Policy makers in South Carolina are considering changes to retirement benefits for public employees to make employer costs more predictable, reduce risk and uncertainty, and to shore up long-run funding of their pension system. These reforms typically involve prospective benefit reductions, contribution increases, or both for future entrants. Typically, alternative proposals are assessed in terms of their effects on pension cost and risk. We evaluate the impact of a range of potential pension reforms on employee retention. Using estimated economic models of the retention behavior of South Carolina public employees, we simulate the retention responses of these employees.
We find that changes in pension design, especially changes in the design of the defined benefit (DB) plan, create behavioral responses that influence how long these employees remain in public service. Further, reforms to the state's DB plan influence its relative value compared to the state optional alternative retirement plan, a defined contribution plan (DC), leading to shifts in employee choice between these plans.
We also considered recent reform proposals which have included an enhancement to the DC plan and a hybrid DB/DC plan. Both plans are predicted to lead to relatively small changes in employee retention, suggesting limited impact to employee turnover and longevity relative to the status quo. We find that our example hybrid plan comes closest to achieving no change in employee retention relative to the current DB plan with a slight increase or decrease in the average length of service depending on the group of public employees examined. The key implication of our analysis is that pension reform proposals should consider not just the effects of the proposals on pension funding and risk but also the effects on retention and the experience mix of the workforce.
Table of Contents
Overview of the South Carolina Public Pension Systems
Data and Sample
Model and Fit
Analysis of Alternative Policies
Conclusions and Implications
Technical Details of Model
Estimating Earnings Trajectories