Essays on environmental regulation and robust control
MetadataShow full item record
The present dissertation is divided into three essays. The first essay develops an analytical model with three goods: clean agricultural, clean manufacturing and dirty manufacturing. The production of agricultural goods has land as a fixed factor. The model is calibrated using data for Mexico. The introduction of environmental taxes and a tax cut in the manufacturing goods to keep government revenue unchanged increases welfare but at the cost of a more unequal distribution of income. The costs of the policy are mostly regressive. The regressivity of the policy is driven by tax cut on manufacturing rather than the environmental tax. The income effects from land for Mexico are relatively small. In the second essay an analytical framework using robust control was developed for the one-state variable, one-control variable model to examine the response of the control to changes in the “free” parameter. As this parameter decreases the uncertainty always increases. Nevertheless, more uncertainty does not always correspond to a more aggressive control variable response. It was found that the response is characterized by a hyperbolic shape. Such response presents a concave shape on the right side of the discontinuity and a convex one on the left side. A reasonable range for the free parameter was provided for the case of one-state, one-control model. In the last essay, I introduce model uncertainty into the regulation of a stock pollutant, as presented in Hoel and Karp (2001), and analyze its effects on the levels of pollution taxes and the pollution stock. Fundamental uncertainty about the decision-maker’s model is an important aspect of global warming policy, and of many other policy issues as well. To analyze this problem, I use robust control, a computational tool that allows the planner to deal with model uncertainty. I find that: 1) an increase in uncertainty about the model increases pollution taxes, while more uncertainty about abatement costs reduces taxes, and 2) the effect on taxes and the pollution stock of introducing model uncertainty is small for medium levels of robustness.