Boosting early retirement in cash-strapped districts doesn’t hurt students’ math and reading scores, according to new studies released at the American Economic Association meeting here, but pension-incentive programs may cost schools some of their most effective teachers.
Separate studies of teachers in California, Illinois, and North Carolina paint a complex picture of the choice increasingly faced by education leaders: Keep your most experienced—and expensive—teachers, or encourage them to retire to ease budget woes.
Cornell University researchers Maria D. Fitzpatrick and Michael F. Lovenheim, both assistant professors of policy analysis and management,in grades 3, 6, and 8 before and after the state’s “5+5" pension- incentive program, which took place in the early 1990s.
The 5+5 program allowed any teacher age 50 or older, and with at least five years of experience, to qualify for pension benefits immediately if he or she retired at the end of the 1992-93 or 1993-94 school year and paid a one-time fee. The average teacher who took advantage of the program had 29 years’ experience, compared with incoming teachers who had on average less than three.
Immediately after becoming eligible for pension benefits, the most- and least-effective teachers in North Carolina are the first to leave, a new study finds. By six years out, however, more-effective teachers are much more likely to retire than less-effective ones.
SOURCE: Patten Priestly Mahler, University of Virginia
Before the program, more than half the teachers in each of the three grades studied were veterans, with at least 15 years’ experience. After, the average experience level at each school dropped by 2.6 percent for every veteran teacher it had before the pension buyout.
Some middle schools even ended up replacing their exiting teachers with fewer teachers overall.
Yet the less-experienced teachers didn’t lead to lower test scores, Ms. Fitzpatrick and Mr. Lovenheim found. In fact, 8th grade scores in both mathematics and reading improved slightly, but the difference was significant.
Moreover, the poorest and lowest-performing schools saw the biggest test-score gains.
“We were very surprised; it took us a while to convince ourselves this finding was real,” Mr. Lovenheim said. “I don’t think it has been highlighted as a policy question to look at the end of teachers’ careers, but these findings suggest that it may be an important thing to look at.”
The Illinois results are bolstered by another, this time of California teachers, also released at the economics conference Jan. 4-6. Kristine M. Brown, an assistant professor of economics and labor and employment relations at the University of Illinois at Urbana-Champaign, found that teacher retirements between 1998 and 2001 in the state were associated with increases in the proportion of students achieving above the national median in math and reading by 2 percentage points, on average.
In both states, “these teachers who are very close to retirement are just not as effective as younger [incoming] teachers,” Mr. Lovenheim said, raising the possibility that pension systems that backload benefits or require a set number of years of service might demand an “educational cost of keeping teachers who want to retire in the workforce.”
Raising the Middle?
However, average test-score gains may cloak the loss of some of a school’s institutional memory, according to Patten Priestley Mahler, an economics doctoral candidate at the University of Virginia, in Charlottesville.
In a separate, Ms. Mahler found a teacher’s likelihood of retiring jumped 17 percentage points after he or she became eligible for pension benefits, compared with just two years earlier, even though teaching positions often have younger minimum retirement ages than other professional careers. She agreed that many teachers may feel “pulled to stick it out a few more years” in order to receive their full pension benefits, even if they are no longer interested in teaching.
When teachers become eligible for their pensions, the financial incentives to stay reverse themselves, Ms. Mahler said: “They look at it and say, ‘If I quit teaching, I can get 60 percent of my salary, so I’m only getting paid 40 percent to work the same 40 hours a week. My time is not as valued.’ ”
Yet in North Carolina at least, teachers did not respond equally to early-retirement incentives. The most- and least-effective teachers (as judged by their students’ previous math and reading test scores) were 5 percentage points more likely to retire after qualifying for their pension than were teachers whose students had average performance.
Moreover, Ms. Mahler found lower increases in turnover for urban and high-minority schools.
“Those are the schools that are hardest to staff; they have people leaving all the time. These pension incentives just might not be keeping them around as strongly as other teachers,” she said, “but then the ones who actually stick around and stick through it for long enough to be eligible for the pension may have other reasons to be teaching that aren’t affected by the pensions.”
Because so few teachers work under nonpension retirement programs, Ms. Mahler and Mr. Lovenheim both said it is hard to gauge how strongly teachers weigh financial versus other reasons to teach, such as altruism or enjoyment.
Jesse Rothstein, an associate professor of economics and public policy at the University of California, Berkeley, who was not associated with either study, argued that teachers are not good at identifying their own effectiveness, making it difficult to target incentive systems to encourage only less effective teachers to leave.
States vs. Districts
Mr. Lovenheim also cautioned that the test-score increases found in the California and Illinois studies aren’t substantial enough to make pension changes a major force for school improvement, particularly as they can set up a showdown between districts and equally cash-strapped states, which are left footing most of the bill for early teacher pensions.
“Hey, any increase in test scores is good if you can do it for low cost,” Mr. Lovenheim said, “but this is not necessarily a low-cost change, because of the dramatic increases to the state budgets. I’m not sure the test-score increases are enough to pass the cost-benefit test.”
Even without early-retirement incentives, more than a third of educators in the classroom today are older than 50, at or approaching retirement age. Teacher-pension funds in 41 states do not have enough money to meet their obligations, according to a study released in December by the National Council on Teaching Quality. The group estimates a $390 billion shortfall nationwide in 2012.
More states have been pushing back against district early-retirement promotions, the council found. Since 2008, half of states have increased the retirement age for teachers, 22 have lowered retirees’ cost-of-living increases, and 21 have both reduced benefits and required teachers to contribute more to their plans.
Teachers may retire just to avoid dealing with district belt-tightening, even if they don’t get a golden parachute.
For example, the research organization WestEd’s Regional Educational Laboratory West found, in a separate 2012 study of California teachers, that for every $1,000 cut from per-student spending, teachers in the state were 4 percent more likely to retire.
Coverage of policy efforts to improve the teaching profession is supported by a grant from the Joyce Foundation, at
A version of this article appeared in the January 23, 2013 edition of Education Week as Loss of Veterans Doesn’t Hurt Scores