Summary: | Risk is central in the study of rural development. To cope with risk, smallholder farmers rely on a range of formal and informal insurance mechanisms: an extensive literature has explored their interactions. Yet, our understanding of the implications of these interactions for smallholder farmers’ decision-making is incomplete. This thesis addresses this scholarly gap by shedding new light on the risk-related decisions of smallholder farmers and the mechanisms through which networks affect these decisions. To do so, it relies on a combination of experimental and non-experimental economic analyses. The first chapter draws on a framed field experiment in Gujarat, India to explore the effect of selling weather index insurance to groups (as opposed to individuals) on the investment decisions of the insured. The analysis reveals that group pressure reduces risk taking among individuals with group insurance in contexts with perfect information about peer investment decisions. Group insurance thus suffers from the same potential pitfall as group microcredit. The second chapter examines the extent to which informal transfers can explain take-up of individual weather index insurance. It aims to disentangle two channels through which informal transfers influence decisions to purchase insurance: (i) informal risk sharing and (ii) moral hazard. As in the first chapter, the study draws on a framed field experiment in Gujarat. The main finding of this experiment is that redistribution norms reduce take-up: moral hazard leads to lower levels of insurance coverage. The final chapter builds on these results with a nonexperimental analysis of panel data from a rural household survey in India. It examines how cultural obligations to redistribute within networks affect investments in selfprotection. The empirical evidence suggests that increases in individual income lead to higher investments, but increases in network income lead to lower investments due to moral hazard. Collectively, the three papers nuance our understanding of how redistribution norms affect the risk-related decisions of the rural poor. While not negating their consumption smoothing benefits, this thesis indicates that networks also affect decision-making via group pressure and moral hazard. Such externalities could be forestalled by targeting insurance in rural areas with weaker redistributive norms or modifying insurance policy designs. Further research on the welfare implications of such approaches is thus recommended.
|