Eating Crow and Crowing Over Performance Pay
Rewarding teachers with bonus pay, in the absence of any other support programs, does not raise student test scores, according to a new study issued today by the National Center on Performance Incentives at Vanderbilt University’s Peabody College of education and human development in partnership with the RAND Corporation.
With that simple paragraph, the whole world of performance pay is thrown on its ear.
I can’t, like Rick Hess, say the experiment tells us nothing. Nor can I say, like Eduwonk, that the results were unsurprising. A lot of people are running for cover and it’s discouraging to see. If we want to evaluate teachers on their performance, we should be prepared for performance pay programs to be evaluated on their performance.
In this experiment, fifth- to eighth-grade math teachers were given bonuses of up to $15,000 annually for the specific purpose of raising student test scores. Unlike most education research studies, this one was conducted scientifically, with random assignment and control groups. It had every opportunity to succeed and it didn’t.
Were conditions perfect? Of course not. The teacher attrition rate over the three years was atrocious – 50%. There were unexplained significant gains by fifth-graders that disappeared in the higher grades. But there are no indications that these factors had an effect on the overall result. There’s no getting away from the conclusion reached by Matthew Springer, executive director of the National Center on Performance Incentives:
“If teachers know they will be rewarded for an increase in their students’ test scores, will test scores go up? We found that the answer to that question is no.”
Naturally, the unions are all over this. The AFT response was measured:
This study, the most robust to date, concludes that individual performance pay based on test scores doesn’t work. It isn’t the motivational carrot its advocates believed. Education reform that actually improves teaching and learning requires a much more comprehensive approach, not just the implementation of one reform. It’s time to end our love affair with simplistic strategies that don’t get us where we need to be, in order to provide a great education for all children. There is a role for performance pay as part of a larger education reform plan that should also include providing teachers with the necessary tools, resources and conditions to do their job and using a robust curriculum. As this and several other studies show, performance pay doesn’t work by itself to boost test scores.
NEA portrays the report as the latest in an unbroken series of poor results for performance pay in a story headlined “New Study: Merit Pay Does Not Boost Student Achievement“:
The study — being billed as the first scientific study in the U.S. of teacher performance pay — is only the latest blow to merit pay, which the Obama administration continues to advocate as part of its education reform strategy. The $4.5 billion “Race to the Top” competitive school funding grant program encourages states to offer merit pay as an incentive.
It’s funny to see NEA suddenly equating “student achievement” with the results of a “single standardized test” (page 13) – in this case, the Tennessee Comprehensive Assessment Program (TCAP) math test.
The Tennessee results aren’t all rosy for the teachers’ unions, however. For one thing, the burden of proof is always on the performance pay side. The control group didn’t outperform the bonus group, so the report isn’t exactly a ringing endorsement of the traditional salary schedule. Second, if your goal is to raise student math scores, and a $15,000 bonus to math teachers didn’t do it, why would giving all teachers more money have any effect?
What happens next is just as much a political question as an education one. In the meantime, those of us who believe in the performance pay concept should avoid the temptation to rationalize the outcomes. We should take our lumps and learn from them, as a way to improve our own performance.
Wednesday, September 22nd, 2010
