Researchers extend RAND's dynamic retention model for Defense Department civil service workers by reestimating pay profiles, adding costing capability, obtaining estimates for separate and combined cohorts, and allowing estimates to differ for veterans and nonveterans. They apply the extended model to simulate the retention and cost effects of mandating workers to pay a larger share of contributions into the Federal Employees Retirement System.
An Enhanced Capability to Model How Compensation Policy Affects U.S. Department of Defense Civil Service Retention and Cost
- How can policymakers and planners better understand how changes in compensation and personnel policy affect the U.S. Department of Defense civilian workforce?
Policymakers need a rigorous capability to evaluate the retention and cost effect of possible changes to U.S. Department of Defense (DoD) civilian workforce compensation. The research reported here extends earlier work by estimating dynamic retention models (DRMs) for entry cohorts from 1988 to 2000; estimating new age–earning profiles for federal civilian and private-sector workers for use as an input to the DRM; examining DRM parameter estimates by cohort to check for consistency and possible trends, such as a change in the preference for DoD civilian employment across cohorts; and developing a costing model for comparing the cost of alternative policies along with their retention effects. The earning analysis finds that estimates based on recent cross-sectional data from the American Community Survey closely reflect earning profiles constructed from synthetic cohorts of Current Population Survey data. Differences in the preferences for DoD civilian employment across cohorts are small, and the use of a DRM estimated from a sample combining the cohorts is justifiable. The DRM fits the data well for each cohort and the combined sample. The new capability is demonstrated by simulating the retention and cost effect of a policy change mandating that workers pay a larger share of the total contributions to the Federal Employees Retirement System.
The Revised Model Fits the New Data
- Estimates based on recent cross-sectional data from the American Community Survey closely reflect earning profiles constructed from synthetic cohorts of Current Population Survey data.
- Differences in the preferences for U.S. Department of Defense civilian employment across cohorts are small, and the use of a dynamic retention model estimated from a sample combining the cohorts is justifiable.
- The dynamic retention model fits the data well for each cohort and the combined sample.
- The new capability is demonstrated by simulating the retention and cost effects of a policy change mandating that workers pay a larger share of the total contributions to the Federal Employees Retirement System.
- The dynamic retention model (DRM) capability can be used to simulate the retention effects of other policies of interest.
- The DRM capability could also be extended to other occupational areas within the U.S. Department of Defense (DoD), including the cyber workforce; to other pay systems, such as the science, technology, engineering, and math workforces in the various demonstration programs and the wage grade workforce; to specific demographic groups, such as women and minorities; and to specific locations of interest, such as Hawaii.
- The DRM capability could be applied to civil service workforces in other agencies within the federal government, including the U.S. Department of Veterans Affairs, the U.S. Department of Homeland Security, and the various agencies that make up the intelligence community.
- Incorporate into the model the fact that some civil service personnel leave and then return.
- Incorporate changes in expectations about future policy changes. Pay freezes and furloughs can lead to changed expectations about the likelihood of future pay freezes and furloughs.
- The model could be extended to create a "total force" model of DoD workforce dynamics and the effects that compensation has on those dynamics, in which total force includes active and reserve military personnel and DoD civilians (but not contractors).
- RAND has already developed a unified DRM capability to provide logically consistent and empirically based estimates of the effects that compensation policy has on active-component retention and reserve-component participation. This capability could be extended to include DoD civil service employment.
Table of Contents
Entry Cohort Characteristics
The Dynamic Retention Model: Overview and Extensions
Estimates of Federal and Private-Sector Pay Profiles
Dynamic Retention Model Estimation Results
Estimation of Pay Profiles
Entry Cohort–Specific Dynamic Retention Models with Fixed Personal Discount Factors