Market Frictions in Developing Countries

In my dissertation, I study how market frictions shape the production decisions of firms and farms in developing countries. I employ a variety of empirical techniques, including randomized experiments and structural production function estimation.;Chapter 1: Do Factor and Financial Market Frictions Interact to Constrain Growth? Evidence from Firms and Farms.;Researchers and policymakers have devoted significant attention and resources to expanding access to finance in developing countries. The underlying theory is that providing credit or insurance should raise investment and allow the most productive firms and farms to grow. However, this simple theory neglects the fact that frictions in factor markets may also limit investment. In particular, for small firms and farms, supervision costs for hired labor make it costly to add labor. For farms specifically, thin land markets make it difficult to add land. Despite a long literature in development highlighting the importance of these factor market frictions, the literature on the impact of relaxing financial frictions has generally ignored this interaction.;In this paper, I combine a model with frictions in both factor and financial markets and two randomized experiments that relaxed financial frictions for firms and farms to estimate the extent to which frictions in factor markets weaken the effect of relaxing financial frictions.;The randomized experiments I use come from De Mel et al. (2008), who provided cash grants to small firms in Sri Lanka, and Karlan et al. (2014), who provided access to rainfall insurance and cash grants to farmers in Ghana. In Sri Lanka, De Mel et al. (2008) show borrowing constraints limit firm investment and that their treatments relaxed this constraint. In Ghana, Karlan et al. (2014) show uninsured risk limits farm investment and that their treatments relaxed this constraint. In both settings, however, there is reason to believe that constraints in markets for labor and, for farms in Ghana, land may be at work, too.;To test whether labor and land market frictions weakened the effect of relaxing financial frictions in these experiments, I develop a model of firm or farm production with financial frictions and potentially factor market frictions. My model predicts that if factor markets are perfect, relaxing financial frictions through the treatments causes firms or farms to increase all inputs "in proportion," such that the ratios of the marginal revenue products of these inputs are unchanged. If, however, there are frictions in markets for certain inputs, such as labor or land, relaxing financial frictions causes firms or farms to increase these inputs "disproportionately less" than inputs that only face financial frictions, such that the ratios of the marginal revenue products of these two types of inputs increases. These ratios of marginal revenue products capture the shadow prices of the inputs with factor market frictions. An increase in the shadow price suggests there are frictions in the market for that input that prevent firms or farms with relaxed financial frictions from buying from firms or farms without relaxed financial frictions.;I implement this test using the experiments. In the Sri Lanka experiment, I find that the smaller cash grant had no effect on the shadow wage. Relaxing financial frictions even more with the larger cash grant, however, led to a 46 percent increase in the shadow wage, suggesting firms hit constraints on labor. In the Ghana experiment, I find that insurance alone led to an 11 percent increase in the shadow wage but had no effect on the shadow price of land, suggesting that farmers were able to acquire enough additional land. However, in the more powerful combined insurance and cash treatment, the shadow price of land rose 22 percent, indicating that farmers began to run into a land constraint, and the shadow wage increased even more (by 36 percent). These results suggest that frictions in labor and land markets did in fact weaken the effect of the treatments, especially those that provided the greatest reduction in financial frictions.;Implementing my test requires assuming a functional form on the production function to estimate the ratios of the marginal revenue products. Specifically, I assume the production function is Cobb-Douglas. I test this assumption using a novel approach. In each setting, I estimate a Cobb-Douglas production function using a dynamic panel approach, which uses lagged inputs as instruments. I then note that the experiments provide additional instruments. These additional instruments permit overidentification tests that I use to test the appropriateness of the Cobb-Douglas functional form assumption. In both cases, I fail to reject the overidentification test.;Overall, my results suggest that frictions in labor and land markets may diminish the effect of credit and insurance programs, as well as other interventions designed to increase firm and farm growth in developing countries.;Chapter 2: Farm Production Function Estimation: Lessons from Firms and Implications for Measuring Misallocation.;This paper extends the recent literature on estimating firm production functions to farms. I argue that the assumptions necessary to estimate and identify production function coefficients using the dynamic panel approach developed in the firm production function estimation literature fit farms well, especially in developing countries. I apply the dynamic panel approach to a panel of Indian farmers in the original ICRISAT survey. I find that the production function coefficients estimated using the dynamic panel approach differ in expected ways from coefficients estimated using OLS and OLS with farmer fixed effects. Using the estimated coefficients to back out the "gap" between the marginal revenue product of inputs and their prices across farmers, I show that the different estimation approaches also lead to substantially different inferences about misallocation and frictions facing farmers. These results suggest that the dynamic panel approach may provide a valuable tool for research on misallocation, frictions and inefficiency in agricultural settings.;Chapter 3: Measurement Error and the Farm Size-Productivity Relationship: An Instrumental Variables Approach Using Self-Reported Land Area and GPS Estimates .;This paper estimates the extent to which measurement error bias explains the inverse farm size-productivity relationship. Traditionally, researchers have measured farm size using farmers' self-reported land area, which is prone to (non-classical) measurement error. In response, recent papers use land area estimates from Global Positioning System (GPS) devices. However, GPS estimates may also have measurement error. I show that instrumenting GPS estimates with self-reported land area resolves bias from measurement error in both measures, provided that GPS estimates have classical measurement error. I apply this approach to data from Tanzania and find that the inverse farm size-productivity relationship remains. OLS with GPS estimates overstates the inverse relationship, relative to my approach, consistent with measurement error in GPS estimates. In fact, OLS with GPS estimates leads to more bias than OLS with self-reported land area. These results are robust to accounting for other explanations for the inverse farm size-productivity relationship as well.