Learning First Alliance

Strengthening public schools for every child

Is Merit Pay Merit-Less?

obriena's picture

Over the past few years, the idea of paying teachers a bonus based on student performance (typically on standardized tests) has been called into question for a number of reasons. Some education organizations have expressed concern about the focus it puts on tests they are not convinced accurately reflect student learning. They also question the underlying theory: That teachers can be motivated to work harder for more money; in other words, that they are not already working as hard as they can.

Some outside the education industry share this skepticism. Behavioral economist Dan Ariely expresses concern that these pay systems create odd incentives for teachers and points out that “If you teach, you want to focus on teaching and not on how your salary is changing every day. Not on your chance for a bonus.” Business writer Dan Pink questions how they motivate, believing that educators more than most respect the difference between intrinsic and extrinsic motivation.

Research also challenges the effectiveness of these systems. Last year, in what many considered the first controlled study of the issue, researchers found that Tennessee’s Project on Incentives in Teaching (POINT), which awarded bonuses of up to $15,000 to teachers who raised student standardized test scores, had no overall impact on student performance – “It simply did not do much of anything.”

This week we got further evidence suggesting that perhaps this is not the path to improved student performance, with a new study by RAND.

In evaluating New York City’s Schoolwide Performance Bonus System (SPBP), RAND found no positive effects of bonuses on student achievement (as measured by performance on state accountability tests) and no statistically significant differences between school progress report scores (which include measures of school environment, performance, progress and more) of SPBP schools and control schools or SPBP schools and other eligible schools over the three years the program was implemented. It also did not impact teachers’ self-reported attitudes or behaviors.

Why didn’t the program work? The study’s authors posit a number of theories, including that the program was too new to produce effects (though they aren’t convinced that’s actually the case) and that several factors thought to be important in pay for performance schemes were lacking in this program, including accurate understanding by school staff about the program [the amount of the award, the performance target needed to receive the award, and so on] and uncertainty teachers could act in ways that would allow their schools to meet performance targets. They also question underlying theory of action (maybe performance pay programs aren’t enough to bring around improvement) and the low motivational value of the bonus, especially given the other pressures of accountability that educators were already facing.

Based on the study, authors offer policymakers a few suggestions. Some are based on implementation of pay for performance strategies (such as improved communication with participating schools about how the program operates and transparency as regards award distribution criteria and plans), but they point out that there is no evidence that suggests implementing those strategies will improve the effectiveness of the program. They also recommend that before implementing a performance-based pay program, policymakers ensure stability of the measures on which the bonuses rely (instability in those measures in New York drastically reduced the number of schools receiving the bonus in the third year, which may impact educators' feelings towards it).

But the authors hone in on the role that motivation plays in changing educator behavior and thus student outcomes. While some argue that the reason these types of pay for performance systems don’t work is that bonuses are not motivating to educators, evidence in this study suggests otherwise. Here, teachers reported that the possibility of a bonus is a motivator. However, it is not as high a motivator as a good progress report grade, which has high stakes in terms of public opinion and potential negative consequences. The authors hypothesize that concerns about negative sanctions are a greater motivator than a monetary bonus of unknown size.

In addition, the authors suggest that just having a motivator is not enough. Perhaps educators lacked the capacity to change or identify new teaching strategies that might have improved student performance, despite being motivated to do so. The bonus did not impact that capacity.

So what does all this mean? In New York City, it means that SPBP will be no more – officials cancelled it based on the results of this study. And growing concern about the effectiveness of the programs may contribute to the willingness of state officials to defund it in times of tough budgets. Texas recently slashed the budget for what used to be the nation’s largest merit pay program.

Yet some government officials are convinced that merit pay is the wave of the future. With Race to the Top, the federal government incentivized the development of performance pay systems. And states like Florida and Ohio are working to develop them.

So is merit pay merit-less? In theory, I don’t think so. I firmly believe that we have to develop new ways of compensating teachers. Differentiated pay is an important part of that. And I think many educators agree.

But as is becoming more and more clear, the promise of merit pay alone won’t improve student performance. Such pay systems have to be carefully designed and perceived by school staff as fair. They should be accompanied by resources to help school staff build capacity to change practice. They should be based on something other than scores on standardized tests that are widely recognized as flawed. And until we can ensure that these (and perhaps other) conditions are met, we should proceed with caution in adopting and implementing new pay systems.