The revolving door and worker flows in banking regulation

Abstract

This paper traces career transitions of federal and state U.S. banking regulators from a large sample of publicly available curricula vitae, and provides basic facts on worker flows between the regulatory and private sector resulting from the revolving door. We find strong countercyclical net worker flows into regulatory jobs, driven largely by higher gross outflows into the private sector during booms. These worker flows are also driven by state-specific banking conditions as measured by local banks’ profitability, asset quality and failure rates. The regulatory sector seems to experience a retention challenge over time, with shorter regulatory spells for workers, and especially those with higher education. Evidence from cross-state enforcement actions of regulators shows gross inflows into regulation and gross outflows from regulation are both higher during periods of intense enforcement, though gross outflows are significantly smaller in magnitude. These results appear inconsistent with a “quid-pro-quo” explanation of the revolving door, but consistent with a “regulatory schooling” hypothesis.

Introduction

According to a prominent narrative of the recent financial crisis, weakness in banking regulatory oversight and, in particular, regulatory capture was a key contributing factor in the buildup of risk ahead of the crisis.¹ In conjecturing this link, a much talked about source of regulatory capture is the revolving door of regulatory personnel between the official and private sector that allegedly incentivizes regulatory personnel to soften their regulatory stance due to the prospect of future lucrative employment in the private sector.² Future job opportunities may, on the contrary, incentivize regulators to toughen their oversight and favor complex regulations as these might enhance the subsequent value of regulators, should they transition to the private sector. Other commentators have downplayed distortions of the revolving door in either direction and presented a more benign viewpoint of revolving doors as a means for regulatory agencies to attract higher ability and skilled workers. Despite the continuing debate on this issue and numerous policy prescriptions, little systematic evidence exists on the incidence and drivers of the revolving door — i.e., worker transitions between the banking regulatory and private sectors.

In this paper we attempt to fill this gap with two goals. First, we provide basic facts on worker flows between the banking regulatory and private sectors by examining worker transitions resulting from hirings and separations. Second, we characterize these flows in terms of their business cycle variation, job creation and destruction in the two sectors (Davis et al., 2006) and worker characteristics such as human capital and seniority. As well, we study worker flows as a function of intensity of regulatory activity to assess if worker flows in the aggregate can shed light on different views of revolving door.

The main obstacle in studying worker flows between the regulatory and private sector is a lack of available data. A large literature has studied direct job-to-job flows (Blanchard and Diamond, 1990, Fallick and Fleischman, 2004), and more closely to this paper, inter-industry job transitions (for example, Artuç et al., 2010) with longitudinal and cross-sectional data surveys. However, because banking regulators account for only a tiny fraction of all US workers, none of these sample surveys is detailed enough to construct worker flow statistics for the regulatory sector.³ We circumvent this challenge by constructing a unique dataset of career paths of more than 35,000 former and current regulators across all regulators of commercial banks and thrifts — the Federal Reserve Banks (Fed), the Federal Depository Insurance Corporation (FDIC), the Office of Comptroller and Currency (OCC), the Office of Thrift Supervision (OTS), and state banking regulators — that have posted their curricula vitae (CVs) on a major professional networking website. Our sample spans the past 25 years and provides a unique view into the process of selection and transition of personnel from these regulatory agencies to the private sector. After detailing our approach to construct worker transitions, which is new in the literature, and contrasting its shortcomings and strengths to labor data surveys, we study worker flows both in the aggregate and using panel regressions at the worker level.

Our data reveal clear evidence of countercyclical net worker inflows into the regulatory sector. Net worker flows from the private sector to the regulatory sector fall significantly and are often negative in good times, which may be the result of higher gross inflows to the regulatory sector in bad times or higher gross outflows in good times. Looking at the data, we find that higher gross outflows to the private sector in good times are a key driver of the countercyclical regulator net worker flows. We also find evidence of higher gross inflows in regulation in bad times as well as in the past few years, likely due to strong regulatory demand linked to enhanced banking supervision following the financial turmoil.

We investigate in more detail the potential sources of inflows and outflows from regulatory jobs by assessing the relationship of worker flows with relative demand-side proxies for job-creation or destruction in the banking and regulatory sector. Job creation in the regulatory as compared to the banking sector is likely to be positively related to measures of banking stress and inversely related to bank profitability, with opposite patterns expected for job destruction. Indeed, after controlling for aggregate economic conditions, net regulatory inflows tend to be higher in periods and states where local banks have lower ROAs, higher non-performing loan ratios, as measured from Call Reports, and higher occurrences of bank failures. Overall, we find regulatory job mobility to be much lower than in the private sector. On average, gross-worker flows between the regulatory and private sector is less than half the job-to-job transitions in the private sector as measured from a benchmark sample we employ and data from the Current Population Survey (CPS).⁴

Next, using information obtained from the workers’ CVs and the fact that our dataset is longitudinal, we examine the selection of individuals into and out of the regulatory sector by assessing which individuals enter and exit banking regulation over the business cycle, as a function of their human capital (i.e., their education levels) and skills/connections (i.e., their seniority in the regulatory organization). We find that the best talent, as proxied by higher human capital, has shorter regulatory spells because of higher outflows to the private sector. We also find that more senior staff, not surprisingly, spend more time in regulation. While we find no significant differences in the business cycle sensitivity across different human capital levels, overall, the regulatory sector appears to face a retention challenge when it comes to individuals with higher human capital based on their shorter regulatory spells. Consistent with this finding, we also find that regulatory spells have been declining in the past 25 years. For example, while about 88% of workers that started working in regulation in 1988 spent three or more years in regulation, only 64% of workers that started working did so two decades later.

We briefly turn next to an analysis that explores aggregate regulator mobility as a function of regulatory actions. According to one prominent view (the “quid-pro-quo” view), future employment opportunities in the private sector affects, as a quid pro quo, the strictness of actions of a regulator while the individual is employed in the regulatory sector. This hypothesis implies that we should observe lower gross outflows from regulation to private banking during periods of high enforcement activity. Understanding whether this hypothesis may be at play is important since destabilizing banking crises have often been ascribed to weak regulatory oversight. Based on an alternative view of the revolving door (the “regulatory schooling” view) regulators may instead have an incentive to favor complex rules because “schooling” in these regulations enhance regulators future earnings, should they transition to the private sector.⁵ Such a hypothesis would imply high gross inflows into regulation at times of higher regulatory intensity as workers are schooled in the new rules as well as high gross outflows from regulation into the private sector as regulators earn returns from schooling in the new rules. According to the regulatory schooling view, inefficiencies may derive not from laxity, but on the contrary from more complex regulations.

In order to shed light on these hypotheses, we relate worker flows to intensity of supervisory activity in the data. We measure each regulator’s strictness using their formal enforcement activity in a given state and year and focus on the most severe of these actions: terminations or suspensions of deposit insurance as well as cease-and-desist orders and prompt corrective action directives, for which failure to comply are all grounds for receivership. While enforcement activity is not a measure of regulatory complexity per se, as discussed in Agarwal et al. (2014) the final impact of rules on regulated entities is the result of both regulations and the manner in which these are enforced through supervisory activity. Empirically, we find a positive association between the intensity of strict actions over time and across states and the net inflows into the regulatory sector. Looking more closely at gross flows, we find that this relationship is driven by more inflows into regulatory jobs in periods of high enforcement/more intense regulatory activity. Gross outflows from the regulatory sector are, in fact, higher around periods of higher enforcement activity. We find similar patterns when we focus on cross-state variation only by including time fixed effects. This implies that the association between regulator activity and worker flows is identified even when we use variation across states and regulators and is not just capturing aggregate economic conditions. Based on the discussion above, the patterns we uncover are opposite of what would be implied by a quid pro quo story but are instead consistent with the regulatory schooling story. These patterns are consistent with Agarwal et al. (2014) who find higher turnover of state regulators into private sector if these regulators give harsher supervisory ratings to banks under their supervision.

This evidence is admittedly naïve to some extent and not a ubiquitous test of the quid-pro-quo or regulatory schooling hypothesis. For instance, it cannot rule out the presence of distortions in banking regulations at a more microlevel or whether regulations are more complex as a result of the schooling hypothesis. Uncovering whether revolving doors distorts regulatory effectiveness at a smaller scale or leads to regulators favoring excessive regulatory complexity remain open questions requiring more research. A recent paper by Shive and Forster (2014) takes a promising step in this direction.

This paper is most directly related to work studying revolving doors in banking regulation. Most of this work (for example, Johnson and Kwak, 2010) is normative in nature but relies on case studies or other anecdotal evidence. Our analysis, which relies on a new data collection, is instead mainly positive, and we therefore see it as complementary the existing work.⁶ More generally this paper is also connected to a vast literature in both political economy and banking on the role of regulation and its potential distortions (for examples, specific to banking: Kroszner and Strahan, 1999, Berger and Hannan, 1998, Barth et al., 2004, Agarwal et al., 2014, and more in general: Besley and Coate, 2003). Finally, our work is related to the labor literature on job-to-job transitions and intersectoral job transitions, mentioned above, and to the modern approach to labor flows (Davis et al., 2006) from which we borrow methods and definitions, while proposing a new approach to computing worker flows.