At the same time Georgia is racing to develop merit pay systems for its teachers, emerging research is questioning both the fairness and effectiveness of basing teacher pay on student performance.
Under merit plans being considered around the country, student test scores in core subjects would count for as much as half of a teacher’s evaluation. Twenty-six Georgia counties are expected to pilot merit pay under the state’s federal Race to the Top grant.
Now, Georgia rewards teachers for years in the classroom and degrees, neither of which has been linked to improved student performance. But is merit pay determined by student scores a better alternative?
The case against that form of merit pay was bolstered with the recent release of a much-anticipated study out of Nashville that shows no impact on student performance from teacher bonuses.
Produced by the National Center on Performance Incentives at Vanderbilt University, the study examined the test scores of 300 middle school math teachers who agreed in 2006 to participate in the Project on Incentives in Teaching. Teachers could earn between $5,000 and $15,000 annually, depending on how their students performed on state exams.
The project was a three-year randomized experiment that tested the assumption — an assumption that undergirds Race to the Top and many of the reforms coming out of Washington — that teachers will work harder and produce greater student gains if they are rewarded for it.
Not so, according to the findings, which come on the heels of a policy brief by the nation’s top education researchers challenging the merits of merit pay.
“We tested the most basic and foundational question related to performance incentives — does bonus pay alone improve student outcomes? – and we found that it does not,” says Matthew Springer, executive director of the Vanderbilt center.
The study found little evidence that merit bonuses improve student achievement, noting improvement only in fifth-grade scores in two of the three years.
There was zero impact on the scores of sixth- and eighth-graders, causing the researchers to conclude that merit pay did not “yield consistent and lasting gains in test scores. It simply did not do much of anything.”
Teachers told researchers that they were already working as hard as they could and the bonus wouldn’t affect their behavior.
The Nashville study reinforces a policy paper issued a few weeks earlier by a blue ribbon panel of education researchers, including Eva L. Baker, Paul E. Barton, Linda Darling-Hammond, Edward Haertel, Helen F. Ladd, Robert L. Linn, Diane Ravitch, Richard Rothstein, Richard J. Shavelson, and Lorrie A. Shepard
The 27-page paper published by the Economic Policy Institute criticized the use of test scores, commonly called the “Value Added Model,” to judge teacher performance.
The brief contends that there’s no research basis for merit pay. It cites studies that have found little consistency in teacher performance over time, noting, “... a teacher who appears to be very ineffective in one year might have a dramatically different result the following year. This runs counter to most people’s notions that the true quality of a teacher is likely to change very little over time and raises questions about whether what is measured is largely a ‘teacher effect’ or the effect of a wide variety of other factors.”
The brief advocates a more comprehensive approach to evaluating teachers, including classroom observations, a portfolio of student work and peer reviews.
But does any school system have the time, resources or staffing to conduct the thoughtful and deeper evaluations that these researchers recommend, especially now when teachers complain their strapped schools can’t even afford paper towels or pencils?
Many new teachers contend that their principals pop their heads in their classrooms only once or twice a year to evaluate them and then disappear back down their administrative rabbit holes.
In an interview, Matthew Springer, the co-author of the Nashville study, noted that his project only examined merit pay in the context of student scores on a single test. His study didn’t consider whether merit pay might improve the recruitment of better teaching candidates or retain talented teachers longer.
“While it’s nice to have some hard evidence on the issue, there are many questions left unanswered,” Springer said.
As for the Race to the Top funded pilots of merit pay, Springer says they must have research at their core.
“The current system of compensating teachers isn’t efficient,” he said. “We have been debating merit pay for a century now, but we still don’t know a better way. It is critical we learn from these pilot initiatives.”
He says pilots should acknowledge the shortcoming of relying on a single snapshot of student achievement to determine a teacher’s effectiveness and incorporate broader, more comprehensive views that may entail classroom observations and parent and student ratings.
As education research moves forward, Springer says we ought to look at multiple measures to assess teachers and provide continual feedback and professional development to them.
“This next generation of pilot pay reforms needs to take into account a more complete set of factors and that we pair these efforts with scientifically rigorous research,” says Springer. “If we don’t learn as we go, we will continue to repeat the failures of the past.”