the merit of merit pay

Big new study finds that performance bonuses for teachers boost test scores (a bit)

PHOTO: Megan Witucki
Megan Witucki, a teacher at Compass Montessori School in Wheat Ridge, works with students.

A school district leader with money for teacher bonuses faces a choice: Should she spread the money around to all teachers equally or give more to teachers who have performed best?

A new study, released by the federal government, suggests that merit-based bonuses are the way to go, as they help raise student test scores without making a significant dent in teacher morale. It offers the latest evidence that programs of this sort can help schools and students, despite the common perception that they are ineffective.

The research focuses on a federal program known as the Teacher Incentive Fund, and compares schools that gave all teachers an automatic 1 percent bonus to those schools that gave bonuses based on classroom observations and student test scores. About 130 schools across 10 districts were randomly assigned to one of the groups.

The result? The schools that gave performance bonuses boosted student test scores throughout the four years of the study, between 2011 and 2015.

The differences between the two groups of schools weren’t big, and they were only sometimes statistically significant. For instance, in year three of the study, performance pay increased student test score performance by 2 percentile points.

A graph showing the share of teachers getting different size bonuses in districts that implemented performance pay.

But because the program was relatively cheap to implement, performance bonuses were very cost effective. They offered a better bang for the buck than class-size reductions, the researchers estimated.

And what about fears that performance pay would dampen teacher morale or reduce collaboration as teachers competed with each other for a fixed pool of bonus money? There was little evidence of that. According to teacher surveys, the program had only a limited impact on their job satisfaction, interactions with colleagues, or school morale. In the initial year of performance pay, there seemed to be small dips in morale and collaboration, but in year three the effects were actually positive.

The program also led to small increases in teacher retention, which may help explain the positive findings.

The study comes with an important caveat: It relies on test scores alone as a measure of learning. If the incentives pushed teachers to raise scores by less desirable means, like cheating or “teaching to the test,” that wouldn’t say much about the success of the policy.

The research adds to the hotly contested debate about how teachers should be paid. High-profile research in New York City and Tennessee from several years ago found disappointing results for merit pay, solidifying a conventional wisdom that the approach doesn’t work for teachers. Other districts, like Denver, have struggled with the design of a performance pay program and have seen inconsistent results.

But some recent studies have re-opened this debate. An overview of research last year showed that performance pay leads to small boosts in test scores, and a number of studies have found that added pay can help keep more effective teachers in the classroom.

It’s not entirely clear why the studies have reached such different conclusions. But performance-pay programs that judge teachers by test scores alone and only measure short-term effects tend to be disappointing. Programs like the Teacher Incentive Fund and those in Minnesota and Washington, DC, which use performance pay as part of a broader system for evaluating teachers, often produce more positive results.

That doesn’t mean that districts will heed the latest research. In fact, even in the new federal study — which found benefits for students — less than half the districts said they planned to continue offering performance bonuses after the Department of Education grant ran out.

Are Children Learning

Second study shows Indianapolis charter students fare better on tests

PHOTO: Dylan Peers McCoy

The second study in a week shows strong test scores for students at Indianapolis charter schools, bolstering the claims of advocates in a city where school choice continues to expand.

Indianapolis elementary students who attend mayor-sponsored charter schools beginning in kindergarten — and remain in the same schools — make bigger improvements on state tests than their peers in traditional schools across the city, according to the latest study.

The study contributes to emerging research that suggest that charter schools that are well managed and have good instruction can be successful, said co-author Hardy Murphy, a clinical professor of educational leadership and policy studies at the IUPUI School of Education.

The results of the study indicate Indianapolis charter school students are doing better than they would’ve done if they hadn’t enrolled in charter schools, Murphy said.

“This does not appear to have happened by chance,” he said. “I believe that the school experiences and the instructional teachers of those schools they are enrolled in are actually a big part of the results that we are seeing,”

The educational landscape in Indianapolis is defined by school choice. About 18,000 students who live in Marion County attend charter schools, and thousands more transfer to nearby districts or attend private schools with vouchers, according to state data. In recent years, the state’s largest district, Indianapolis Public Schools, has also become a national model for partnerships with charter schools. That makes understanding school performance essential for parents — but unpacking whether schools actually help boost student achievement can be particularly thorny for researchers.

With this study, Murphy said he and co-author Sandi Cole, director of the Center on Education and Lifelong Learning at Indiana University Bloomington, hope to disentangle one factor that makes studying charter schools difficult: the dips in test scores that students often experience after transferring to new schools. Murphy’s research focuses on students who began in charter schools in kindergarten and compares them to similar students in traditional public schools in Indianapolis.

“It’s time to move beyond the debate about whether or not charter schools are effective and start talking about, when they are effective, why, and for whom?” Murphy said, adding that successful approaches can be used in other settings.

The study focuses solely on students who attend charter schools authorized by the mayor’s office. For the control group, the study included township districts as well as Indianapolis Public Schools. The researchers plan to present their results to the education committee of the Indianapolis City-County Council and the 2019 Conference on Academic Research in Education.

The findings add to a growing body of research on Indianapolis charter schools. Last week, the Stanford-based group CREDO released a report that found that students at charter schools had test score gains that mirrored the state average, while Indianapolis Public Schools students made smaller gains on math and reading tests than their peers across the state. Another recent study found that when students moved to charter schools their test scores held steady.

To Do

Tennessee’s new ed chief says troubleshooting testing is first priority

PHOTO: (Caiaimage/Robert Daly)

Penny Schwinn knows that ensuring a smooth testing experience for Tennessee students this spring will be her first order of business as the state’s new education chief.

Even before Gov.-elect Bill Lee announced her hiring on Thursday, she was poring over a recent report by the state’s chief investigator about what went wrong with TNReady testing last spring and figuring out her strategy for a different outcome.

“My first days will be spent talking with educators and superintendents in the field to really understand the scenario here in Tennessee,” said Schwinn, who’s been chief deputy commissioner of academics in Texas since 2016.

“I’ll approach this problem with a healthy mixture of listening and learning,” she added.

Schwinn’s experience with state assessment programs in Texas and in Delaware — where she was assistant secretary of education — is one of the strengths cited by Lee in selecting her for one of his most critical cabinet posts.

The Republican governor-elect has said that getting TNReady right is a must after three straight years of missteps in administration and scoring in Tennessee’s transition to online testing. Last year, technical disruptions interrupted so many testing days that state lawmakers passed emergency legislation ordering that poor scores couldn’t be used to penalize students, teachers, schools, or districts.

Schwinn, 36, recalls dealing with testing headaches during her first days on the job in Texas.

“We had testing disruptions. We had test booklets mailed to the wrong schools. We had answer documents in testing booklets. We had online administration failures,” she told Chalkbeat. “From that, we brought together teachers, superintendents, and experts to figure out solutions, and we had a near-perfect administration of our assessment the next year.”

What she learned in the process: the importance of tight vendor management, including setting clear expectations of what’s expected.

She plans to use the same approach in Tennessee, working closely with people in her new department and Questar Assessment, the state’s current vendor.

“Our job is to think about how to get online testing as close to perfect as possible for our students and educators, and that is going to be a major focus,” she said.

The test itself has gotten good reviews in Tennessee; it’s the online miscues that have many teachers and parents questioning the switch from paper-and-pencil exams. Schwinn sees no choice but to forge ahead online and is quick to list the benefits.

“If you think about how children learn and access information today, many are getting that information from hand-held devices and computers,” she said, “so reflecting that natural experience in our classrooms is incredibly important.”

Schwinn said computerized testing also holds promise for accommodating students with disabilities and provides for a more engaging experience for all students.

“When you look at the multiple-choice tests that we took in school and compare that to an online platform where students can watch videos, perform science experiments, do drag-and-drop and other features, students are just more engaged in the content,” she said.

“It’s a more authentic experience,” she added, “and therefore a better measure of learning.”

Schwinn plans to examine Tennessee’s overall state testing program to look for ways to reduce the number of minutes dedicated to assessment and also to elevate transparency.

She also will oversee the transition when one or more companies take over the state’s testing program beginning next school year. Former Commissioner Candice McQueen ordered a new request for proposals from vendors to provide paper testing for younger students and online testing for older ones. State officials have said they hope to award the contract by spring.

In Texas, a 2018 state audit criticized Schwinn’s handling of two major education contracts, including a no-bid special education contract that lost the state more than $2 million.

In Tennessee, an evaluation committee that includes programmatic, assessment, and technology experts will help to decide the new testing contract, and state lawmakers on the legislature’s Government Operations Committee plan to provide another layer of oversight.

Spring testing in Tennessee is scheduled to begin on April 15. You can learn more about TNReady on the state education department’s website.

Editor’s note: This story has been updated with new information about problems with the handling of two education contracts in Texas.