This dissertation studies topics related to labor markets and agricultural cooperatives in developing countries.

In the first chapter, I study the Employment Tax Incentive (ETI), a South African wage subsidy to firms for newly hired young workers. This chapter provides a unique contribution to the literature on the effects of wage subsidies in developing countries. The ETI is one of only a few nationally implemented wage subsidies. Additionally, South Africa has the highest youth unemployment rate in the world. I use a difference-in-difference strategy based on age-eligibility and the start date of the policy to estimate its effects on youth employment outcomes in the first year of implementation. I find that even though the take-up rate of subsidies by firms was higher than expected, youth employment did

not improve. I discuss the potential structural factors responsible for the seeming failure of this policy.

In the second chapter, I explore the importance of ethnic capital, a proxy for social capital, in determining labor market outcomes for South Africa’s internal migrants. While the role of social networks for the labor market outcomes of international migrants has been extensively studied, internal migrants have not received the same attention. Arguably, internal migration is an important phenomenon for developing countries. I look at how ethnic capital affects the employment probability and the occupation choice of South African internal migrants. I find that the employment probability of an internal migrant is driven by the education distribution of his ethnic group, even after accounting for local labor market conditions and unobserved ethnic group effects. I also find that occupation choice is related to the education distribution and the gender composition of the ethnic group at the destination.

In the third chapter, coauthored with Angelino Viceisza and Tanguy Bernard, we study how communication can be improved to solve coordination issues within agricultural cooperatives, using both a lab-in-the-field experiment (LFE) and a randomized controlled trial (RCT). Theory and conventional lab experiments suggest that non-binding pre-play communication can enhance coordination of interdependent agents in situations of strategic uncertainty, and that coordination should be easier to achieve in smaller groups. We test for these hypotheses in the context of Senegalese agricultural cooperatives which seek, but mostly fail, to jointly sell their outputs in order to secure higher unit prices. We show that revealing farmers’ intended sales yielded enhanced coordination in larger groups where coordination is initially more difficult, and to higher income for small-scale farmers. We also show that participation in the LFE affected future behavior in the RCT.