The Politics of Public Sector Pensions

Abstract

As Laswell (1950) famously wrote, politics is the study of who gets what, when, and how. My dissertation explores this question in the context of public sector pensions in the United States.

I suggest that the relationship between political parties and activists is too complicated to represent with a simplistic ``I'll scratch your back if you scratch mine'' heuristic. Democratically controlled state governments do not always act generously towards state employee pensions, and Republican ones do not always act to dismantle them. In the context of political parties, interest groups, and public pensions, Laswell's dictum takes some unraveling.

Chapter 2 investigates the ``what'' and ``when'' of pension policy. I examine the determination of liability and asset levels of these public pension funds by focusing on specific policies that have led to the expansion and contraction of funding levels since 1995. Building on an original dataset constructed from state personnel and pension benefit data, I demonstrate that the roles that legislative and executive partisanship play in these policy decisions is not straight forward and strongly dependent upon the policy environment. Importantly, public sector unions are not substantively responsible for contributing to state fiscal solvency woes, at least via pensions, and often act as bulwarks against poor pension performance when paired with friendly co-partisan state governments.

The ``how'' of pension policy is explored in Chapter 3. I investigate how and why some public pension plans underestimate their liabilities and overestimate their assets to a greater degree than others. Utilizing an original dataset that expands Novy-Marx & Rauh (2011)'s calculations to encompass 17 years of pension funding, I investigate the distinct influences that transparency and the presence of attentive stakeholders have on the likelihood of accurate financial disclosure. I demonstrate that the degree to which the government distorts the health of a public pension is not just a function of the actual health of the pension, but also of the information environment of the state, the administrative structure of the pension board, the degree to which multiple types of state employees are covered by a single pension plan, and the strength of public sector unions. This suggests that the type of subtle corruption inherent in improper accounting and reporting behavior is not simply a result of political culture or amateur policymakers, but a rational response to concrete political trade-offs.

Next, Chapter 4 explores what actual and perceived state-level fiscal crises might mean for public policy and fiscal sustainability, and how policies in other arenas might have their ``who, what, when, how'' scrambled by irresponsible pension policy. Using historical S&P’s ratings and revised pension data that has been stripped of reporting bias, I investigate whether self-reported or revised data are better at predicting one-year ahead general obligation bond ratings. Though both data fare reasonably well in predicting ratings, the self-reported data appear to outperform revised data in terms of explanatory power. Analyzing borrowing costs directly, however, suggests that creditors may be more savvy, and that actual liability data is better at predicting the size of the interest burden relative to outstanding debt.

Finally, Chapter 5 summarizes and concludes with directions for future research.