Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Graduate School


Public Policy and Administration

First Advisor

Dr. Dwight V. Denison


Defined Benefit (DB) state pension trust funds are an integral component of state finances and play a major role in the country’s labor and capital markets. The last decade though has seen a substantial growth in unfunded pension obligations and a seeming inability by states to make the contributions needed to cover funding shortfalls. When coupled with even larger unfunded retirement health benefits, the looming threat of insolvent state retirement systems pose both current and long-term fiscal challenges to state governments already struggling with the ongoing economic downturn and billions of dollars in budget deficits. The convergence of these factors have led states to undertake various reform strategies in an attempt to move their respective public pension plans towards a more sustainable funding path.

Using an asset-liability framework to describe the DB plan funding structure and process, this dissertation advances the discussion over major pension reform efforts currently implemented or considered by states. I show analytically the link between various pension reform categories and specific DB plan funding components, and how this in turn, affects DB plan funding outcomes. From this analytical framework, I derive the study’s hypotheses on the relationship between DB plan reform-linked funding components and outcomes of interest.

This study looks at three DB-plan reform-linked funding components: (1) plan member employee contributions, (2) plan employer contributions, and (3) retirement benefit payments. Four major funding outcomes are evaluated: (1) the employer contribution rate, (2) flow funding ratio, and (3) stock funding ratio, and (4) relative size of plan unfunded liability.

Utilizing a unique panel dataset of 100 DB state retirement systems from 50 states covering a nine-year period of FY 2002 to 2010, I empirically test the following hypothesized funding relationships: (1) States as DB plan sponsors have underfunded their plans as indicated by their failure to meet annual employer funding requirements; and (2) Increasing the employee and employer contribution rate and reducing the cost of retirement benefits are associated with higher plan stock funding ratios and lower unfunded pension liabilities.

Results from my fixed-effects (FE) panel regression analyses provide the clearest empirical evidence to date that state DB pension plan sponsors underfunded their required annual employer contributions. The financial condition of a state’s budget is also shown to have a significant effect on the amount states are able to contribute into their pension funds. I find empirical support for the crucial function of employer contributions in determining the overall funded status of state pension plans. This finding is further reinforced when I estimate plan stock funding ratios using a dynamic system GMM (sGMM) panel regression model. The results from static FE and dynamic sGMM models suggest no significant effect on overall plan funding levels from changes in the employee contribution rate or the average retirement benefit cost. Lastly, the results lend evidence to the significant influence of past funding levels on current funding levels. It is recommended that future empirical research account for the dynamic nature of public pension funding and related endogeneity issues. This dissertation concludes by discussing the implications of the empirical findings for policy makers seeking to improve the funded status of their respective state DB retirement systems.