Farming under Weather Risk: Adaptation, Moral Hazard, and Selection on Moral Hazard

Farmers in the American Midwest decide on agricultural land use (cropping pattern) and crop insurance in springtime after observing pre-plant precipitation. We examine cropping-pattern adaptation to pre-plant precipitation as a natural experiment. In tandem with the weather experiment, we also exploit a quasi-experiment created by a federal program that sharply reduced insurance deductibles to examine both risk-taking in cropping pattern as a moral hazard of insurance and selection of insurance coverage in response to the risk-taking. Using a 2001–2014 panel of high-resolution spatial data on land use and weather, we present evidence of heterogeneous adaptation in cropping pattern across the large agricultural states of Illinois, Iowa, Nebraska, and North Dakota. We also find evidence of heterogeneous risk-taking in cropping pattern during the federal program in 2009–2011, with farmers in Nebraska and North Dakota much more responsive to pre-plant precipitation in both their adaptation and risk-taking than farmers in Illinois and Iowa. Using a 2001–2014 panel of county-level data on crop insurance expenditures, we find limited evidence of selection on moral hazard in insurance expenditures in response to pre-plant precipitation. Farmers in Illinois and Iowa increase (decrease) the rate of insurance expenditures on corn when they increase (decrease) corn acres. They do so to a lesser degree with soybeans. The interaction of adaptation, moral hazard, and selection on moral hazard provides new insight into incentives, hidden actions, and hidden information in major cropland and insurance markets.