Chain of Command v. Who’s in Command: Structure, Politics and Regulatory Outputs

• In Event: The Politics of Regulatory Rule Making

Abstract

At the outset of new political administrations, there is immense intrigue about who gets appointed to what administrative agency. Supporters typically view the political appointment process as an opportunity to consolidate an election victory into new policy initiatives, whereas opponents perceive it as creating a risk of undoing past policy achievements. To be sure who leads an agency has consequential influence over not only day-to-day agency activity but also on long-run strategic planning and important societal outcomes. Newly elected political coalitions seek policy influence by manipulating agency budgets and staffing (Carpenter 1996; McCubbins and Schwartz 1984; Weingast and Moran 1983; Weingast 1984). But these are not the only avenues through which politics may shape policy. Administrative structures and procedures, that at first glance appear rather innocuous, can nevertheless shape policy delivery in meaningful ways (McNollGast 1987; 1989).

Indeed, there is abundant anecdotal evidence that politicians manipulate structure and process with an eye toward shaping policy. In the empirical context of this paper - environmental enforcement - three recent anecdotes are particularly noteworthy. In 2013, in an attempt to bring its environmental agency more in line with their preferences, North Carolina’s Governor expanded the number of agency employees exempt from civil service protections from 24 to 179 (Gabriel 2014). In another example, in 2011, Tennessee’s Commissioner of the Department of Environment and Conservation merged three water protection divisions into one, centralizing the process and empowering the department political leadership’s role in enforcement (Reicher 2016). Most recently, in April of 2018, the EPA’s Assistant Administrator for Enforcement and Compliance Assurance issued memorandums that interjected herself as an intermediary between EPA regional offices and the Department of Justice on civil enforcement cases (Grandoni 2018). Each of these cases highlights the attempted use of bureaucratic procedure to shape policy delivery by adding a veto player and altering where within an agency the line between political appointees and civil servants falls.

But when a political appointee is added to an agency’s command structure – two features change simultaneously: a political appointee rather than a civil servant is inserted into the agency’s decision-making stream and an additional layer of bureaucracy, another veto player, is added. If we were to observe, as a consequence of this institutional change, a shift in agency output, it is not clear what inference we should draw. Is the agency’s output shift due to the addition of a political agent, the addition of another veto player, regardless of politics, or some combination of the two? After all, the two institutions are intimately intertwined. Political appointees tend toward the top of agency’s command structure but the top of the command structure also, by definition, has many offices beneath it – offices that typically have a role to play in the long line of signatures required to advance an enforcement action against a facility. So when a political appointee is added to an agency’s command structure is it more consequential that a political appointee has been added or that the total number of actors that weigh in on such actions has increased?

Yet, studying who’s in command v. chain of command is immensely challenging because of data constraints. In this paper, we leverage newly collected data that captures the rich institutional diversity of state pollution control offices. In particular, we examine whether and to what degree political appointees or the workflow of actions moving through an agency’s decision-making structure competes to insert politics into agency enforcement. Specifically, we test our hypotheses with an empirical exercise centered on regulation of about 16,000 major air pollution sources under the U.S. Clean Air Act between 2003 and 2012. We combine facility-level regulatory action data for all major stationary sources in the U.S. with novel data we have collected on institutional designs for state air pollution control regional offices and local air agencies. We generate a dataset on the internal institutions of all state and local air agencies with respect to the location of political appointees and the extent to which decision-making over enforcement decisions is centralized or decentralized (or somewhere in between).

Our preliminary results suggest that, while political appointees influence regulatory output in expected ways, centralized workflow, regardless of party, impedes output. These findings contribute literatures in political science, public administration, and environmental politics.

Authors

• David Konisky, Indiana University, Bloomington

• Christopher M. Reenock, Florida State University

• Matthew Joseph Uttermark, Florida State University