When markets are incomplete, cultural norms may play an important role in shaping economic behavior. In this paper, we explore whether income shocks increase the probability of child marriages in societies that engage in bride price payments -transfers from the groom to the bride's parents at marriage. We develop a simple model in which households are exposed to income volatility and have no access to credit markets. If a daughter marries, the household obtains a bride price and has fewer members to support. In this framework, girls have a higher probability of marrying early when their parents have higher marginal utility of consumption because of adverse income shocks. We test the prediction of the model by exploiting variation in rainfall shocks over a woman's life cycle, using a survey dataset from rural Tanzania. We find that adverse shocks during teenage years increase the probability of early marriages and early fertility among women.⇤ We thank Jérôme Adda, Marcella Alsan, Natalie Bau, Caterina Gennaioli, Veronica Guerrieri, Eliana La Ferrara, Costas Meghir, Claudia Olivetti, Michèle Tertilt, Marcos Vera-Hernandez and participants at the workshops at COSME, Barcelona GSE Summer Forum, SED meetings,"Families and the Macroeconomy" in Mannheim, EDePo, and AEA meetings, and at seminars at Stockholm School of Economics, Bocconi and Queen Mary for helpful comments. We are grateful to Helene Bie Lilleør and Sofya Krutikova for useful discussion on the Kagera Health Development Survey and to Kalle Hirvonen for the clarification regarding weather data. Simone Lenzu and Jorge Rodriguez provided outstanding research assistance. This research was partly funded by a Research Grant from the Social Sciences Division at the University of Chicago.† Queen Mary, University of London and IFS. Email: l.corno@qmul.ac.uk ‡ The University of Chicago and NBER. Email: avoena@uchicago.edu.
1The relationship is stronger in villages where bride price payments are typically higher. We use these empirical results to estimate the parameters of our model and isolate the role of the bride price custom for consumption smoothing. In counterfactual exercises, we show that parents heavily rely on child marriages and bride price payments to smooth consumption. Without credit markets, bans on these practices are costly for a daughter's parents. However, ensuring access to credit limits parents' cost, making bans more likely to succeed.