Scholar seeks ways to balance legislative, bureaucratic power

Legislatures that make laws and the bureaucrats who execute them: it is a fundamentally complicated relationship in our democracy, which sometimes works harmoniously and sometimes is hopelessly stormy. Sean Gailmard, Assistant Professor in the Harris Graduate School of Public Policy Studies, has focused an aspect of his current scholarship on using information economics to better understand the legislative-bureaucratic interaction to find ways to improve the design of public bureaucracies.

How power should be balanced between legislatures and bureaucrats is an enduring question in policy-making, Gailmard said. “Since policy-making in advanced democracies is fundamentally about these sorts of linkages between institutions, this question is not only enduring but important.”

Gailmard set out to create a research model in which a bureaucrat performs a project for multiple legislative principals. His analytical model took into account that a legislature need not be passive in its relations with bureaucrats, that bureaucrats typically know more about what they do than legislators, and that legislators can use oversight to obtain outside information and limit the inevitable deficiencies in their knowledge of an agency in question.

“Bureaucrats care about variables besides their budgets,” Gailmard said. “They care about their reputations, their relationships with superiors and the stability of the bureaucratic structure in which they operate.”

He explained that focusing on budgets as the primary lever legislators can pull is not always appropriate. “And perhaps most important, a bureaucrat (the agent) usually faces not just one legislative overseer (the principal) but many. At the federal level, we have bicameral legislative bodies with the U.S. House and Senate, specific committees that handle the appropriations authorization process and other committees that monitor all or part of the agency’s ongoing operations.”

His research model recognizes that a single common agent being influenced by multiple principals is subject to their overt attempts to control policy choices, but also takes hidden actions and processes hidden information because of its specialized expertise. “My model showed that these multiple sources of oversight dilute legislative control over bureaucracy,” Gailmard said. “Conflict among legislators that bureaucrats can ‘play off’ against each other is a commonly cited reason for that. But in my model, it isn’t necessary: the loss of control is not due to conflicts among the sources of oversight; it is due instead to the problems created by collective action where control and oversight authority are diffused.”

The conclusion drawn from this is that the institutional structure of the overseeing bodies is just as important as the structure of the bureaucracy being overseen, said Gailmard. His analysis revealed that this diffused oversight–with multiple principals, each with its own ability to try to influence the agent, and oversee and gauge its performance–can sometimes ironically result in an inefficiently low measurement of performance overall.

Gailmard hopes his ongoing research can provide tools for policy-makers to reach more informed decisions about the critical need to design their oversight effectively. “And my model could also apply, with some modification, to regulation–bureaucrats themselves in the role of principals monitoring the firms over which they have oversight,” he concluded.