Tennessee Teacher-Retention Bonus Paid Dividends, Study Finds
Tennessee teachers with top teacher evaluation ratings were more likely to continue teaching in low-achieving schools when given a substantial pay incentive, according to a new working paper released released this morning.
But there's a catch: The bonus program only seemed to affect teachers working in the "tested" grades, not those teaching other grades or subjects, noted the study, conducted by the Tennessee Consortium on Research, Evaluation, and Development.
"It appears tested subject teachers are driving the effect, which isn't unexpected given the amount of weight Tennessee's teacher-evaluation system attributes to school-level performance for untested subject teachers," the study concludes. That may be because in Tennessee, teachers of nontested grades and subjects get part of their evaluation score based on schoolwide performance. It's meant to encourage collaboration, but has been controversial and even led to a lawsuit because teachers say they don't have control over how students they don't teach will fare.
The 2013 initiative gave $5,000 to teachers with high "value added" scores, based on student test scores, who agreed to stay in the state's "priority schools" for another school year. Tennessee's 83 priority schools are defined as those in the bottom 5 percent based on student-proficiency rates.
For the study, the authors looked at the 56 priority schools that agreed to participate in the program. About 375 teachers in those schools had a "level 5" rating, the highest, on their teacher evaluation rating and agreed to participate. (The program also offered $7,000 to effective teachers who transferred into such schools, but only 59 teachers took up that option, and that part of the program wasn't studied.)
To estimate how the program affected teachers, the scholars compared the retention rates of the teachers who qualified for the program to those taught by teachers who fell just below the qualification bar. They accounted for factors, such as student demographics, that could skew the results. The quasi-experimental research design helps to establish a cause-and-effect relationship, though it isn't considered as conclusive as a random-assignment study.
Here's a rundown of the results:
- Teachers who received the bonus were more likely to remain teaching in a priority school compared with those just below the cutoff.
- For teachers in "tested" grades and subjects, the participating teachers were 24 percent more likely to remain compared to those below the cutoff. But for nontested subject teachers, there was no statistically significant effect.
- Using recent research on the impact of high "value added" teachers, the scholars estimated that the state would in the long run recoup all but 5 percent of program costs through taxes on the higher income potential of students taught by these effective teachers.
There has been some research on retention bonuses, but much of it has found mixed results, probably because of the varying size of the bonuses and their context (some were pieces of larger initiatives, others were stand-alone efforts). As a result, there hasn't been much of a consensus on how these programs ought to be structured.
For example, last year, I wrote about a federally financed initiative that, at least initially, improved teacher-retention rates for teachers who transfered to low-achieving schools and received a substantial bonus. Conversely, a Washington incentive program tied to the district's teacher-evaluation system boosted teacher performance but didn't have a noticeable impact on teacher retention for the most effective teachers.
This study falls squarely in the "pro" camp, providing preliminary evidence that significant financial bonuses can help to keep the top teachers on the job.
The researchers cautioned, though, that the program was announced in April 2013, after some schools had already started to fill vacancies. And second, the study found "considerable noncompliance" with the rules for distributing the bonuses, which means that findings might be partly attributable to other factors.