UC Berkeley

This dissertation looks at the effect of participating in crop and livestock insurance on many aspects of farmers' business decision making. In Chapter 2, we investigate the effect of crop insurance enrollment on contract terms and farmers' participation in marketing contracts. Our model shows that improved terms of crop insurance (lower premium, higher subsidy) make contracts less appealing to farmers as mechanisms to mitigate risk. Therefore, intermediaries may revise their contract offers so that they are more attractive. However, improvements in contract terms are limited by their cost to the intermediaries, and will not lead to expanded participation in contracts. Thus, improved insurance terms may not crowd out participation in contracts, but will not expand participation either. In Chapter 3, we use various data sources and 2SLS to empirically test the predictions out of the model we developed in Chapter 2. Specifically, we look at how crop insurance enrollment affects farmers' share of value of production under marketing contracts. We construct a pseudo panel out of the repeated observations from the agricultural and resource management survey (ARMS) data. We use lagged county level subsidy and total precipitation as IVs for insurance enrollment. The second stage result indicates that, in peanut sector, farms with higher possibility of insurance enrollment will have more share of value of production under marketing contracts as well. In Chapter 4, we explore the relationship between farmers' use of livestock insurance and debt use. Risk balancing hypothesis (RBH) suggests that an exogenous reduction in farm's business risk would result in farmers taking more financial risk. We utilize a natural experiment from Jiyuan city in Henan province, China and regression discontinuity design approach to test the RBH. Our empirical results suggest that, farmers that are selected into the compulsory hog insurance program are 20 percent more likely to be willing to take debt than those who are not. Meanwhile, the difference of probability of taking debt between the two groups is insignificant. The combination of the two results indicates that farmers are willing to take more debt after insurance is available. However, the presence of credit constraint prevents them from taking more debt.