Advocates of teacher pension plans claim that pensions help retain teachers in the profession. That is, they argue that the very structure of defined benefit pension plans encourages workers to stick around and devote their lives to the profession.*
The problem is, researchers can't find much evidence this is true. For example, when St. Louis spent $166 million to enhance the retirement benefits it offered to teachers, it saw a temporary, one-year boost in retention among teachers already eligible for retirement. But all other groups of teachers acted as if nothing had changed; after all that money, they were no more or less likely to leave St. Louis schools. Similarly, state plans themselves do not assume that qualifying for a pension has an effect on teacher retention. If pensions were really a retention incentive for workers, we'd see evidence of teachers hanging on just long enough to qualify for a pension. We don't.
The latest piece of evidence on this question comes from a BLS Working Paper on Oregon. Prior to 1996, Oregon offered its teachers what the authors call "one of the most generous pension plans ever devised." The state had two formulas, and upon retirement Oregon teachers automatically received the better of the two. The first was a traditional defined benefit pension plan awarded by formula, and the second was a "money match" pension plan that gave teachers an amazing investment promise. In good years, teachers received the actual market return (minus a small fee), and in bad years teachers were guaranteed at least an 8 percent return. The state offered unlimited upside while shielding teachers from any potential losses.
This was an incredible deal, and eventually Oregon legislators realized they couldn't afford it. They subsequently enacted retirement plans for new teachers that were worth half to one-third as much as was given to those hired before 1996.
This provided an excellent real-life experiment to test the theory of pension advocates. If they were right, Oregon's pension cuts should have led to a mass exodus of teachers. Oregon went from having one of the most generous retirement plans ever created to a run-of-the-mill plan equivalent to what the typical state offered. Teachers hired after 1996 had much less reason to stick around, so theory would predict they'd have higher turnover rates.
That's not what happened. Despite having much less generous retirement plans, retention rates for early- and mid-career teachers didn't change at all. For late-career teachers, turnover actually fell. Regardless of plan type or teacher experience level, Oregon's teacher turnover rates looked pretty much identical to those in nieghboring Washington State. The authors concluded that, "Oregon’s policymakers and citizens allocated substantial resources to its retirement system and, in return, received little economic benefit in the form of promoting longer teacher tenures."
What was going on here?
The authors suggest opacity as one major driver. Oregon teachers received annual statements on the value of their traditional defined benefit formula, but they never saw estimates of the value of their money match formula. Only when they neared retirement did they see the full value of the state's money match promise. That "windfall" may have pushed teachers to retire earlier than they otherwise would have.
The opacity argument may explain part of Oregon's extreme situation, but it doesn't explain results like the one in St. Louis, or what we see in state pension plan assumptions around vesting periods. Another explanation is simply that teachers, like most humans, are not that worried about retirement until they get close to retirement age. That's what was happening in St. Louis, and a forthcoming paper on the state of Missouri suggests a similar pattern. Teachers just aren't that sensitive to retirement plans until very, very late in their careers. That would also explain why teachers seem to retire based on when the retirement plan nudges them to do so.
In short, the best way to characterize the research literature thus far is that pensions don't enhance teacher retention. In fact, because traditional pension plans push out veteran teachers, and because those veterans tend to be better than their replacements, pension plans are actively harming overall teacher quality. Contrary to the theories of pension plan advocates, shifting to alternative retirement plans that didn't push out veteran teachers would be better for students.
*Pensions could also be serving as a recruitment incentive by attracting certain types of people to the teaching profession. That's possible, but half of all new teachers won't qualify for any pension at all, and 80 percent won't stay long enough to reach the full normal retirement age. The back-loaded nature of pensions may be deterring some otherwise talented people who would prefer to receive upfront compensation rather than waiting 20 or 30 years for the promise of a lucrative retirement benefit.