I study the effect of political transitions on regulatory policymaking. I start with a simple model of dynamic policymaking, where regulators (and their overseers) make regulatory decisions based on expectations about the probability and consequences of a political transition. When transitions are likely, political actors may pay a cost to "lock in" their policies. When transitions are less likely, such investments are typically not worth it. I analyze the regulatory activity of federal agencies from 1984 to 2012 and find differences in presidential and congressional transitions. Presidential transitions resemble the managed transition equilibrium. Controversial regulatory proposals are locked in prior to transitions. Following transitions, proposals inherited by the new administration appear non-controversial and are unlikely to be withdrawn. By contrast, congressional transitions resemble the unmanaged equilibrium. The transitions are not preceded by a similar lock in and they are followed by an increase in withdrawn proposals. The withdrawals also appear to be politically motivated. Stepping back, the analysis has implications for understanding the efficiency and responsiveness of policymaking in the administrative state.

Regulatory agencies are often in conflict with the White House. A key reason why is that agencies develop policies that reflect the preferences of career bureaucrats, not just political appointees. One approach to inferring agency preferences is through their stated preferences, such as by surveying employees. Stated preferences can be insightful, but typically struggle with how to weight the relative influence of appointees and careerists. I propose a revealed preference---or behavioral---approach that sidesteps this challenge. My focus is the White House's regulatory review program. Agencies make proposals that the White House can review, or audit. My theoretical model clarifies conditions under which White House auditing reflects policy disagreement between presidents and agencies. Intuitively, proposals from "adversaries" are audited more frequently than proposals from "allies." Using data on the behavior of presidents (their strategic audits) and agencies (their strategic proposals), I statistically estimate the auditing bias presidents exhibit toward individual agencies. The resulting audit scores reveal policy disagreement between presidents and agencies across time.

In response to the growth of the administrative state, presidents have developed a system of centralized review to selectively audit regulatory actions for political scrutiny. How does centralized review make agencies more responsive to presidential priorities? Clearly centralized review allows presidents to single out non-compliant regulations for revision or termination. More consequential, however, might be how centralized review shapes agency behavior in anticipation of an audit---the deterrence effect. In this paper, I develop a simple theory to show how the threat of auditing should induce agencies to invest in accommodating presidential preferences, regardless of whether a particular regulatory action is eventually audited. Across nearly two hundred regulatory agencies and three presidential administrations, I show how patterns of targeting and deterrence have moved in unison, demonstrating the pervasive influence of centralized review in holding agencies accountable to presidential priorities.

The Office of Information and Regulatory Affairs (OIRA) is the locus for centralized presidential efforts to control agency rulemaking, predominately through the process of selectively auditing and then ordering the revision of regulatory proposals. We begin by modeling OIRA's targeting decisions, analyzing two distinctly different models. Our first approach centers the ideological conflict in rulemaking between conservative under-regulators and liberal over-regulators. Under this framework, Republican administrations target large proposals from overzealous liberal agencies and Democratic administrations focus on scaling up modest proposals from conservative agencies. Our second approach reorients the ideological conflict to the familiar spatial model, whereby OIRA is used to limit policy damage from ideologically distant agencies by scaling down their more ambitious proposals. To test our theoretical predictions, we conducted the first systematic analysis of OIRA's targeting decision, using data across three presidential administrations and over 30 regulatory agencies. Our analysis reveals that Republican OIRAs disproportionately targeted large proposals from liberal agencies, a finding compatible with both theoretical models. Democratic OIRAs, on the other hand, did not disproportionately audit conservative agencies, either by targeting under-regulation (per our extension) or the more ambitious proposals (per the baseline model). Rather, they targeted large regulations with apparent disregard for ideology. We speculate what objective (if any) might lead Democratic administrations to rationally employ such a regulatory auditing strategy.

Presidents rely on two core strategies for controlling administrative agencies: a decentralized strategy to staff agencies with political appointees and a centralized strategy to manage agencies from the Executive Office of the President, such as by reviewing, changing and vetoing their regulatory proposals. How do these two strategies interact? In particular, are they substitutes or complements? I develop a theoretical model of policymaking to show how centralization and decentralization act as complements. The model highlights the role of appointees in signaling agency activity to the president, as well as the challenges faced by appointees---loyal or otherwise---in anticipating the preference of president. Appointees must balance signaling too much (Type I errors) with not signaling often enough (Type II errors). The key proposition finds that centralization (presidential review of agency regulations) and decentralization (politicization of agencies with appointees) should be positively correlated. I find empirical support for the proposition by looking at patterns of presidential review and politicization across two presidential administrations and over a hundred regulatory agencies.

Why are statutes that delegate lawmaking authority to federal agencies typically passed by near-majority coalitions with the support of the president? More specifically, why does this pattern hold for policy areas that tend to evoke partisan discord? In this paper, I present a new theory of delegation that is derived from vote-buying models of legislative behavior in order to highlight the incentives that lawmakers have to cede lawmaking authority to agencies. The theory presents delegation as a means of creating a market for agency oversight activities, which allows politicians to capture more rents from interest groups. The theory is distinct from the two dominant, though not mutually incompatible, explanations in the literature for why elected officials delegate policymaking authority to unelected bureaucrats: policy insulation and expertise. The goal of the paper is not to invalidad these theories, but rather to provide a richer menu of theoretical possibilities from which to explain delegation, particularly in the context of agency lawmaking (rulemaking). Critically, the model highlights the tension between the legislature and executive over the control of interest group rents and, relatedly, provides predictions about whether delegation should transfer policymaking authority to an independent agency or an executive branch agency.