In the United States today, political bureaucracies make the overwhelming majority of public policy decisions. This is not to deny the importance of policy-making by the other branches of government, yet the sheer amount of policy-making done by the bureaucracy dwarfs the output of these other institutions. To cite just one measure, the Federal Register, which lists all rules and regulations issued by the bureaucracy, usually covers more than 50,000 pages per year.
In this investigation, the researcher uses a formal model to derive testable propositions about the influence of other political actors on the bureaucracy. He develops a spatial model of political influence that is more explicit in its assumptions and predictions than are previous studies of the relationship between bureaucrats and other political actors. In addition, the theory highlights substantial problems with previous methodological approaches and introduces a more appropriate approach for estimating political influence, one, which acknowledges that different political regimes will have different effects on agencies.
The investigator provides preliminary support of the model by analyzing the monitoring activities of the Food and Drug Administration. He then collects data on a variety of other agencies. After collecting the data, he estimates the influence on bureaucracies by using approaches he has developed. To allow for comparison of different types of agencies, a variety of agencies are examined, including the Environmental Protection Agency, the Federal Trade Commission, the Food and Drug Administration, the Food Safety Information Service, and others.
The investigation makes several significant contributions to an understanding of the role of bureaucracies in a representative democracy. First, it develops a rigorous theoretical model that can be tested. Second, it points to problems with previous empirical tests of political influence that do not account for the existence of different regimes. Third, it proposes a methodological approach that follows from the theoretical model. Finally, through the estimation and testing of the model it gives political scientists a more accurate picture of the relationship between agencies and other political actors.