Theories of agency behavior are tested via the empirical application of hydropower project relicensing by the Federal Energy Regulatory Commission (FERC). In the relicensing of each project, fish and wildlife afencies make formal recommendations to FERC on fish and wildlife protection at the project. FERC then enters a two-stage deliberation process during which it accepts; amends then accepts; or rejects each recommendation. We apply a count data model to explain the number of recommendations made per project and a bivariate probit model to explain FERC's disposition of the recommendations. The analysis covers 933 fish and wildlife recommendations made for 72 projects relicensed during 1980-96. Independent variables for the benefits and costs of the recommendations cannot be constructed because of FERC's deficient generation of economic information.
A background section thus details the more serious limitations of FERC's application of economic analysis. In explaining outcomes, a tension exists between the hierarchical control of Congress or the executive branch (the Progressive Reform model of agency behavior) and an agency's bureaucratic resistance to change (an evolutionary model of agency behavior). Key findings indicate that a new law (the Electric Consumers Protection Act of 1986) substantially altered FERC's fish and wildlife decisions, while a new administration (the Clinton administration, beginning 1993) exerted a mixed effect. Both events influenced the fish and wildlife agencies, as the number of recommendations made, per project, increased significantly at these junctures.