The research presented in this documented briefing provides estimates of the effects of early retirement incentives, buyout incentives, and retention allowances on the retirement rates of federal employees covered by the Civil Service Retirement System (CSRS) in the Department of Defense (DoD). The research grew out of two interests.
The first was a general interest in understanding the effects of financial incentives, particularly those provided by retirement and pension plans, on retirement and separation behavior among workers. Our study provides information on these effects using workers covered by CSRS as a case study. Federal employees covered by CSRS are a useful case study because the CSRS pension formula is relatively simple to compute, and a large number of federal employees are covered by CSRS but not by any other retirement plan, such as Social Security.
The second interest related to the increasing numbers of retirements from federal employment that are predicted to occur over the next decade owing to the aging of the federal workforce. Most of these retirements will be among workers covered by CSRS. To the extent that federal agencies want to influence the timing of these retirements, there are a variety of financial incentives that could be used. Early retirement incentives and voluntary separation incentives (or “buyouts”) can induce individuals to leave sooner than they would under the normal retirement option, by making the incentive for early departure more lucrative and the incentive to leave later relatively less lucrative. Retention allowances can be used to induce the deferment of retirement to a later date, thereby reducing retirements in the near term. By controlling the timing of retirements, personnel managers can better ensure that retirements and separations occur when they will have the least cost or have the least undesirable effects. Timing control can also help managers coordinate the hiring of qualified replacements, if such hiring is needed, with the timing of separations and retirements. Thus, these programs can help achieve effective workforce planning.
Our research consists of two separate strands of analyses. The first focuses on using administrative data on federal workers in DoD who are covered by CSRS to estimate alternative models of retirement behavior. These models provide estimates of the effects of financial incentives on retirement behavior. The results of this part of the analyses are more technical and are presented in a companion document, The Retirement Behavior of Federal Civil Service Workers, by Beth Asch, Steven Haider, and Julie Zissimopoulos, RAND and Michigan Retirement Research Center, UM-1–09, 2002.
The second strand of analysis focuses on applying our estimates of the effects of CSRS incentives on retirement behavior to predict the effects of buyout and early retirement incentives as well as the effect of retention allowances . . .