Relying on data gathered from surveys and interviews of top political officials in the U.S. EPA, Lisa Schultz Bressman and Mike Vandenbergh (both at Vanderbilt) have written Inside the Administrative State: A Critical Look at the Practice of Presidential Control. I saw this paper presentation at AALS in D.C. and it is well worth the read for those interested in presidential influence and accountability in agency decision-making. The Abstract:
From the inception of the administrative state, scholars have proposed various models of agency decision-making to render such decision-making accountable and effective, only to see those models falter when confronted by actual practice. Until now, the presidential control model has been largely impervious to this pattern. That model, which brings agency decision-making under the direction of the President, has strengthened over time, winning broad scholarly endorsement and bipartisan political support. But it, like prior models, relies on abstractions - for example, that the President represents public preferences and resists parochial pressures - that do not hold up as a factual matter. Although recent empirical analyses purport to validate the model, they fall short because they examine how the White House exercises control without considering how agencies experience control. This Article is the first to study the practice of presidential control from inside the administrative state. We interviewed the top political officials at the Environmental Protection Agency from the George H. W. Bush and Clinton Administrations during 1989-2001. Our data, which do not vary substantially between respondents of different presidential administrations, suggest that White House involvement is more complex and less positive than previous accounts acknowledge. But we do not conclude that the presidential control model lacks merit. Indeed, our respondents recognize that the President has a role to play in controlling agency decision-making. We therefore conclude that the presidential control model requires reworking to remain valid in practice as well as in theory. We identify next steps in that direction.