A handful of new studies that scrutinize performance-pay initiatives nationwide have found mixed results on how they affect student achievement. That conclusion reinforces views that more work is needed on such plans, despite a recent surge in their popularity, before they can replace the traditional single salary schedule.
At a recent conference here organized by the National Center on Performance Incentives at Vanderbilt University, some researchers concluded that positive effects of performance-based pay were small or none, while others found varying levels of increases in student achievement.
The most positive results were reported by researchers who studied a small-scale pilot program in Little Rock, Ark., that paid bonuses to teachers for simply raising student test scores—a model some educators actually warn against because it can force teachers to teach to the test.
Among the programs that did not show large-scale gains was the Missouri career-ladder program, in place since 1987, which uses a mix of teacher performance, tenure, and extra responsibilities to determine bonuses. Researchers, who examined the program’s effects on student achievement by looking at the mathematics and reading data of more than 500 schools over a nine-year period, concluded that while the program did result in small positive increases in math scores, no significant improvement in reading occurred.
“If there is a positive effect, it is going to be very small,” said Kevin Booker, a researcher with Mathematica Policy Research in Princeton, N.J., who co-wrote the study.
It concludes that a possible outcome of the program could be that the career ladder could help the districts attract and retain higher-quality teachers.
Calls for Experimentation
In recent years, the single salary schedule, which is fiercely protected by teachers’ unions and which sets salary largely by seniority and education credits, has attracted criticism from policymakers under pressure to meet standards under federal education law and eager to hold teachers accountable for student achievement. A number of states—Florida, Minnesota, North Carolina, and Texas—and districts—Denver, Houston, Nashville, and New York City, among them—now offer varying versions of performance-pay programs. More states and districts are weighing them.
But observers and educators present at the conference sounded a warning note against rushing to adopt such plans without adequate experimentation.
“I worry that our enthusiasm for professional pay has gotten way ahead of our knowledge,” said Kate Walsh, the president of the National Council on Teacher Quality, a Washington-based research and advocacy group, pointing to the “many failures across the country and some glimmers of hope.”
“Both [national] unions are really not against such plans, but we are against full-scale implementation without experimentation,” Howard Nelson, the lead researcher at the American Federation of Teachers, said of his organization and the National Education Association.
Evidence emerging from the reports presented at the Feb. 27-29 conference demonstrates that some plans can indeed have unintended consequences. For instance, in North Carolina, researchers found that a program that rewards bonuses of $1,500 to teachers in schools that show test-score gains above a specified threshold could have resulted in higher rates of teacher turnover in low-performing schools.
While 70 percent of the schools got bonuses each year, some schools, such as those with higher numbers of minority and high-poverty students, were disproportionately unlikely to get bonuses. As a result, researchers said, teachers could view jobs in advantaged schools as more attractive.
The study also found that the program, in effect since 1996, resulted in an improvement in overall test scores, particularly in math, but there was less evidence to show that achievement gaps had narrowed.
Despite those drawbacks, Jacob L. Vigdor, an associate professor of public policy and economics at Duke University in Durham, N.C., said the program has been politically popular. Teachers in the right-to-work state, he said, also widely support the program and want to see the bonuses expanded. “They want every school to participate,” he added.
In Little Rock, meanwhile, a three-year performance-pay pilot program did appear to improve student achievement, more so among students whose teachers were previously less effective at producing learning gains.
Researchers at the University of Arkansas looked at the program implemented at five public schools from 2004 to 2007. The program, said co-author and researcher Jay P. Greene, was remarkably simple: A cash bonus was assigned to an increase in student achievement. There was no evaluation, no professional development, and no supervisory assessment involved.
The results showed that student achievement improved when the teacher became eligible for the bonuses. “There is much to be learned, but the initial results are encouraging,” said Mr. Greene.
Some conference participants offered the perspective that models involving teachers from the outset and not relying solely on test scores are the most successful.
A study by Julia E. Koppich, an education consultant based in San Francisco, concluded that the most comprehensive programs, such as those in Denver, Minneapolis, New York, and Toledo, Ohio, use multiple measures to gauge the extent to which teachers are contributing to increases in student learning. For instance, in Denver, Minneapolis, and Toledo, teachers are paid based on knowledge and skills and the responsibility they assume, as well as student growth.
Ms. Koppich said that all four programs are not punitive, retain echoes of the single salary schedule, and reflect careful planning and transparency.
What’s more, all are products of joint union-management undertakings, she said.
“Teacher involvement in developing and implementing these programs is essential. Those districts that choose to go behind closed doors and implement programs found they didn’t work very well,” Ms. Koppich said, noting the importance of bringing teachers’ unions on board while considering such plans.
A version of this article appeared in the March 12, 2008 edition of Education Week as Performance-Pay Studies Show Few Achievement Gains