Teacher Merit Pay: Beyond the Rhetoric

The Canadian education blogs were buzzing last month about Erin Anderssen’s commentary on merit pay for teachers, published in The Globe and Mail. Too often merit pay generates polarized rhetoric rather than thoughtful discussion. To Anderssen’s credit, she introduces some of the research and contextual issues that frame the debate about this reform strategy.

Anderssen notes, for instance, that merit pay, which rewards individual teachers for higher levels of performance, is not particularly popular with teacher unions. Unions have long argued that individual incentive plans force teachers to compete, rather than cooperate, creating a disincentive for teachers to share information and teaching techniques (National Education Association, undated). Frank Bruseker, president of the Alberta Teachers Association, also has argued that compensation for student achievement will remove the incentive for teachers to work with struggling students (Mendleson, 2009).

The unions have some evidence on their side, since research on individual compensation pay plans is mixed. In an Israeli study reported in 2002, Lavy did find a correlation between an individual incentive program and student performance on standardized tests. He evaluated a particular incentive program experiment, in which teachers were rewarded with cash bonuses for improvements in their students’ performance on high school matriculation exams. His analysis supported the hypothesis that teachers’ monetary performance incentives had a significant effect on students’ achievement in English and math. However, other studies have not been as complimentary to merit pay schemes. Eberts, Hollenbeck, and Stone (2002), for instance, presented case study evidence from a U.S. county where one high school had piloted a pay for performance system to reward student retention. Comparing this high school with another in the same county that used a traditional compensation system, the researchers found that individual incentive programs for teachers were associated with a significant fall in drop-out rates, but were also associated with no apparent effect on students’ grade point averages, reduced average daily attendance, and an increased percentage of students who failed.

Teachers generally favor school- or group-based performance award programs, in which performance is measured and rewarded at the school or group level (National Governors Association, undated), since this approach is more consistent with the social norms of equality and collegiality in the teaching profession. Group goals are established, especially in the areas of student achievement (level or growth), and graduation and attendance rates, and bonuses are paid to teachers and other staff according to the degree of goal attainment. Typically, the single salary structure remains intact (Heneman, Milanowski & Kimball, 2007).

There are some applied research findings that appear to offer support for use of group-based compensation plans to improve student achievement. Lavy (2002b) matched Israeli schools eligible for a school-based incentive program with schools in similar communities. The study’s empirical results suggest that group monetary incentives contributed to significant gains in many dimensions of students' outcomes (Lavy, 2002b). Glewwe, Ilias and Kremer (2003) studied the effects of a school-based teacher incentive experiment in rural Kenya, where every teacher in grades 4 to 8 in a winning school got the same bonus. The authors demonstrated that students in treatment schools had higher test scores. However, there was little evidence that students in treatment schools retained these gains after the end of the program, consistent with the authors’ hypothesis that teachers seeking performance bonuses focused on manipulating short-run scores.

Knowledge- and skill-based compensation plans reward individual teacher behaviors or competencies thought to be linked to high-quality teacher performance (Heneman, Milanowski & Kimball, 2007). Teachers receive bonuses or increases in base salary for the acquisition of new competencies, classroom performance mastery, and/or certification by the National Board for Professional Teaching Standards (NBPTS). Often, changes are made to the single salary structure when this approach is utilized. For example, the number of steps in the salary structure may be reduced (Heneman, Milanowski & Kimball, 2007).

Knowledge- and skill-based pay plans are often based on the premise that teacher knowledge and competencies are positively correlated to student achievement. However, research on this relationship has produced mixed results. National Board certification, for example, has recently come under research scrutiny. A study of more than 300 fifth-grade teachers in North Carolina found that the distinction of being a National Board Certified Teacher does not necessarily translate into greater student success in the classroom. McColskey, Stronge, and colleagues (2006), for instance, found no significant relationship between National Board certification status and student achievement gains in the classroom. Another study, also conducted in North Carolina, produced quite different results. Using more than 600,000 student records from North Carolina schools, Goldhaber and Anthony (2004) demonstrated that students of National Board Certified Teachers improved on elementary math and reading tests more than pupils whose teachers did not achieve National Board Certification. The effects of NBCTs on students who were younger or low-income were even greater.

The success of a merit pay plan depends, in large part, on a reliable and valid system for measuring performance. Knowledge-based incentive systems require the simplest measurement system; in most cases, all that is required is documentation that the teacher has completed a professional development activity or achieved certification. Individual- and group-based compensation systems require a more complex measurement system that assesses teachers’ classroom performance, such as a standards-based evaluation system, and to varying degrees, correlates teacher performance to student achievement. This type of system typically utilizes a comprehensive set of standards and rubrics with the intention of enhancing instruction and strengthening educational accountability (Danielson, 1996).

A related measurement issue not frequently discussed within our borders is the efficient tracking of data related to teacher performance and student achievement. Many states south of the border, including Florida, Georgia, and South Carolina, have already taken steps to add teacher data to existing student unit record systems, making it simpler to track this information on a state- or school board-level. These systems also make it possible to identify which students and which courses are being taught by teachers with different levels and types of preparation, certification, professional development, and seniority and how these experience levels are related to the academic growth of the students. The federal government in the U.S. is making substantial investments in these enhanced data systems, seeing them as the foundation from which states and school boards can make data-driven decisions to improve student learning, as well as to facilitate research to increase student achievement.

Too many commentators glibly push (or decry) merit pay for teachers as the best (or the worst) thing to improve education. Anderssen, to her credit, takes the higher road and attempts to sketch the complexity of the issue—the range of merit pay strategies and the interrelated issues of measurement and data systems. It is time to open a richer discussion on merit pay, drawing upon international experience in merit pay to inform a new-generation Canadian experiment in rewarding teacher performance.