It is increasingly apparent that public defined-benefit (DB) pension plans, including teacher plans, across the United States are in a difficult financial situation. Chicago Public Schools is now routinely in the news for its precarious finances, owing in large part to pension problems. In a recent paper covered previously in this blog space, we document that on average across state plans, over 10 percent of salaries for new teachers are being collected and used to pay down previously accrued pension debts. Even more, there is growing awareness that pension plan reports are based on actuarial assumptions that make their financial conditions appear better than they actually are. The Citrus Pest Control District No. 2 in California learned this lesson the hard way when it tried to convert its workers from the state pension plan, CalPERS, to a 401(k) plan.
What do pensions buy us? A look at St. Louis Public Schools
OK, so pensions are expensive. They require real and substantial resources that could be used for something else. But what do they buy us? An oft-cited theoretical benefit of DB pension plans is that they help to retain workers. Improving retention is of value to school systems because it reduces turnover costs, including indirect costs to student learning, and can increase the experience of the workforce, which also benefits students.
In a recent paper, coauthor Brett Xiang and I study a large and very expensive enhancement to the pension formula for teachers in St. Louis Public Schools. The enhancement strengthened teachers’ financial incentives to remain in covered employment until full-retirement eligibility. We leverage the enhancement to provide empirical evidence on the retention effects of changes to teachers’ pension incentives.
In brief, the enhancement we study increased the “formula factor” or “multiplier” in the St. Louis plan from 1.25 percent to 2.0 percent per year of service (it changed nothing else). This represents a 60 percent benefit increase for all workers. The enhancement was implemented retroactively in one fell swoop. For example, in 1998, the year before the enhancement, a pension-eligible teacher with 30 years of service would have been able to retire and collect a pension that replaced 37.5 percent (30*1.25) of her final average salary annually for the rest of her life; the next year her replacement rate rose to 62 percent (31*2.0). Thus, even though 30 of the service years were accrued under the old pension formula, all 31 years are rewarded at the higher pension rate. (Note that teachers’ pension benefits in St. Louis are in addition to Social Security, although nationally many teachers are not covered by Social Security.)
The enhanced pension formula increased the value to teachers of remaining in St. Louis Public Schools until attaining retirement eligibility—that is, it made their retention incentives stronger. Moreover, the retroactive implementation of the enhancement created a situation where some teachers (those closer to retirement eligibility) had much more to gain than others. We set up models to test whether teachers whose pension incentives were most affected by this substantial enhancement were more likely to remain in the system due to the enhanced benefit formula. However, we find no evidence that the pension enhancement generated differential increases in employee retention in St. Louis.
There are several possible explanations for our findings, including that (a) teachers do not greatly value, and/or do not fully understand the values of their pensions, and (b) the policy was poorly targeted. On the first point, note that the enhancement did not change the retirement timetable for teachers—it only changed the value of reaching pre-existing milestones by changing the formula factor. Teachers may not have understood just how much this mattered.
The second point refers primarily to the retroactive implementation of the enhancement. What this meant in practice is that rather than using the enhancement resources primarily to incentivize newer teachers to stay in the workforce longer, many teachers close to or at retirement simply received much larger pensions than they would have otherwise. A stark example is a teacher who was already retirement eligible when the enhancement was enacted. There were many such teachers in the workforce, and they essentially received a pure windfall. It is quite common for enhancements to pension plans to be implemented retroactively. Perhaps the motivation for this is political, as senior teachers and administrators benefit most from retroactive implementation. However, it is hard to argue that rewarding late-career teachers liberally and unexpectedly for time they’ve already served is motivated by the goal of improving the workforce in a way that benefits students in schools.
Ultimately, the enhancement we study was very expensive for the district—in 2013 dollars, we estimate the cost in present value was $166 million just for the single cohort of teachers working when it was enacted. This is over one quarter of the entire operating budget of the school district at the time. What’s more, the enhancement also obligated the school district to provide richer pensions for future cohorts of teachers—meaning an even bigger price tag. Yet, we do not find evidence of a meaningful retention response among teachers.
Our findings support concerns that the substantial resources directed toward DB pension plans for teachers are not generating commensurate benefits to the education system. All indications point toward teachers not valuing these benefits as much as they cost to provide. It seems likely that the funds currently devoted to support teacher DB plans could be redeployed in a more strategic manner to promote the highest quality workforce for students in K-12 schools.
— Cory Koedel
Cory Koedel is an Associate Professor of Economics at the University of Missouri.
This post originally appeared on the Brown Center Chalkboard.