Journal Issue: Excellence in the Classroom Volume 17 Number 1 Spring 2007
Policy Implications of the Evidence on Teachers' Incentives
Research evidence suggests, although not conclusively, that pay-for-performance incentives can improve teachers’ performance. Yet these incentives can also lead to unintended and undesired consequences growing out of their inherent structural challenges, such as measurement issues or the possibility of teachers’ directing their efforts exclusively to rewarded activities. Although there are no magical cures for any of these problems, it is possible to draw lessons from the experience and evaluation of the better performancerelated pay programs. In this section I offer some general guidelines for designing effective programs.
Which Outcomes to Reward
Teachers should be evaluated on the basis of their true performance, not random variation in performance. Therefore, the performance measures chosen should exhibit relatively little random variation. Because teachers are expected to direct their efforts according to the incentives provided, the performance measures should cover all outcomes of interest, including quality and quantity (relevant examples are the two outcome measures used in the Israeli pay programs—passing rate and average score). Outcome measures should be as close as possible to the ultimate outcomes, difficult to game, and easy to monitor regularly.
Close supervision and monitoring are needed to minimize gaming and shading the truth, if not outright fraud, in measured output. Because monitoring can be costly, an appropriate strategy should balance the trade-off between penalties and probability of detection. The matriculation examination system in Israel minimizes gaming: students are tested twice—in a school and a state examination— in each subject, and their final score is a weighted average of the two. Large gaps between the scores invite penalties. Nonrandom gaps between the two scores are audited at the class and school levels. Obviously, even the best-designed measurement and monitoring system may lead to inappropriate reallocation of effort because no single system is likely to capture all important behavior. One way to avoid this constriction in effort is to include two outcomes, for instance, test scores and processes (for example, attention directed explicitly to disadvantaged students) and teacher practices (for example, teacher absenteeism).
Teachers can be compared according to student performance levels or to some form of gain in performance. Value-added measures, in particular, are more appropriate for measuring individual teacher effectiveness and ensuring fair ranking. Such methods can substantially affect school or teacher rankings, in addition to the behaviors engaged in to game the system. Performance levels create incentives to pay greatest attention to highestperforming students; value-added criteria may increase attention to lowest-performing students. It is thus important to use performance measures that can be adjusted for contributing factors, such as student socioeconomic characteristics, but not at the cost of complex value-added measures that lack transparency and are difficult to understand and accept. In addition, goals should be attainable and not be perceived as too ambitious.
Rewarding Both Short- and Long-Term Measures
Performance-related pay programs typically focus on key short-term schooling objectives— such as increasing the number of students reading at a given grade level—rather than on long-term objectives related to postsecondary education and labor market outcomes. This emphasis is due in large part to the difficulty of measuring long-term objectives, as well as to an almost universal belief that achieving short-term goals is a necessary condition for promoting long-term goals. But medium-term outcomes, such as dropout rates, can be affected by performance-related pay. The team-based program developed in Israel demonstrates how a mix of short- and medium-term goals can be achieved simultaneously in an effective performance-related pay program.
Subjective and Objective Evaluation
Subjective evaluation invites the problem of performance measure inflation (namely, that teachers under the pressure of incentives tend to give higher than warranted scores), while objective metrics invite the problems of measurement cost and the inability to encompass all school targets and goals. Given these limitations, it may be appropriate to use both—each imperfect but still informative. For example, selective emphasis can be assigned to each subjective and objective evaluation criterion, depending on subject taught or specific task, and associated with individual, team, and school performance. The evaluation process and its results should be transparent to all parties involved.
School and Individual Incentives
Because some degree of teamwork characterizes all schools, incentives should balance individual rewards with school incentives. Design of these incentives should foster a cooperative culture, but not at the cost of an aggravated free-riding problem, a condition likely to arise when only group incentives are used.
Monetary and Other Incentives
Money is not alone in motivating performance and promotions. Improved working conditions and increased decisionmaking authority may also be attractive to teachers, especially when combined with bonuses or salary increases. The Israeli school incentive program included modest monetary bonuses but large rewards in the form of media attention and enhanced reputation for the winning schools. Recognition and prestige are, it appears, highly effective motivators.
Eligibility and Size of Awards
All teachers should be eligible for the incentive offered. Only when the majority of teachers are eligible to enjoy the benefits of hard work and improved outcomes will the incentive scheme be effective. However, only a subset of teachers should be rewarded in practice. If too many teachers are rewarded, teachers may not need to exert much effort to benefit.
A performance-related pay program can motivate teachers and schools even more effectively if it also allows teachers to make decisions regarding instruction, curriculum, and other aspects of schooling that contribute to attaining the desired outcomes and goals. Creativity flourishes in the absence of bureaucratic constraints. The search for ways to improve teaching technology arises both because teachers and schools have a direct stake in the matter and because merit pay plans reduce the need for regulation of inputs and thus give teachers and schools more freedom. It is therefore essential that performance- related pay plans be accompanied by the increased flexibility required for their success.