Why don\u27t small-scale producers supply French bean export markets in Kenya?

In this dissertation, I explore some of the issues which could explain why small-scale producers do not supply French bean export markets in Kenya. The main result from these essays is that Kenya\u27s French bean market has structural features - including imperfect contract enforcement, partner search frictions, and limited persistent benefits from contracting - which may help deter small-scale producers from participating in the market. In the first essay, I examine the role of imperfect contract enforcement as a barrier to entering the market and potential explanation for exit decisions. I find that imperfect contract enforcement deters entry, complicates exit, and reduces land allocation on the intensive margin. I also find that cost factors do not deter market participation, but help sort entrants into different market segments. In the second essay, I examine the role that partner search frictions play in conjunction with imperfect contract enforcement to serve as a barrier to entry and potential explanation for exit decisions. I find that search frictions are present in the market which primarily limit farmers\u27 abilities to match with potential buyers and not vice versa. I also find that search frictions are a potential barrier to entry for the fresh submarket and may be a factor in farmers\u27 exit decisions from the processed market. Lastly, I find that search frictions are a potential barrier to market in areas where buyers are more reliable on average, suggesting that there may be a trade-off for policymakers depending on whether they design an intervention to target contract enforcement or search frictions. The results are relevant for policy-makers who want to improve on existing programs to connect small-scale producers to French bean export markets in Kenya. This dissertation also makes several methodological contributions to the empirical contract and search literatures. First, I introduce a choice experiment design that is capable of assessing the effects of imperfect contract enforcement on the intensive and extensive production margins, and also of assessing the degree of imperfect contract in a market. Second, I demonstrate the efficacy of a geographically-nearest neighbor algorithm, which has low data requirements to implement and ensures that matched pairs experience similar market conditions. Third, I introduce a novel method for measuring search frictions based on variations in the local spatial density of firms and farmers in the market. These contributions should be helpful for research on the design, adoption, and use of contracts in a variety of intra- and inter-firm applications where it is otherwise challenging to disentangle the effects of unobservable heterogeneity and endogenous matching of agents to contracts