By Barbara Butrica & Dan Doonan

Workforce turnover has implications for employees and employers. On the employee side, turnover can involve lost wages and benefits. On the employer side, turnover can involve the loss of experienced employees as well as the investments made in training them. It also generates additional costs associated with recruiting and training replacement workers. Understanding the timing of turnover and how retention evolves over employees’ careers is important for economic security, workforce planning, public service delivery, and benefit design. Previous studies of workforce turnover and retention have relied primarily on household surveys with self-reported turnover and tenure (Haverstick et al. 2010; Munnell, Haverstick, and Sanzenbacher 2006), records from individual employers or pension systems that focus on particular states, employers, or workforces (Brown 2013; Clark, Hanson, and Mitchell 2016; Goda, Jones, and Manchester 2017; Goldhaber, Grout, and Holden 2017; Quinby 2020; Quinby and Wettstein 2021), or cross-sectional analyses that provide a snapshot of worker tenure at a single point in time (BLS 2024). One notable exception is Munnell et al. (2014), which uses actuarial termination and retirement assumptions from a sample of public pension plans to examine the probability that workers vest and, once vested, remain until the earliest full retirement age. These studies provide important insights into public-sector retention, which is generally higher than in the private sector (BLS 2024). However, much less is known about how retention evolves over public employees’ careers or how retention varies across public workforces and retirement systems.

To address this gap, this report examines workforce turnover and retention in the public sector using insights from public pension plans. The analysis draws on information collected from state-administered public pension plans and supplemented with selected information from the Public Plans Data (2001-2025). It relies on turnover assumptions published in actuarial reports, which are based on administrative census data describing the historical experience of pension plan members, to characterize longrun turnover and retention patterns among public-sector workers. Because these assumptions are developed from large administrative datasets and smoothed to represent expected long-run experience, they provide another perspective on how workforce retention evolves over employees’ careers.

This report extends prior research by using administrative turnover assumptions from a large sample of public pension plans and employee groups to characterize long-run retention patterns among publicsector workers.

The report focuses on public pension plans that cover general employees, state employees, local employees, teachers, and police officers and firefighters. The sample includes 136 state-administered public pension plans representing 203 plan-employee groups and 12.8 million active members. The actuarial studies that set termination assumptions typically use five years of census data, suggesting that these assumptions draw on roughly 60 million person-year observations across the 136 plans.

This report first compares workforce tenure patterns nationally and in the sample. The analysis then examines overall retention patterns, compares retention across workforce and plan characteristics, considers variation in retention across individual plans, shows how retention patterns translate into expected years of service, and briefly discusses and summarizes the findings.

Get the report here