Modeling investment uncertainty in the costs of global carbon dioxide emission policy.

Increased concentrations of atmospheric carbon dioxide threaten the stability of terrestrial weather patterns and have led many people to conclude that governmental intervention is necessary to induce the world's economy to rely less upon carbon based fossil fuels. This dissertation investigates how the United States economy will evolve during the next 110 years if policies are implemented which limit the production of carbon dioxide as a by-product of economic activity. It investigates this question using a stochastic economic model and develops new bounding techniques to derive solutions to the model. Many deterministic macroeconomic models have been developed to try to determine what specific policies are necessary to cause the United States economy to shift away from fossil fuel use as well as what the cost of such measures are in terms of lost economic output. This dissertation develops a nonlinear stochastic macroeconomic model with continuously distributed random variables that determines optimal policy while explicitly modeling the uncertainty inherent in new technology development and exhaustible resource exploration. It shows that policy derived using this stochastic model leads to significantly better economic performance than non-optimal policy derived under the assumption of a certain future with known returns to investment. Because the stochastic model with continuously distributed random variables is very difficult to solve, this dissertation bounds the true solution. It uses a parallel implementation of a nonlinear decomposition algorithm to obtain an upper bound and develops a new lower bounding method based on the linear response function technique for stochastic linear programs to derive a lower bound. The quality of the resulting bounds is explored as well as the benefits of parallelization to improve solution times.PhDIndustrial and Operations EngineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/103837/1/9409794.pdfDescription of 9409794.pdf : Restricted to UM users only