The effect of regulation on utility decisions about generation

This dissertation empirically investigates the effects of state regulation on utility decisions about generation. First, we examine the relative merits of bidding and negotiation, the two most common methods used to award long-term power purchase contracts to third party generators. Together, auction theory and the transaction cost literature suggest that when information predicting supplier performance is accessible, bidding should be the preferred procurement method. We test this hypothesis using data from power purchase contracts between utilities and independent generators. Results indicate that bidding is generally more cost-effective than negotiation but that utilities should be allowed to negotiate with unproven suppliers that win competitive solicitations.;Second, we provide the first statistical analysis of the effect of regulatory policy on utility decisions to own or buy new capacity. Two inequalities in the treatment of utility owned plant and third party generation are likely to bias these choices. First, returns allowed on purchased power and utility owned plant differ because purchased power is excluded from the utility's rate base. Second, in many states, purchased power receives more comprehensive treatment under automatic adjustment clauses than does the utility's own generation. Empirical results demonstrate that these asymmetric rules influence input choices and highlight the importance of providing utilities with incentives to make efficient investment decisions.;Third, we address an important methodological problem in the empirical literature investigating the effects of regulation on utility behavior. Previous studies do not adequately capture differences in state regulation. First, studies that examine cost effects of individual regulatory practices do not control for other policies affecting utility behavior. Second, studies analyzing the effect of profit restrictions on utility decisions use endogenous or incomplete proxies for the profit constraint. It is often infeasible to include all policies comprising the utility's regulatory environment in estimation. Hence, we test whether regulatory climate, a regulatory policy index commonly used in the finance literature, captures quantitative differences in state regulation better than a single policy. Results show regulatory climate provides a serviceable summary measure of policy and that equations containing a single policy suffer from omitted variable bias.