Does Performance-Based Pay Improve Teaching?
Teacher pay has been a flashpoint in the education reform debate, with free-market-minded reformers claiming that merit-based approaches could improve educational outcomes by attracting and motivating high-quality teachers. Yale SOM economist Barbara Biasi studied what actually happened when some school districts in Wisconsin started paying partly based on effectiveness. The changes improved teacher quality for some students but could exacerbate inequalities between rich and poor school districts.
By Roberta Kwok
In 2011, the Wisconsin education system underwent radical changes. Until then, schools had paid teachers based almost entirely on seniority and education level. But a controversial new law called Act 10 weakened the teachers unions’ bargaining power and gave districts much more latitude to decide how salaries should be calculated.
Barbara Biasi, an economist at Yale SOM, wanted to know how districts responded. Would high-quality teachers earn more? And would those changes drive talented teachers to move to districts with new pay schemes?
Biasi found that some districts, which she called “flexible pay,” did appear to revise salary criteria. In these districts, teachers who scored better on a measure called “value-added”—a metric estimating how much a teacher contributes to student test score improvements—tended to earn higher salaries than those who scored poorly. In those districts, “better teachers are now paid more,” she says.
Biasi also found that effective teachers were more likely to move from districts that stuck with the seniority-based pay system to flexible pay districts, driving up the average quality of teachers in those areas. Flexible pay districts tended to be wealthier than seniority pay districts, and the influx of good teachers may have given these schools even more of an advantage over poorer counterparts. Increased teacher quality “is coming at the expense of other districts,” Biasi says. “And the districts that are winning are districts that were better off to begin with.”
Read the study: The Labor Market for Teachers Under Different Pay Schemes
To conduct the study, Biasi examined employee handbooks from Wisconsin school districts and categorized each one as “flexible pay” or “seniority pay.” Flexible pay districts were more likely to be located in suburban areas with higher property values, although the amount of money spent per pupil was similar across the two types of districts, she found.
After 2011, clear differences in salaries emerged. In flexible pay districts, Biasi saw bigger gaps between the 25th and 75th percentiles of salaries for teachers with a given seniority. For instance, in the flexible pay district Green Bay, that gap was $13,127 for teachers with five to six years of experience. In Racine, a seniority pay district, the gap was only $7,923.
These changes were linked to the value-added metric. Biasi compared the teachers who scored in the top 10% on this measure to those in the bottom 10%. In seniority pay districts, the most effective teachers earned 1.1% more than the least effective ones. But in flexible pay districts, that gap increased to 2.5%.
Next, Biasi investigated whether the changes drove teachers to switch schools. Before the reform, 0.27% of teachers moved from seniority pay to flexible pay districts each year. Afterward, the fraction of high-quality teachers who did so was 0.34 percentage points higher than that of low-quality teachers. And ineffective teachers were more likely to quit the profession altogether if they worked in flexible pay districts.
Does that mean that all districts should switch to flexible pay? Not necessarily. Biasi created a mathematical model to predict what would happen in that scenario. While overall teacher quality improved because ineffective teachers left the profession, the increase was much smaller than what Biasi saw in the current flexible districts.
However, the model did not account for changes in new teachers entering the system. It’s possible that flexible pay could attract more talented teachers, she says, but it also might discourage people who want predictable salary increases. Biasi is now studying how the reform affected the types of people entering the profession. “It calls for some caution,” Biasi says.“It really depends on who’s replacing the teachers who leave.”
One interesting observation from the results is that “districts are able to identify who the good teachers are,” Biasi says. While that might seem obvious, some observers have argued that merit-based pay won’t work because schools can’t identify talented teachers. And the results suggest that effective teachers want to earn higher salaries, contrary to arguments that teachers aren’t motivated by salary and value other factors such as job security more.
The study was not meant to determine whether Act 10 improved education in Wisconsin overall. That question is very much up for debate. The Center for American Progress Action Fund has reported that median teacher salaries and benefits have fallen, and more teachers are quitting. Other studies have looked at how test scores may have changed after the 2011 reform but haven’t yet produced clear results. Biasi used Act 10 as “a setting to learn how teachers’ labor markets function,” a question which has implications for any state or authority considering reforms aimed at improving teacher performance.
Biasi also isn’t claiming that basing salaries on test score improvements is the correct or only way to improve teacher pay. Other factors, such as whether a teacher works well with low-income kids, also are important, she says. Still, “a pay scheme that just rewards seniority and education is outdated,” she says. “We have to rethink the way that teachers are paid.”