UC Berkeley

This manuscript is comprised of three independent essays in applied microeconomics. Each essay examines a distinct challenge or trend impacting African youth. First, I look at the sports betting industry which has exploded across many African countries over the past decade, targeting young men as its primary consumers. In Kampala, Uganda, the site of my study, nearly one in three young men age 18 to 40 participate in sports betting, spending 8-12% of their weekly income on these activities. The second chapter presents results from an impact evaluation of a sports and youth group program in Monrovia, Liberia. The use of sports and youth programs has become extremely popular for many multilateral and international non-profit organizations' efforts to engage and improve the conditions of marginalized or at-risk youth in developing countries. The final chapter looks at the impact of rainfall shocks on marriage decisions among youth from Burkina Faso, a setting where marriages are determined by households and influenced by financial constraints and economic shocks.

In the first essay, "Betting, Saving, and Lumpy Expenditures: Sports Betting in Uganda'', I present evidence from my field study on one of the causes of high betting demand among young Ugandan men. In my paper, I show that financial constraints push many men towards betting in the hopes of payouts desired for unmet liquidity needs. I use a range of experimental and quasi-experimental methods and present four pieces of evidence in support of this claim. Winnings from betting increase the size and likelihood of making large, lumpy purchases, with strongest effects among those with limited ability to save. I then use a randomized intervention to improve an alternative strategy of liquidity generation to betting by distributing a basic commitment savings technology and find that it lowers recipients' demand for betting. Next, I use an experimental prime or nudge, increasing the salience of a desired large expenditure, and find that it increases demand for betting tickets, particularly among those with low ability to save. And finally, I use a randomized budgeting exercise and find that it lowers demand for betting among those who learn that their ability to save is better than previously believed. For people whose ability to save and borrow is inhibited or costly, betting provides an enticing alternative way to generate meaningful sums of liquidity despite imposing considerable losses.

My second essay, "Do Sports Change Lives? Evidence from a Randomized Control Trial'', is the product of a joint project with Lori Beaman, Niall Keleher, and Jeremy Magruder. In this project we assess the impact of a popular form of international development program that uses sports both as a direct intervention and also as a point of entry to facilitate engagement with vulnerable youth. The stated ambitions of sports for development programs are typically both admirable and lofty, but we find only limited evidence that the impacts of these programs match their promises. We find evidence of some modest impacts on psychosocial behaviors for young men with moderate improvements in measures of self-esteem and aggressive behavior. We also see an increase in labor force participation for both men and women, although earnings among those working remain unchanged. We also explore whether the research structure required by randomization may have hindered program efficacy. However, it appears that program effects were likely stronger among late registrants who are more likely to have been excluded in the absence of the study. And finally, permitting endogenously formed sports groups may have increased pre-existing social network connections on sports teams. We find that presence of friends does impact program attendance but does not significantly impact program outcomes.

In my third essay, "Marriage Markets and Rainfall Shocks: Evidence from Burkina Faso'', I analyze the impact of rainfall shocks on marriage outcomes for young men and women in Burkina Faso. In particular, I use quasi-random variation in rainfall to provide evidence that low rainfall from two years ago causes an increase in the likelihood of marriage for young women by 15%. These effects are strongest among women aged 13-16. In addition, I present evidence that closed marriage markets, where womens' partners are most likely to come from the same local geographic area, respond less to rainfall shocks than areas where marriage markets are likely to be more geographically integrated. This finding is consistent with basic trade theory on market structure and response to simultaneous shifts of supply and demand in opposite directions.