Assessing Pension Plan Options and Teacher Retention: Q&A with David Knapp

David Knapp

Researcher Spotlight David Knapp

David Knapp is an associate economist at the RAND Corporation. His research focuses on worker recruitment, retention, and retirement and their effects on households, employers, and workers.


States across the country have begun legislating reductions in teachers' retirement benefits as a way to address large unfunded liabilities of their pension plans. But policymakers have limited understanding of how these pension reforms and changes to teacher compensation will affect retention in a school district.

Knapp, along with a team of RAND researchers, analyzed the relationship between compensation, including retirement benefits, and retention over the length of the careers of public school teachers in Chicago. The structural modeling approach they used—the dynamic retention model (DRM)—was first developed at RAND for studying the relationship between military compensation and the retention of military personnel. They recently published Retirement Benefits and Teacher Retention: A Structural Modeling Approach, a RAND venture that was funded in part by philanthropic contributions.

He describes here how the model was extended to public school teachers in order to demonstrate the trade-offs between cost savings and teacher retention.

You used data on Chicago Public Schools (CPS) teachers to model their retention decisions under different pension policies. What did you find?

We chose four popular pension reforms to analyze—decreasing the cost-of-living adjustment; increasing the number of “high pay” years used to calculate benefits; increasing the vesting, early, and full-benefit collection ages; and decreasing the “benefit multiplier,” which is the percentage of a teacher's salary to be received in retirement. Several of these policies had already been put into place for new CPS teachers hired after 2010.

What we found was that certain policies had significantly more of an impact on retention, while others, like decreasing the cost-of-living adjustment, had a very small effect. As an example of a large response, we found that increasing the vesting and retirement eligibility ages would induce more teachers to leave earlier in their careers, in the first 10 years, but the ones who stayed, who made it to 27 years, would actually remain in the profession beyond that time. By changing the pension benefit in this way, you would end up with a larger fraction of newer teachers and much more senior teachers.

We were also able to help CPS make a decision about how to deal with its pension liabilities. CPS was recently in the midst of contract negotiations and was considering a voluntary retirement incentive. The reason for offering an incentive was to entice more teachers to retire, and they wanted to know how many teachers would take their planned one-time incentive. What we showed them was that approximately 600 teachers would take the incentive, but about 450 would have retired anyway.

What are the implications of the findings?

Changes to compensation can change teacher behavior. While school districts have focused on reducing their unfunded pension liabilities by shifting how they compensate teachers, there's no sense in fixing the pension liability problem if doing so creates a whole new set of problems. The failure to retain the right teachers is a concern. It's costly to replace a teacher and lower retention rates can lead to lack of continuity in instruction, less expertise in the classroom, and fewer teachers in the pipeline to become principals and administrators. Conversely, teachers who stay in teaching longer are more costly.

Our dynamic retention model (DRM) can serve as a practical tool for future contract negotiations and policymaking—including exploring policies aimed at reshaping the experience composition of the teaching workforce. With this tool, we can help stakeholders weigh policy alternatives and establish efficient compensation systems.

Can the model be adapted to assess other policies?

Yes. The DRM was originally developed by RAND to understand the relationship between compensation and retention in the military, but we demonstrated that this model works for the teaching profession. We'd like to extend our work to other states and districts, and to incorporate school characteristics into the model to explore how non-financial incentives—for example, smaller classrooms or new teaching methods—affect teacher retention. We're hopeful the model can be applied to other public employee sectors—any sectors with a unified pay and pension system.

How did philanthropy make this research possible?

This project would not have happened without donor support. It's so important to be able to think ahead and be ready when policymakers need to address issues like pension reform. CPS benefitted from our ability to quickly apply our model and analyze their options. Many school districts are already underfunded and don't have extra dollars for research or the time to line up grants or other funding sources. We were able to say, “We have a tool. How can we help you?” This foresight and the support for RAND ventures are what make RAND unique in the research and policy analysis world.

Learn more about David

Related Resources