Supermarket contracts and household welfare in the small farm sector: Panel data evidence from Kenya

In many developing countries, the food retail sector is rapidly modernizing. With rising urban middle classes, supermarkets are increasingly replacing wet markets and other traditional retail outlets as the preferred place of food purchases. These shifts in the food retail sector also have implications for agricultural supply chains. While in smaller towns supermarkets primarily focus on the sales of processed foods, in larger cities they offer their customers a variety of fresh fruits and vegetables. In order to manage regular supply of fresh and high-quality produce, supermarkets often do not rely on wholesale markets alone but also contract farmers directly. In many countries, such supermarket contracts also involve smallholder farmers. The existing literature shows that the rapid expansion of supermarkets in developing countries can provide interesting new marketing opportunities for smallholder farmers. Smallholder farmers, who make up the majority of the rural poor, can benefit from supplying supermarkets through higher and more stable prices, as well as better access to information and technology. Hence, the expansion of supermarkets may potentially contribute to poverty reduction and rural development. However, available impact studies show mixed results. While a few studies confirm positive effects of supermarket contracting on farm productivity and household incomes in the small farm sector, others find that marginalized farms are unable to participate in supermarket channels due to technical, financial, or other types of constraints. Smallholder farmers who manage to overcome these constraints often benefit significantly. However, their participation in supermarket contracts is dynamic, with high drop-out rates observed over time. Against this background, this dissertation entails two essays that cover two key questions (i) what types of supermarket contracts do smallholder farmers prefer? And (ii) what are the effects of smallholder participation in supermarket channels on household income and diets? For the first question, inclusion and successful participation of smallholders seem to depend on contractual design. This is difficult to analyze based on observational data alone, because variations in contractual design rarely occur in the same setting. In the second chapter of this dissertation, we address this research and knowledge gap using a choice experiment that we carried out with smallholder farmers in Kenya in 2015. For the second question, one shortcoming of the existing research is that most studies used cross-sectional data, making the identification of causal effects difficult. A few studies used panel data, but only looked at purely economic effects. We contribute to the literature by using two rounds of panel data to examine the effects of smallholder participation in supermarket channel on household income and diets as elucidated in the third chapter of this dissertation. Addressing the two knowledge gaps would be beneficial in understanding the types of contracts that facilitate smallholder participation in such high value channels as supermarkets, and in unravelling the welfare effects of supermarket contracting on farm households in terms of income and diets. Recent statistics show that 800 million people suffer from hunger and another 2 billion suffer from micronutrient malnutrition. The prevalence of undernourishment and micronutrient malnutrition is high among the rural populations that are also largely smallholder farmers. Hence, our study is relevant in the context of assessing the welfare effects of supermarket contracting on smallholder farmers. We focus on supermarket contracts with smallholders in Kenya as one of the countries with the fastest growth of supermarkets in Sub-Saharan Africa. Currently, supermarkets accounts for 10% of national grocery sales in Kenya and the share is projected to grow with increased urbanization. The first essay in chapter two entails a discussion of the choice experiment results on vegetable farmers’ preferences for supermarket contracts. We used D-optimal design to design the choice cards where we varied design attributes of contracts with supermarkets and other traders to examine farmers’ contractual preferences and how they vary with socioeconomic characteristics. The five attributes included were; output price for the vegetables sold, place of sale, form of sale, timing of sale, and payment mode. We then specified and estimated mixed logit models to account for preference heterogeneity and finally calculated farmers’ willingness to accept (WTA) output price changes for certain contract attributes. The results show that farmers generally prefer supplying traditional marketing channels without a contract. Farmers that currently supply traditional channels are more contract-averse than either the current or previous supermarket suppliers. Farmers that recently entered the supermarket channel have significantly more positive attitudes towards contracting. The results further underscore the importance of contract transparency, and trust between farmers and buyers. Supermarkets contracts usually involve delayed payments but this has evolved over time from payment based on quantities delivered to payments based on actual sales to consumers, without a proper monitoring mechanism. This involves different types of risk. The results indicate that farmers would accept significantly lower prices if they were either paid based on quantities delivered or if they had an option to verify actual customer sales. Farmers’ contractual preferences are also influenced by socioeconomic variables such as age and level of education of household head, and income levels. The second essay in chapter 3 discusses the effects of supermarket contracting on household income and diets. We focus on household dietary diversity, consumption of calories, zinc, iron, and vitamin A as dietary indicators. Data on household food consumption were collected using 7–day recalls and conversion made using food conversion tables for Kenya. We estimated fixed effects and random effects models for panel data analysis. The results show that supplying supermarkets has increased farm household income by 66%. Supermarket contracts also contribute to higher levels of dietary diversity, calorie consumption and zinc consumption. We do not find significant effects on vitamin A and iron consumption. Using the estimated dietary effects for simple simulations suggests that wider participation of farmers in supermarket channels could help to reduce the prevalence of undernourishment by 8% and zinc deficiency by 12%. The final chapter of this dissertation concludes by presenting a summary of the key findings and the policy implications, limitations of the study and suggestions for future research. We conclude that designing contracts with; lower transaction costs, more transparent quality grading system, sales verification mechanisms, and fairer risk-sharing arrangements could enhance smallholder participation in supermarket channels and other emerging high-value markets. Group marketing by farmers could be an interesting option to reduce individual risks and transaction costs. Infrastructural and institutional improvements could also facilitate smallholder access to high-value markets. Further, supplying supermarket channels increases household incomes and improves diets. In summary, smallholder participation in high value chains such as supermarkets is worthwhile and efforts towards facilitating their engagement would not only contribute to poverty reduction, but also to more economic and social sustainability of high value chains in an environment where smallholders dominate agricultural production.