Friday, October 11, 2013

When Merit Pay Is Worth Pursuing

Performance-based pay is controversial. But in five Arkansas elementary schools, it's proving effective.
Virtually every U.S. school official, education researcher, or interested citizen has a preferred remedy for low student achievement, particularly among disadvantaged learners: instructional coaches, smaller classes, more time on task, and so on. Many argue that improving student achievement is only a matter of having the political will to adopt these strategies. But here's the rub: There's little conclusive evidence that any of these strategies consistently leads to improved student performance.
However, evidence does show that one element of the education mix improves student performance: the presence of a high-quality teacher. Indeed, many researchers maintain that improving the quality of the nation's teaching force is the best policy intervention for raising student achievement (Goldhaber, 2002; Rivkin, Hanushek, & Kain, 2005). Director of the Institute of Education Sciences Grover J. Whitehurst reported—drawing on research by Jordan, Mendro, and Weerasinghe (1997); Rowan (2002); and Sanders and Rivers (1996)—that a student who has effective teachers for three straight years is likely to score more than 50 percentile points higher on standardized tests than a student who has ineffective teachers.

So, how can policymakers increase the likelihood that students consistently have effective teachers? We contend that one of the most direct systemic changes we could make to reach this goal is to change the compensation structure for educators, moving toward some form of merit pay. Our support of merit-based compensation stems from our experiences evaluating the Achievement Challenge Pilot Project (ACPP) in Little Rock, Arkansas, and advising school districts and policymakers who are considering similar reforms.

How Are Teachers Currently Compensated?
The single-salary system, which operates in 95 percent of all public schools in the United States (Protsik, 1996), rewards teachers financially on the basis of each teacher's years of experience and number of degrees. This system assumes that teaching ability improves with more years of experience and higher degrees. However, many researchers argue that the benefits of teacher experience plateau after several years and that additional degrees do not translate into enhanced student performance (for example, Goldhaber, 2002; Hanushek, 2003). Thus, the current teacher-compensation structure is not likely to move the field forward in terms of boosting student achievement.

Further, perverse incentives are built into the single-salary system that might exacerbate teacher shortages in key areas and encourage the best teachers to leave the system. Because the current system includes no monetary rewards directly tied to effectiveness, many effective teachers seek more “compensation” through better working conditions, often choosing to leave schools with a high population of disadvantaged students and challenging teaching conditions for schools serving more advantaged students. Good teachers might also increase their compensation by leaving the classroom for an administrative position or, worse, leaving the field of education entirely.

Finally, the current system gives policymakers no freedom to address teacher shortages in such key areas as special education and upper-level math and science by adjusting salaries upward for these particular groups. In the current structure, teachers in saturated areas are essentially overpaid, whereas teachers in critical shortage areas are underpaid.

Merit Pay as a Lever for Change
Any policy change aimed at retaining effective teachers must alter the structure of pay, not just the magnitude. Across-the-board raises will not fix the problems inherent in the current uniform salary structure. We believe policymakers should consider using merit pay as a tool to recruit and hold on to effective teachers. Such programs could improve the teaching corps in three ways.

First, the possibility of financial rewards tied to increased student achievement would motivate existing teachers to focus increased effort and innovation on student achievement.
Second, merit-pay programs would draw a more talented pool of candidates into teaching because individuals who were confident that they would thrive under a pay-for-performance arrangement would be more likely to enter the profession.

Third, under a merit-based scheme, the most effective teachers would consistently earn large bonuses and ineffective teachers would earn smaller bonuses. Through natural selection, more competent teachers would remain and less effective ones would leave.

One of the sticking points in implementing merit-pay plans is defining “merit,” and deciding when and to whom we should provide bonuses. What should be rewarded: Teacher characteristics? Teacher behaviors? Measurable student achievement? There is no single ideal merit-pay plan; plans can and should be tailored to each individual situation. Merit-pay plans using differing criteria and reward systems have been implemented and evaluated around the United States. Although the research base is small, evidence suggests that compensation plans that reward teachers for student achievement lead to improved student performance (Figlio & Kenny, 2006; Ladd, 1999; Podgursky & Springer, 2006).

In our evaluation of the Little Rock's Achievement Challenge Pilot Project, we saw substantial learning gains on standardized tests for students in participating schools. The research literature on merit pay often highlights the opposition such plans meet and the negative feelings teachers have about merit pay. During the three years it operated, we did not find that the Little Rock project created a negative atmosphere among teachers, however—quite the opposite.

A look at the effect of the Little Rock program on participating schools shows how a well-constructed plan can capitalize on the potential of merit pay to enhance student achievement.

How It's Working in Little Rock
In the 2004–05 school year, the performance-based pay plan was implemented at Meadowcliff Elementary School in Little Rock. It expanded to four other district elementary schools by the 2006-07 school year. The schools chosen for the project had a high percentage of ethnic minority students and a high percentage of students who qualified for free and reduced-price lunch. During the first year, a private foundation funded the project; in the subsequent two years, Little Rock School District funded the plan at some schools, and private funders (the Brown Foundation, Hussman Foundation, and Walton Family Foundation) did so at other schools.

Teacher merit, in this plan, was determined exclusively by student achievement gains on the Stanford Achievement Test or the Iowa Test of Basic Skills. Students' gains were calculated on the basis of the change in normal curve equivalent scores from the pre-test to the post-test. One pilot school administered the pre-test in the fall, and the other four pilot schools administered pre-tests in the spring of the school year before the post-test. All post-tests were given in the spring. Student growth was also compared to growth of comparable students at nonparticipating schools.

Teachers used the results of the pre-test to target instruction toward students' needs throughout the year. At most of the participating schools, merit bonuses for classroom teachers were based on the average learning gains for all students in their classrooms. Merit bonuses for administrators, supplementary teachers, and other staff were based on the learning gains of the school as a whole.1 
An evaluation after the first two years of the Achievement Challenge Pilot Project showed that schools implementing the program achieved average gains of approximately seven percentile points for students in mathematics and reading. Scores of students in the pilot schools improved, whereas those of students in comparison schools decreased.

The evaluation also compared teacher attitudes in schools implementing the pilot with teacher attitudes in Little Rock schools that did not participate, matched for student characteristics. Teachers in the pilot schools reported increased satisfaction with their salaries without increased feelings of negative competition or a sense of a negative environment. Although pilot teachers did not indicate that they were working harder or were more innovative than comparison teachers indicated, they were more likely to view teaching low-performing students as an opportunity to demonstrate their teaching ability, and less likely to view teaching low-performers as “a burden.” Perhaps most important, the teachers who participated in the pilot project indicated that their students' performance improved, which suggests that teachers saw measurable academic improvement.

Characteristics of a Successful Program
Not all merit-pay plans are created equal; the devil is, in fact, in the details. The Achievement Challenge Pilot Project includes the four following strategies that we believe will foster success in any performance-based bonus program.
1. Use a Straightforward Formula
A key criticism of merit-pay plans is that teachers view these plans as secret formulas and have no idea why merit bonuses were or were not awarded. If a merit-pay plan is to successfully elicit excellent teaching, it must be based on a transparent payout scheme and a clear understanding of what kind of outcomes the plan rewards.
2. Tie Merit Pay to Student Growth
As evidenced by the U.S. Department of Education's recent move to allow states flexibility in using growth models to calculate average yearly progress, there is a growing consensus that indicators of teacher effectiveness should be based on student growth rather than on predetermined levels of performance. Any plan that determines merit by absolute achievement levels risks creating an incentive for teachers to leave behind students in low-performing schools or classrooms and head for the higher-performing hills.
3. Build in Incentives for Collaboration
The fear that poorly designed merit-pay plans will create unhealthy competition among teachers is a legitimate one, particularly with any plan that divides a fixed amount of reward dollars among a set number of teachers. We recommend that plans foster collaboration in two ways.
First, allocate awards to teachers based on the extent to which they meet their individually set goals, regardless of how other teachers perform. This creates a far better situation than one that pits teachers against one another. Second, include a reward component based on schoolwide student improvement. The “merit” score for classroom teachers with tested students should be based, in part, on schoolwide student achievement growth. In this type of plan, teachers have an incentive to focus on not only the achievement of students in their own classrooms, but also on the achievement of all students.
4. Give Substantial Awards
Successful merit-pay plans should give out significant bonuses to teachers who accomplish their goals. Education officials cannot expect teachers to shift their behavior toward excellence for the proverbial peanut bonus. In the first year the Arkansas project was implemented, bonuses for individual teachers and support staff ranged from $1,800 to $8,600.

The Key Question
Our experiences with the ACPP plan and performance-based plans in other districts have convinced us that it is possible for such plans to be fair and to create incentives that push teachers to focus on student learning growth. In fact, it's not at all clear that the current uniform-pay scale is fair to teachers—or to students. In any case, the objective of performance-based pay is not to create a salary scale that is “fair” to all teachers. Ultimately, the key question of whether a merit-pay system is worth pursuing is, Does the system increase student achievement and encourage effective teachers to remain in the field? If so, this is the type of system we want.
Barnett, J. H., Ritter, G. W., Winters, M. A., & Greene, J. P. (2007). Evaluation of year one of the Achievement Challenge Pilot Project in the Little Rock Public School District (Department of Education Reform Working Paper). Fayetteville: University of Arkansas. Available
Figlio, D. N., & Kenny, L. (2006). Individual teacher incentives and student performance (NBER Working Paper 12627). Cambridge, MA: National Bureau of Economic Research. Available:
Goldhaber, D. (2002). The mystery of good teaching. Education Next, 1, 50–55.
Hanushek, E. A. (2003). The failure of input-based resource policies. Economic Journal, 133(485), 64–68.
Jacobson, S. L. (1988). The distribution of salary increments and its effect on teacher retention.Educational Administration Quarterly, 24, 178–199.
Jordan, H. R., Mendro, R., & Weerasinghe, D. (1997, July). Teacher effects on longitudinal student achievement: A preliminary report on research on teacher effectiveness. Paper presented at the National Evaluation Institute, Indianapolis, IN.
Ladd, H. F. (1999). The Dallas school accountability and incentive program. Economics of Education Review, 18, 1–16.
Protsik, J. (1996). History of teacher pay and incentive reforms. Journal of School Leadership, 6,265–289.
Podgursky, M. J., & Springer, M. G. (2006). Teacher performance pay: A review. Nashville, TN: National Center on Performance Incentives. Podgursky_and_Springer_2006_Revised.pdf
Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005). Teachers, schools, and academic achievement.Econometrica, 73(2), 417–458.
Rowan, B. (2002). What large-scale, survey research tells us about teacher effects on student achievement: Insights from the prospects study of elementary schools. Ann Arbor: University of Michigan.
Sanders, W., & Rivers, J. (1996, November). Cumulative and residual effects of teachers on future student academic achievement. Knoxville: University of Tennessee Value-Added Research and Assessment Center.
1  In the last year of the pilot, some schools also measured growth through students' scores on the Iowa Test of Basic Skills. See Barnett, Ritter, Winters, & Greene (2007) for more details on the program and teacher payouts.

Joshua H. Barnett is Assistant Research Professor at Arizona State University;

Gary W. Ritter is Director of the Office for Education Policy at the University of Arkansas at Fayetteville;


No comments:

Post a Comment

Note: Only a member of this blog may post a comment.