Education is awash in rhetoric and sloganeering – often wildly divorced from the evidence. An interesting aspect of the teacher pension issue is how much these large and complicated retirement systems bump up against much of the rhetoric we commonly hear in the teacher quality debate. Here are three ways and some basic ideas about fixes.

Turnover is bad! Teacher turnover is an evergreen concern. You, of course, want some turnover in any high-performing organization but too much is disruptive and costly. But while we hear a lot of concern about turnover – especially from the teachers unions – no one bothers to mention that today’s teacher pension systems are dependent on turnover to survive.  The basic math of pension systems is that a lot of small losers (people who leave before vesting or collecting end of career benefits) pay for the benefits of those few who survive (pdf).* At the extreme in states like Colorado only 13 percent of teachers earn full benefits. Nationwide only about one in five teachers achieves that mark. If a lot more teachers persisted these pension plans would crack financially.  They need the turnover to survive. They count on it. To be clear, I’m not saying they encourage it, just that they need a fair amount of turnover to function.

Veteran teachers are best! There is a lot of rhetoric about veteran teachers but the data show that while the returns to experience are important to a teacher’s growth they level off after a few years. Nonetheless, the wisdom and experience that 25 and 30 year veterans can bring to schools is very valuable and schools should cultivate effective veterans – especially those who want to continue giving to kids. Yet because of the way teacher pension systems work once someone earns full benefits it becomes economically against their interest to remain in the system. The data show that people respond to this incentive and retire. In other words, despite all the rhetoric about valuing veteran teachers today’s teacher retirement schemes are largely set up to push them out after a certain period of service – and despite whether they still want to teach or their school or district wants them to stay on.

Social Security is sacrosanct! Social Security demagoguery is a pretty effective political tactic and teachers union leaders are pretty good at it. But rarely is it mentioned that 40 percent of America’s teachers are not covered by Social Security. Adding to the retirement insecurity that today’s pension systems create for teachers, teachers in states or districts that do not participate also lose out on this portable and progressive social insurance program. Adding to the perversity, the reason some places don’t participate is because doing so is a bad deal for high-income earners (which teachers are not). I was talking with a state teachers pension fund leader recently who said they were against participating because their state would be sending more money to Washington than it gets back because it has so many high-income earners.  OK, but that progressively is part of what makes Social Security important and it would still be good for their members, and whatever happened to all this concern over the middle class working class Americans?

How to address these issues?

There is no single way but here are some guideposts:

Pension systems are based on a very uneven accrual of benefits, the last few years matter a lot, the years leading up to that not so much. Smoothing out the accrual curve and making benefits more portable would help reduce the savings penalties many teachers face. Enrolling teachers in Social Security would – while not a substitute for a good retirement system – add another leg to the stool for the teachers who are not enrolled and help with the portability issue.  And more transparency is a must so people understand how the system affects them and policymakers can understand when perverse incentive are built in.

*Important to note that this is not just a problem for people who only teach a few years. For instance a 30 year veteran with fifteen years of service in two separate states is adversely affected as well.