Market days are the pulse of rural, economic and social life in many parts of the world and millions of people rely for their daily sustenance on periodic, typically weekly, markets. At their economic core, they constitute a solution to a complex coordination problem: Bringing buyers and sellers together in the same place at same time. Rural periodic markets are also the final link in the food supply chain for millions of people, either as consumers or as smallholder farmers. The livelihoods of vulnerable market participants on both sides of the transaction depend on reliably finding someone to trade with at a predictable price. This process is challenging in thin markets with few buyers and sellers. This research project addresses this issue by studying how scheduling market days in such a way that they do not clash between neighbouring villages can allow buyers and sellers to cross-attend. The goal of this project is to evaluate whether this approach is impactful in raising market attendance and if so whether this has effects on other measures of economic development.
This study leverages a natural experiment in Western Kenya, where the scheduling of market days has evolved over the past century. The study formalises a notion of market schedule coordination proposed in the geography literature. A simple regression analysis that relates outcomes of interest to such a measure of coordination quality would not identify the causal effect of schedule coordination on these outcomes. There may be other factors such as population density or geographic fundamentals that both induce some markets to synchronise their schedule more with their surroundings while simultaneously affecting the outcomes of interest. An in-depth literature review in anthropology, sociology, geography, history, and economics, as well as over 1200 detailed and extensive key informant interviews with village elders and market administrators in the study area allow the researcher team to reconstruct the market scheduling process. Market day scheduling was a function of the distance to existing markets at the time, region-wide weekday preferences, as well as idiosyncratic preference shocks. The study uses this contextual knowledge to separate the observed variation in coordination quality into a component that was to be expected given the known geography and scheduling process, and a component of quasi-random deviations from this expectation. In the analysis that follows, the study uses this latter component to identify the causal effect of coordination quality and frictions on various outcomes of interest.
Food markets in many developing countries are thin and thin markets are volatile: with few buyers and sellers on either end of the transaction, supply, demand, and prices are likely to fluctuate as match rates are low, which makes it difficult for both buyers and sellers to reap the full potential of market participation. Market integration can improve the performance of price systems and the reliability of markets to all participants and ease matching. A number of creative and insightful studies have addressed agricultural markets. Throughout, market integration has been shown to have meaningful effects on people’s lives and livelihoods. While large gains in market integration will certainly be reaped from modern information technology, physical infrastructure, and microcredit, it should be noted that none of these have existed nearly as long as the simple act of synchronizing schedules, which limits our ability to assess their effects in the very long run, and that the latter, if timed at the inception of a new market, can be done essentially costlessly.
Results indicate that quasi-random variation in schedule coordination did indeed influence attendance to this day as well as a number of measures of economic development. The effect is large, indicating that given today’s transport technology, a lot of people live close to more than one market on a given day. The effect size appears to be driven by markets with a road.
