“Within the ten evaluation districts, pay-for-performance led to slightly higher student achievement in reading and math by the second year of implementation,” the study states. “Student reading achievement was higher by 2 percentile points at the end of the first year in schools that offered pay-for-performance bonuses than in schools that did not. The total difference remained at 1 to 2 percentile points across the subsequent three years and was statistically significant in most years.”
The study, conducted by Mathematica Policy Research for the U.S. Department of Education’s Institute of Education Sciences, examined performance in schools participating in the Teacher Incentive Fund (TIF), which Congress established in 2006 to provide “grants to support performance-based compensation systems for teachers and principals in high-need schools,” the report states. “Evaluation of the Teacher Incentive Fund: Final Report on Implementation and Impacts of Pay-for-Performance Across Four Years,” released by DOE in December 2017, examined TIF programs in more than 130 districts awarded grants in 2010, performing in-depth examinations of 10 districts.
The researchers performed a random assignment study of the pay-for-performance component of TIF in the 10 evaluation districts. “Within those districts, this evaluation provided a more in-depth examination of TIF implementation and measured the impacts of pay-for-performance bonuses on educator effectiveness and student achievement,” the study states.
‘Still Have a Lot to Learn’
Matthew G. Springer, an associate professor of public policy and education at Vanderbilt University’s Peabody College and director of the National Center on Performance Incentives, says the study’s results are not as cut-and-dried as they may appear.
“We still have a lot to learn in terms of what is an optimal design for a merit-based pay system,” Springer said. “The Mathematica study adds important knowledge to the existing research base showing that merit pay systems can have a positive impact on student test scores. However, how we design those systems and what the optimal system will look like over time still needs to be decided and, to some extent, is context-specific. As we think about the impact of merit pay systems more generally, it is also critical that we also talk about the impact on the teacher labor market. I believe the greatest potential benefit of merit policies will be their ability to recruit and retain high-performing educators and counsel low-performers to seek ways to improve or exit the profession.
“Merit pay systems must incorporate multiple measures of teacher effectiveness,” Springer said. “They need to take classroom observations into consideration. They need to take value-added measures of teacher effectiveness into consideration, both at the individual and group or team levels. Some merit pay systems even include student feedback as well as individual student learning objectives. If we can begin to move toward using multiple measures to evaluate teacher performance, we are going to get more robust and arguably more effective merit pay systems in the long run.”
‘It Is Not That Easy’
Corey DeAngelis, a policy analyst at the Center for Educational Freedom at The Cato Institute and a distinguished doctoral fellow at the University of Arkansas, says even incentivizing teachers isn’t enough to overcome the harms of a monopolistic arrangement, and the system requires an injection of real, market-based choice.
“Being a free-market advocate, I initially strongly supported merit pay programs,” DeAngelis said. “Market supporters hear words like ‘incentives’ and ‘merit’ to describe pay-for-performance programs and inappropriately think that the reform is a step in the right direction. But we should not expect that putting one aspect commonly associated with the free market— incentives—into the broken system of residentially assigned government schools will lead to better lives for our children.
“Unfortunately, it is not that easy,” DeAngelis said. “The most important aspect of a market is the ability for customers to choose among alternatives. Sadly, anything short of that is unlikely to work. If families do not value test scores, and merit pay programs entice teachers to become test-score-maximizers, the result could be a lower-quality educational product overall. In the current system, since families do not have an exit option, they cannot incentivize schools to change their performance pay practices. The unfortunate reality is that the government, not the customer, decides what types of metrics will be used for the performance pay system.”
Prescribes Parent Power
DeAngelis says the system would be better if parents were allowed to decide what factors they value most about schools.
“Researchers like Jay P. Greene at the University of Arkansas are frequently pointing out that test scores are not good proxies for the long-term outcomes that society actually cares about,” DeAngelis said. “For example, some charter school evaluations have found huge positive impacts on test scores with no effects on graduation rates or college enrollment. On the other hand, other recent school choice studies have found little or no test score gains with much larger positive effects on graduation rates, college attendance, earnings, and crime reduction.
“Because non-cognitive skills are so difficult to capture with objective metrics, the only way for performance pay to work is if we combine it with a market that allows customers to exercise educational choices based on their unique and subjective evaluations of teacher quality,” DeAngelis said.
Kenneth Artz ([email protected]) writes from Dallas, Texas.
National Center for Education Evaluation and Regional Assistance, “Evaluation of the Teacher Incentive Fund: Final Report on Implementation and Impacts of Pay-for-Performance Across Four Years,” Institute of Education Sciences, December 2017: https://heartland.org/publications-resources/publications/evaluation-of-the-teacher-incentive-fund