunerative markets. It assists farmers in connecting to output markets and often provides inputs, credit, or agricultural extension (Eaton and Shepherd, 2001; World Bank, 2007; Da Silva and Ranking, 2013). These services can be not only provided by private firms, but can also come from, or be facilitated by, multi-actor partnerships between companies, governments and NGOs (Prowse, 2012). Estimates of the incidence of contract farming in developing countries are unreliable and differ markedly between countries; they are generally below 10% of total area under production (Minot and Ronchi, 2015). A rapidly growing number of firms – at least in modern market channels – are increasingly relying on contract procurement (Da Silva and Ranking, 2013). Companies may offer a contract only to those farmers who comply with some minimum requirements (for example land ownership, irrigated lands, minimal plot sizes, etc.) which may preclude smaller farms from benefitting directly. Even if these arrangements are beneficial for smallholders directly or indirectly through spill-over effects, there will likely be heterogeneity in impacts, with certain farmers benefitting while others do not or even lose out. It is clear that contracts will not be randomly distributed within a farming community, and contracting farmers will always have special characteristics, or so-called firm-selection and self-selection biases (Minot and Ronchi, 2015; Barrett et al., 2012). Studies that infer quantitative effects of contract farming on income and food security will need a proper research design to control for these biases. Public policies and institutions influence the market strategies and forms of transaction used between farmers and firms. Contracting firms are almost always relatively large processors, exporters, or supermarket chains. Rarely do small-scale traders or even wholesalers offer farmers pre-planting contracts. This is not surprising, given the large fixed costs associated with contracting (Minot and Ronchi, 2015). Contract farming is induced by a firm’s need to source products with specific qualities and in sufficient quantities (Minot and Ronchi, 2015) and the presence of appropriate geographical and political-economic conditions and an enabling business environment (Jia and Bijman, 2013). Relevant geographical conditions are road infrastructure, access to water, soil types, climate, etc. Relevant political-economic conditions are land-rights policies, market regulation, trade policies and the risk of socio-economic shocks. Whether a firm chooses to start offering a contract farming arrangement is also highly influenced by the local business environment, such as financial services, conflict resolution systems, investment subsidies, business development services, brokering services, and farmer organisations. Many of the above-mentioned conditions are influenced by political decision-making. Therefore, policy makers can enable or constrain the opportunities for contract farming, influencing its attractiveness to firms and farmers as a way of organising transactions and embedded services. Since 2007, coinciding with the investor rush for land in sub-Saharan Africa, international development agencies have increasingly presented contract farming as an alternative or complementary development opportunity for smallholder inclusion (Lindholm, 2014). 3 Contract farming is considered by most authors to be a positive development for the inclusion of farmers in markets (Minot, 1986; Eaton and Shepherd, 2001), food security (Bellemare and Novak, 2015), and global poverty reduction (Setboonsarng and Leung, 2014). Yet, there is serious concern whether smallholders are able to benefit from these arrangements, because the relative size of buyers most likely results in an unequal power relationship, which influences the terms of the arrangements (Sivramkrishna and Jyotishi, 2008; von Hagen and Alvarez, 2011). In the earlier literature on contract farming, the issue of power imbalances was especially prominent (Little and Watts, 1994), and the discussion was rather polarised between proponents and critics of contract farming (Oya, 2012). Recent literature, however, has paid more attention to diversity within contract farming arrangements, and, for example, acknowledges the complex interactions involved in a contract farming scheme and how farmers and firms make use of their constrained room for manoeuvre (Vellema, 2002; Minot and Ronchi, 2015). Though, contract farming is the outcome of a complex process influenced by many different factors and in which policies and projects shape the enabling environment in which a firm and a farmer agree on a contract, policy makers, international organisations propose (Lindholm, 2014), and many scholars study (Minot and Ronchi, 2015; Da Silva and Ranking, 2013), contract farming as a way to improve the inclusion of farmer in markets, and consider it a development strategy. Although it is not a well-defined, standard ‘treatment’, we think that a comparison of these effectiveness studies on various empirical instances of contract farming will give an overview of the evidence base behind support to contract farming as a development strategy and can derive useful policy recommendations that have wider application. Many studies in the past compared the incomes of participating and non-participating farmers, or incomes before and after the signing of the contract, and have a high risk of selection bias: e.g., rich farmers may participate more than poor farmers, etc. This results in unreliable effect estimates that are more a reflection of selection bias or confounding factors (e.g., weather or prices) than of the contract farming arrangement itself. The last decade shows a rapid increase of studies on contract farming that assess the effects of contract farming using quasi-experimental research designs, which provide much more reliable estimates. As can be expected, these studies report mixed effects (Minot and Ronchi, 2015). These impact evaluations assess the effects of specific instances of contract farming on smallholders. Therefore, a systematic review of this rapidly growing body of evidence is timely, and may help to distil generalised inferences from these specific instances. To do so, our review follows a two-stage process. First, it identifies and discusses the results of impact studies that applied a research design to reduce selection bias and assess the counterfactual situation (What would have happened to the smallholder farmers had there been no contract farming arrangement?). The review compares the inferences in these studies and assesses the strength of the evidence behind them, as well as the rigour of the methods used. Secondly, it places these contractual arrangements in context. The review maps the relevant contextual conditions and elements of the enabling environment for each of the empirical instances of contract farming covered by these studies, searching for missing information through a targeted information search around each of these instances of contract farming, and making a case-based comparative analysis in order to distill recommendations for policy makers and practitioners