Essays on economics of invasive species outbreaks and spatial information theoretic estimation

Invasive species are non-indigenous species that invade and adversely affect their adopted environment. Their outbreaks cause economic damage and raise concerns among consumers and producers of the affected commodities. Governments provide inspection services to prevent introduction of invasive species and enforce eradication policies in the event of an outbreak. This research shows why mitigation of invasive species outbreaks is difficult and provides policy recommendations on how to design optimal mitigation programs. Two theoretical models of preventative and eradication decisions are presented and applied to livestock disease outbreaks. Additionally, an alternative spatial econometric procedure for the first-order spatial autoregressive model is proposed. The first paper analyzes preventative decisions across countries and shows that spatial spillovers undermine the success of national mitigation programs. A spatial game theoretic framework is applied to compare preventative decisions under the centralized and decentralized decision models. The results show that in the presence of spatial spillovers, the probability of an outbreak is smaller and the expected payoffs are greater under the centralized model. Regional cooperation and monetary transfers could be required to achieve a socially optimal outcome. The model is applied to investigate the epidemics of foot-and-mouth disease in South America. The second paper analyzes eradication decisions for a single country and proposes an optimal control model under uncertainty of a potential entry and spread of foot-and-mouth disease in the livestock sector. Dynamic livestock production, disease dissemination, domestic beef consumption, and international trade are explicitly integrated. Intertemporal welfare estimates are used to evaluate alternative eradication policies. The model is applied to analyze the disease outbreak in Canada, where it is found that the total welfare losses from an outbreak decrease as eradication efforts increase. The third paper proposes an alternative spatial econometric procedure for the first-order spatial autoregressive model. Extensive Monte Carlo experiments are used to compare traditional and three information theoretic estimators. It is found that information theoretic estimators are robust to the specification of spatial autocorrelation and outperform traditional moment based estimators in specific finite sample situations