Research on Teacher Pay-for-Performance Programs 

In this article in Educational Evaluation and Policy Analysis, Kun Yuan, Vi-Nhuan Le, Daniel McCaffrey, Laura Hamilton, and Brian Stecher (RAND Corporation), Julie Marsh (UCLA), and Matthew Springer (Vanderbilt University) report on three randomized studies of programs that gave teachers cash incentives for student-achievement gains. Their research questions were: (a) Did teachers find the incentive-pay programs motivating? and (b) Did the programs bring about changes in teachers’ classroom practices or working conditions? The answer to both questions was No. Here are the details. 

First, the authors examine the logic model for performance pay. Proponents believe that financial incentives linked to student results will have the following results:

• Motivate existing teachers to improve their teaching practices, innovate, modify the content of the curriculum, work longer hours, take more professional development and polish their skills in other ways.
• If pay bonuses are given to teams of teachers, teamwork and cooperation will increase.
• Pay-for-performance will attract a more highly qualified pool of teachers and convince the most talented teachers to remain in the classroom. 

For performance-pay programs to work, teachers have to believe that their personal efforts will lead to higher student achievement, that the goals are clear and moderately challenging, that reaching achievement goals will bring them financial rewards, and that getting the rewards is worthwhile. 

The authors studied three performance-pay programs, all based on the first two bulleted approaches above:

• POINT – This program was implemented over three years in a 73,000-student district. Teachers could earn a bonus based on their students’ year-to-year value-added gains on the statewide math test. Teachers who reached the 80^th percentile of a benchmark received $5,000, those who reached the 85^th percentile got $10,000, and those who reached the 95^th percentile got $15,000. Teachers in the treatment and the control groups all received $750 for participating. Initially, 296 grade 5-8 math teachers volunteered. There was some attrition and the number of bonus winners ranged from 41 to 44, with awards ranging from $9,623 to $11,370. 

• PPTI – Implemented in middle schools in a 43,000-student suburban Texas district, this program awarded bonuses to the highest-scoring teacher teams based on value-added test-score gains. In 2008-09, 67 teachers on 14 teams won bonuses averaging $5,373; in the second year, 52 teachers on 12 teams won bonuses averaging $5,862.

• SPBP – This New York City program rewarded entire school staffs in high-need schools that met Progress Report benchmarks (Progress Reports included test-score gains, graduation rates, student attendance, and school environment). Schools that met their annual performance targets received $3,000 per full-time staff member, and schools that met 75 percent of their targets got $1,500 per staff member (within each school, a four-person compensation committee decided how to allocate the pot and most divided the funds evenly). Over the three-year life of the program, about 193 schools participated, with about 167 schools in the control group. Awards went to 62%, 84%, and 13% of treatment schools averaging $2,857, $2,841, and $2,812 per staff member. 

The researchers found that teachers were not motivated by the programs because: 

(a) some teachers didn’t fully understand the criteria; (b) some teachers had concerns about using test scores to measure teaching performance; (c) some teachers didn’t believe the program was fair; (d) many teachers thought family background was more important than their own efforts in producing student achievement; and (e) even though teachers would have liked to earn a bonus, they saw a bonus as an acknowledgement of their hard work rather than an incentive to work harder. Teachers said that seeing their students do better was the most important reward and therefore the biggest reason to work hard.

The researchers also found that none of the programs resulted in changes in teachers’ instructional practices, number of hours worked, or collegiality (except that those in the POINT program reported a greater emphasis on test preparation and collaboration). 

“It is difficult to obtain teachers’ support of incentive pay programs if they think the performance measure is problematic,” say the authors. “However, given the current emphasis on educational accountability, it is also difficult for incentive pay programs to totally ignore student test scores or test score gains when measuring teacher performance.” 

“Given our findings and the previous literature that finds weak effects of performance pay for teachers,” the authors conclude, “policymakers might favor other reforms.” These might include compensation tied to career ladders and other professional growth goals, compensation for working in challenging schools, or bonuses for implementing specific practices that have proven to be successful. 

“Incentive Pay Programs Do Not Affect Teacher Motivation or Reported Practices: Results from Three Randomized Studies” by Kun Yuan, Vi-Nhuan Le, Daniel McCaffrey, Laura Hamilton, and Brian Stecher (RAND Corporation), Julie Marsh, and Matthew Springer in Educational Evaluation and Policy Analysis, March 2013 (Vol. 35, #1, p. 3-22), 

http://epa.sagepub.com/content/35/1/3.abstract; Yuan is at kyuan@rand.org. 

From the Marshall Memo #479