Self-insurance and self-protection in weed control: implications for nonpoint source pollution

This dissertation investigates producer's weed control decisions in three papers. Using an endogenous risk framework, the papers model weed control decisions as they are affected by uncertainty and as they are affected by policies aimed at reducing nonpoint source pollution;The first paper adapts the theory of self protection and self insurance to weed control decisions in order to model substitutions between fertilizer and herbicides. Our results suggest that an increase in the probability of weed damage will generally decrease both herbicide and fertilizer application rates. We also show that increased probability of weed damage leads to use of less persistent herbicides and less fertilizer. Also, nonpoint pollution policies which restrict fertilizer quantities will tend to reduce herbicide rates and herbicide persistence, while policies which restrict herbicide rates or herbicide persistence will tend to reduce fertilizer rates;The second paper models weed control decisions using the concepts of self insurance and self protection to model substitutions among the types and quantities of herbicides used. Our results indicate that policies which restrict herbicide rates likely lead to use of more persistent herbicides, while policies which restrict the use of persistent herbicides likely lead to increased application rates. We also provide conditions to unambiguously sign the effects of increased application or effectiveness risk on optimal herbicide choices. Generally, risk leads to a tradeoff between herbicide rates and herbicide persistence;The third paper uses a biophysical modeling approach to construct and evaluate the environmental and economic effects of alternative nonpoint source pollution policies. Environmental and economic tradeoffs are compared for six sets of policies including five sets of tax policies and one set of bans. Generally, herbicide targeted tax policies which are based on groundwater exposure values are most cost-efficient for producing relatively small improvements in water quality, while broad-based policies such as flat taxes and bans are most cost-efficient for producing larger improvements in water quality.