| Summary: | The thesis comprises three chapters on the microeconomics of development. The first chapter consists of a lab-in-field experiment in rural Pakistan. I test whether anticipated changes in individuals' liquidity constraints can cause individuals spuriously to appear "present-biased" or "future-biased". The design uses experimental windfalls to create fully exogenous variation in subjects' liquidity constraints. The results suggest that the effect of liquidity constraints on "present-bias" and "future-bias" in monetary tasks exists, is large, and does not operate via changing cognitive function or risk preferences. Importantly, and in contrast with other recent studies, I establish that the causation runs from tighter liquidity constraints to appearing "present-biased"ârather than truly present-biased individuals making choices which lead to tighter liquidity constraints. The findings have important implications for the measurement of time preferences, and for interventions offering commitment. The second chapter - co-authored with Marije Groot Bruinderink, Wendy Janssens and Karlijn Morsink - describes a field experiment in the slums of Matola, Mozambique. We model and test how intra-household bargaining affects adoption of dual-protection contraceptive technologies, when women place a greater weight than men on the health costs of unprotected sex. To do so, we evaluate an intervention which provides women with information about and free access to male and female condoms. We find strong impacts on adoption of female condoms, using both survey data and data from weekly coital diaries. As predicted by the model, take-up of female condoms is higher among women with low household bargaining power at baseline, including women who are unable to convince their partner to use male condoms. The third chapter â co-authored with Marcel Fafchamps â reports the findings from a survey of Village Savings and Loan Associations (VSLAs) in Malawi. We ask whether some individuals adopt VSLAs as a commitment savings technology, and whether they sort into groups with borrowers. Using dyadic regression analysis we find evidence of negative assortative matching on a measure of "present-bias", which suggests that this is the case. This may increase financial intermediation, and welfare, in villages with low access to formal banking.
|
|---|
