Do elections delay regulatory action?☆
Introduction
The theory of political business cycles predicts public officials will be proactive and use fiscal and monetary policy before elections to increase their popularity and thereby their likelihood of reelection (Nordhaus, 1975, MacRae, 1977). Recent research shows elections elicit a different response from public officials when dealing with failing banks, namely public officials delay acting against failing banks before elections (Brown and Dinç, 2005; Liu and Ngo, 2014). We investigate the effect of elections in the US property-liability (P/L) insurance industry and find that the probability of a regulatory intervention falls by 78% in the year leading up to an election. We also find evidence that suggests electoral delays greatly increase the cost of insurer failure.
An important issue in political economy is the accountability structures designed to select public officials. Most public officials are elected, but the heads of many financial regulatory agencies are appointed by politicians (e.g., the Federal Deposit Insurance Corporation (FDIC) and the Federal Reserve). Political economy theory is mixed on whether appointed or elected officials will be more or less insulated from the political process. Deno and Mehay, 1987, Baron, 1988, and Laffont (1996) argue that the method of selection of public officials should have no systematic effect on policy outcomes. In contrast, more recent literature argues that the method of selection matters. Besley and Coate (2003) and Maskin and Tirole (2004) argue elected regulators will be more consumer oriented, while appointed regulators will be more industry oriented, as they can be more insulated from the political process. We find the government structure of political systems affects the magnitude of electoral delays for the interventions of US P/L insurers. Specifically, elected regulators delay action against failing insurers before elections more than appointed regulators.
The US insurance industry provides an ideal laboratory to study the impact of electoral politics on the timing of financial institution interventions because the regulation of US insurers is performed exclusively by the states. Chartering, supervision, and resolution are all done at the state level. An insurer's domiciliary state serves as the primary solvency regulator. Each state has a head regulator, the insurance commissioner. Commissioners are elected by popular vote in some states and appointed by the governor in the others. For all the states but California, the mechanism to select an insurance commissioner was established well before our sample period, removing potential endogeneity in the form of self-section.1 The election dates of commissioners and governors are predetermined, so there is no potential endogeneity in the timing of the election relative to firm failures.2 Moreover, the elections are not nationally synchronized, so there is cross-sectional and time-series heterogeneity in the timing of elections.
We exploit this cross-sectional and time-series heterogeneity in the timing of elections to provide clean identification of the impact of electoral politics on the timing of insurer interventions. The exogenous nature of elections establishes causality (e.g., Atanassov et al., 2015, Jens, 2017). Insurers in states without an upcoming election act as the control group for a treated sample of firms in states about to elect an insurance commissioner, or a governor if the commissioner is appointed. The coefficient estimate on an upcoming election dummy captures changes in the likelihood of regulatory intervention between treated and control groups.
The identification strategy allows us to isolate the effects of elections on interventions from any other state- or year-level factors affecting interventions. The time-series variation compares the same insurers across time, as each insurer will be subject to recurring treatments. The cross-sectional variation contrasts different insurers at a given time, as the timing of the treatment varies across the states. While for any particular state it is possible to imagine that some unobserved factor relevant for regulatory interventions might change at the same time as an upcoming election, such confounding is unlikely across multiple states and time periods.
There are reasons unrelated to electoral politics that could explain cross-sectional variation in insurer interventions (e.g., some states have more aggressive regulators). Other explanations, unrelated to elections, could explain time-series variation in interventions (e.g., catastrophes or macroeconomic fluctuations). None of these reasons, however, could explain why we would observe a relationship before elections but not in nonelection years. Thus, in the absence of politics, the failure of US insurance companies should not exhibit electoral cycles.
To study whether political incentives influence the behavior of insurance regulators, we combine detailed company-level data and failure data from 1989 to 2011 with matched data on the electoral cycles of the insurance commissioner, or the governor if the commissioner is appointed. The sample includes approximately 3,200 firms and consists of 300 separate elections in 50 states over 21 years. We use a hazard analysis to examine the statistical pattern of insurer failures. We control for insurer-specific factors, state-specific factors, differences across time and states, and the solvency screening tools used by regulators to cleanly identify the impact of electoral politics on the timing of regulatory interventions. We find regulatory interventions are much less likely to occur one year before elections. The results are statistically and economically significant. The probability of a regulatory intervention falls by 78% in the year before an election. The results are robust to various specifications and estimation techniques, including controlling for changes in the insurance commissioner, macroeconomic factors, risk measures produced by the private sector, incumbents, term limits, and electoral cycle seasons.
We then examine state insurance department resources around the electoral cycle to understand whether the delays are strategic, rather than the regulator merely being busy. We find that the number of workers charged with monitoring the solvency of insurers does not exhibit an electoral cycle, nor does the number of routine financial exams performed by these employees, indicating that state insurance divisions are neither understaffed nor busier before elections. However, the number of discretionary financial exams, i.e., the exams specifically requested by the insurance commissioner, exhibits a strong electoral cycle even though the number of firms that meet the conditions for a discretionary exam does not. The combined evidence suggests that insurance regulators strategically delay interventions upon failing firms before elections.
We next investigate whether there is a difference between appointed bureaucrats and elected politicians with regard to regulatory interventions. We find no difference between appointed and elected regulators in nonelection years but a significant difference in election years. While both appointed and elected regulators delay interventions in election years, the regulatory supervision provided by politicians (elected regulators) is more prone to incentives to delay action than that provided by bureaucrats (appointed regulators). Observing a difference between appointed and elected regulators in the year before elections, but not in other years, indicates that the difference is at least partially driven by political considerations, rather than differences in skills or preferences of appointed and elected regulators.
We also study the influence of competitive elections on the responsiveness of public officials. The incentive to delay should be especially acute if elections are competitive and not losing votes is particularly valuable. We find that electoral delays increase before tightly contested elections. The competitive election effect is mainly due to the incentives of bureaucrats. Appointed regulators do not delay interventions before elections in which the appointing governor is likely to be elected but do delay before competitive elections. Elected regulators delay interventions before all elections, regardless of the competitiveness.
We next explore whether regulatory governance mechanisms reduce political incentives to delay interventions. The National Association of Insurance Commissioners (NAIC), the association of the state regulators that coordinates solvency oversight among the states, promulgated a model law that mandated regulators take prompt corrective action against insurers with low levels of capital.3 Adopted by all the states, the model law took effect in 1994. The intent of the law was to reduce the discretion of regulators in when and whether to intervene on failing firms. We find that prompt corrective action reduces electoral delays. It is especially effective in reducing the discretion of appointed regulators, suggesting that regulatory rules that constrain the discretion of regulators are more effective when the regulatory function is separated from the political one. Finally, we study whether the ultimate costs of failure increase because of the inaction induced by elections. We find suggestive evidence that election induced inaction increases the cost of failure.
This paper contributes to two streams of literature. The first stream of literature examines the effect of elections on regulatory interventions. Brown and Dinç (2005) find the likelihood of bank interventions in developing economies decreases before elections of the highest level elected official (president or prime minister). Liu and Ngo (2014) find a negative correlation between governor elections and US bank interventions. This paper extends this literature by examining the US insurance industry, a unique laboratory setting that provides clean identification of the effect of elections on interventions because insurers are regulated exclusively at the state level,4 and unlike bank interventions, insurer interventions are not severely clustered in time.5 We find the magnitude of the preelection delay in the US P/L insurance industry is more than twice that observed in the US banking industry (Liu and Ngo, 2014).6 In addition, we find that prompt corrective action reduces electoral delays in the US insurance industry, suggesting that rules designed to govern regulators’ discretion can work. We also find competitive elections increase electoral delays. The unique setting also allows us to examine whether political incentives increase the cost of insolvency—we find evidence that suggests that they do.
The second stream of literature examines the impact of the mechanism to select public officials. A majority of the empirical research on the difference between elected and appointed regulators focuses on the public utility sector (see, e.g., Primeaux and Mann, 1986, Smart, 1994, Formby et al., 1995). For example, Besley and Coate (2003) find elected commissioners are associated with lower energy prices at the state level. Outside of public utilities, Whalley (2013) finds appointed city treasurers lower the cost of borrowing. Lim (2013) and Iaryczower et al., (2013) find appointed judges are more homogenous and make less harsh sentencing decisions. The evidence in the financial sector is limited and mixed. Fields et al., (1997) find the market value of life insurers doing business in California declined when the state changed from an appointed to an elected insurance commissioner. Grace and Phillips (2008) find elected regulators in states with strict rate regulation statutes are associated with higher insurance prices.
We extend this literature by exploiting gubernatorial and insurance commissioner elections to provide causal evidence of the impact of elections on regulatory interventions to determine how elected and appointed regulators' behaviors differ. We find elected regulators are more prone to delay regulatory action before elections than appointed regulators. This finding provides empirical support for recent theoretical work emphasizing the comparative advantage of bureaucratic control for technical policy making (Maskin and Tirole, 2004, Alesina and Tabellini, 2007). Empirically, the results are consistent with Iaryczower et al., (2013) and Whalley (2013). We also find prompt corrective action is more effective in reducing the discretion of appointed regulators than elected regulators.
Section snippets
Institutional setting
In this section, we discuss the institutional details relevant for understanding the US P/L insurance industry. We focus on the aspects of the industry that make it ideally suited for studying the politics of regulation and connect these details to our research design.
The US insurance industry is regulated exclusively at the state level. The main function of each state's insurance department is to monitor the solvency of insurers. Given the multistate nature of many insurance companies and the
Elections and regulatory interventions
The primary sources of the election dates and results are the Insurance Legislative Fact Book & Almanac, The Almanac of American Politics, and the Stateline database (www.stateline.org).13 The Insurance Legislative Fact Book & Almanac and the NAIC's Insurance Department Resources Report provide information about whether the insurance commissioner is elected or appointed.
The list of firms
Elections and regulatory interventions
There are several reasons why regulators could delay intervening on failing insurers before an election (Brown and Dinç, 2005). First, public officials might face questions about their competency when firms under their watch fail and thus have an incentive to defer action until after elections (Maskin and Tirole, 2004, Prat, 2005). Second, public officials want to generate favorable economic news prior to elections (Rogoff and Sibert, 1988). Third, the costs of closing a company fall on a
Cost of electoral delay
We next examine whether regulatory inaction due to elections increases the cost of insolvency. Previous studies examine the difference in the cost of resolving insolvencies across insurers (Hall, 2000, Grace et al., 2009, Leverty and Grace, 2012). Our study differs in that while controlling for the influence of the incentive structure on regulators (Hall, 2000) and managers (Lee et al., 1997, Grace et al., 2009), we examine whether electoral delays increase the costs of insolvency.
The cost of
Conclusion
Policy considerations motivate this research. If regulators intervene to minimize social costs, failures would be determined solely by firm health and economic conditions. This paper demonstrates that this is not the case. We find regulators are reluctant to take over failing firms if it negatively influences their own political or career goals. In particular, we find that regulatory interventions are delayed before elections. Studying the US insurance industry allows us to determine the
References (51)
- et al.
Interaction terms in logit and probit models
Econ. Lett.
(2003) - et al.
Information effect of entry in credit ratings market: the case of insurers’ ratings
J. Financ. Econ.
(2012) - et al.
To elect or to appoint? Bias, information, and responsiveness of bureaucrats and politicians
J. Public Econ.
(2013) Political uncertainty and investment: causal evidence from US gubernatorial elections
J. Financ. Econ.
(2017)Industrial policy and politics
Int. J. Ind Organ.
(1996)- et al.
Guaranty funds and risk-taking: evidence from the insurance industry
J. Financ. Econ.
(1997) - et al.
Elections, political competition and bank failure
J. Financ. Econ.
(2014) The treatment effect, the cross difference, and the interaction term in nonlinear “difference-in-difference” models
Econ. Lett.
(2012)- et al.
Bureaucrats or politicians? Part I: a single policy task
Am. Econ. Rev.
(2007) Financial ratios, discriminant analysis and the prediction of corporate bankruptcy
J. Finance
(1968)
Best's ratings, financial ratios and prior probabilities in insolvency prediction
J. Risk Insur.
The bright side of political uncertainty: the case of R&D. Unpublished working paper
Regulation and legislative choice
Rand J. Econ.
Some consequences of insurer insolvencies
J. Insur. Regul.
Understanding the components of bank failure resolution costs
Financ. Mark. Inst. Instrum.
Enjoying the quiet life? Corporate governance and managerial preferences
J. Polit. Econ.
Incumbent behavior: vote seeking, tax setting and yardstick competition
Am. Econ. Rev.
Does electoral accountability affect economic policy choices? Evidence from gubernatorial term limits
Q. J. Econ.
Elected versus appointed regulators: theory and evidence
J. Eur. Econ. Assoc.
The politics of bank failures: evidence from emerging markets
Q. J. Econ.
Regulatory solvency prediction in property-liability insurance: risk-based capital, audit ratios, and cash flow simulation
J. Risk Insur.
Insolvency experience, risk based capital, and prompt corrective action in property-liability insurance
J. Bank. Finance
Capital and risk in property-liability insurance markets
J. Bank. Finance
Municipal management structure and fiscal performance: do city managers make a difference?
South. Econ. J.
A market based evaluation of the election versus appointment of regulatory commissioners
Public Choice
Cited by (22)
Time to resolve insolvency and political elections
2022, International Review of Law and EconomicsCitation Excerpt :Nevertheless, delaying the resolution of a firm’s restructuring may lead to a deterioration of the relationship with key stakeholders that could harm the firm's likelihood of survival (James, 2016). From this perspective, the implementation of an appropriate regulatory structure and an appropriate system of supervision of insolvency processes has the potential to slow the impact of public policies delays (Leverty and Grace, 2018). Additionally, our econometric approach reports no significant association between the duration of insolvency procedures and the cycle of presidential elections.
The ownership complaint gap: Mutual versus stock intermediaries
2020, Journal of Financial and Quantitative AnalysisFinancial constraints and short-term planning are linked to flood risk adaptation gaps in US cities
2024, Communications Earth and EnvironmentThe Effect of Political Frictions on the Pricing and Supply of Insurance
2024, Review of Financial StudiesPolicy Uncertainty and Bank Mortgage Credit
2023, Journal of Money, Credit and BankingSupervisory power and insurer financial stability: the role of institutional quality
2023, Geneva Papers on Risk and Insurance: Issues and Practice
- ☆
We would like to thank the referee, Mark Browne, Conrad Ciccotello, Keith Crocker, J. David Cummins, Art Durnev, Rob Hoyt, Jim Johannes, Lars Powell, Joan Schmit, Stephen Shore, Art Snow, Justin Sydnor, Mary Weiss, Kenny Wunder, Terri Vaughan, and seminar participants at Georgia State University, St. John's University, Temple University, Virginia Commonwealth University, University of Georgia, University of Wisconsin-Madison, the 2012 American Risk and Insurance Association annual meeting, the 2012 CEAR/MRIC Behavioral Insurance Workshop, the 2014 Risk Theory Society meeting, and the 2016 Western Economic Association annual meeting for helpful comments and suggestions.