In their quest to improve student achievement, educators and policymakers have once again seized on the idea of paying teachers based on performance. Citing the growing evidence of the importance of high-quality teaching, the advocates of revising teacher pay have argued that the single salary scale, which has dominated education for nearly a century, fails to attract the best individuals to the profession or to encourage teachers to excel.
The issue has become so popular that both of the candidates in the 2008 presidential election endorsed some form of performance-based pay. And the American Recovery and Reinvestment Act of 2009 includes $200 million for the Teacher Incentive Fund, which will provide grants to states and districts to develop performance-based compensation systems for teachers and principals.
While the high-level interest is new, the idea of revamping the way teachers are paid has been on the education-policy agenda for a quarter century, and it has been hotly controversial. Numerous states and districts have launched pay-for-performance experiments over the past two decades, but most have withered away in the face of intense criticism. The critics, including the major teachers’ unions, say the methods used to judge teacher performance are unfair because they may reflect factors outside teachers’ control. They also argue that rewards for individual teachers destroy cohesion and collegiality in schools. These critics have been more supportive of plans to provide additional pay or increased responsibilities for teachers who demonstrate superior knowledge and skills, or to offer bonuses to whole schools for improving performance.
|Questions school board members should ask
- Why do we want a pay for performance program? Recruitment? Retention? Student achievement? Hard-to-staff schools?
- Can we sustain the funding? How much money will prove an incentive to teachers?
- Are teachers, principals, and others involved open to conversation?
- Does the school district have the data needed to set up fair measurement and evaluation systems?
As with many issues in education, the intensity of the debate over performance-based pay reflects the relative lack of clear evidence about its effects. And in many cases, poorly designed experiments—plans that were underfunded and did not reward all teachers who qualified, or those that did not provide clear criteria for rewards—strengthened critics’ hands and led to the plans’ demise.
In recent years, though, research on the topic has started to emerge, and policy makers are in a better position to make informed judgments about the idea. This research suggests that performance-based pay can have a positive impact on student achievement, although the effects are relatively modest. However, there is little evidence about what type of rewards are most effective. And there is little long-term research to show whether these types of programs do, in fact, attract better-qualified individuals to teaching or keep them in the profession.
A long-standing debate
The compensation system used in most school districts emerged in the 1920s and has remained intact ever since. Under that system, entering teachers receive the same salary; they can earn increases through additional years of experience or through additional coursework or advanced degrees.
The so-called single salary scale was intended to eliminate favoritism and inequities by ensuring that all teachers were treated the same. Yet the reform movement that began in the 1980s zeroed in on the pay system as an impediment to improvement. A Nation at Risk, the 1983 document that helped launch the reform movement, made a strong case for performance-based pay: “Salary, promotion, tenure, and retention should be tied to an effective evaluation system so that superior teachers can be rewarded, average ones encouraged, and poor ones either improved or terminated” (National Commission on Excellence in Education 1983).
In the wake of that report, half the states adopted some form of performance-based pay system as part of their reform plans. One of the first to do so was Florida, which created a “Master Teacher” program that provided bonuses to teachers who earned high scores on tests and evaluations. However, the program attracted strong opposition from teachers’ unions, who contended that the evaluation system was arbitrary and filed suit to block it. In 1986, the program was recast as a “career ladder” system. In this iteration, it provided additional pay to teachers who took on increased responsibilities, but it remained unpopular and died in 1988 when the legislature declined to fund it.
The Florida program’s fate was typical of performance-based pay plans in that period. One study of district-based plans found that three-fourths of the programs that had been in place a decade earlier were no longer in existence when the study was published in 1994 (Hatry, Greiner, and Ashford 1994). Moreover, the study found, there were virtually no evaluations of the programs to determine their effects.
Despite these meager results, support for some form of performance-based pay continued, and opposition began to soften. In 1985, Albert Shanker, then the president of the American Federation of Teachers, called for rewarding exceptional teachers and strongly endorsed the creation of the National Board for Professional Teaching Standards. The National Board developed rigorous assessments of teachers’ knowledge and skills and certified those who demonstrated high abilities. Nearly all states and many districts offer bonuses to teachers who earn the credential. To date, about 74,000 teachers have become certified by the National Board.
The advent of standards-based reform brought on another type of performance pay: Rewards for schools that show improvement on test scores. Several of the first systems to adopt standards, assessments, and accountability—notably Kentucky and South Carolina—included school-based rewards as part of their accountability systems.
More recently, the idea of rewarding individual teachers for their performance has re-emerged, in part because of the growing recognition of how important high-quality teaching is to improving student performance. In addition, new statistical techniques have enabled researchers to calculate the “value added” by teachers. These techniques compare students’ growth in performance to their prior performance. Since this evaluates gains in student achievement, regardless of students’ prior level of performance, these techniques have answered some of the objections teachers have raised to evaluation methods.
The results from performance-based pay
The growing use of performance-based pay experiments in a number of districts and schools has yielded new research that can suggest whether the policy does in fact achieve the effects supporters claim.
One of the first large-scale studies, by David N. Figlio and Lawrence W. Kenny from the University of Florida, offers some positive results (Figlio and Kenny 2006). Examining a national database and their own survey of districts, they found that test scores are higher in schools with some form of financial incentive for teachers. However, the study could not determine that performance pay produced better student outcomes. It could be that higher-performing schools tended to be the ones that adopted the financial incentive programs.
Another large-scale study surveyed the scant literature on performance-based pay programs, including programs in other countries, and found that the evidence generally shows that such programs do improve student achievement (Podgursky and Springer 2006). However, the study notes that there is little evidence about the kind of incentives that are effective, nor is there evidence about whether such programs in fact attract more qualified people to teaching or help raise the level of teacher quality over time.
Smaller-scale studies of particular programs offer a more mixed picture. One of the best-known performance-pay programs, Denver’s Professional Compensation (ProComp) Plan, underwent an extensive pilot that was evaluated before the program was implemented district-wide. Under ProComp, teachers receive higher pay (not just bonuses) if their students meet or exceed performance goals that are set by the teacher and principal; if student test scores improve; if they receive favorable evaluations; if they complete an approved professional development unit; and if they teach in a hard-to-staff school. An evaluation of the pilot found that it had an impact on student achievement: Specifically, students whose teachers had higher student growth objectives scored higher on standardized tests (Community Training and Assistance Center 2004).
- ProComp: Denver Public Schools’ teacher incentive program.
- TAP: A national program offering multiple career paths and advancement opportunities to teachers.
Studies of the Teacher Advancement Program (TAP), an initiative now in 180 schools in more than dozen states, also show some positive impacts, but the benefits are not universal. Under TAP, teachers receive higher pay and additional responsibility if they demonstrate effectiveness through value-added measures of student achievement and extensive evaluations. A study by the program’s researchers found that TAP teachers in six states generated bigger gains in student achievement than comparable teachers in non-TAP schools, and that TAP schools were more likely than non-TAP schools to show more than a year’s worth of achievement gains in a single year (Solmon, White, Cohen, and Woo 2007).
However, an independent analysis of TAP found that the gains were obtained only in elementary schools. Examining mathematics test scores in 1,200 schools in two states, the study found that TAP schools outperformed non-TAP schools in grades two through five; in grades six through ten, the non-TAP schools outperformed those that participated in the program (Springer, Ballou, and Peng 2008).
The results from paying for knowledge and skills
Some critics of performance pay have argued that rewarding teachers for demonstrating superior knowledge and skills is more effective than methods that use student achievement as a measure of teacher quality. Teachers have more control over their own knowledge and skills, and such approaches, if done well, establish clear standards for teacher practice.
The largest and most closely watched effort to reward teachers for knowledge and skills is NBPTS. An extensive review of the literature on the program found that, overall, students taught by National Board Certified teachers score higher on standardized tests of reading and mathematics than do comparable students taught by teachers who are not Board certified. However, the effects were not uniformly large. And the study found that NBPTS-certified teachers tend to disproportionately teach more advantaged students, and that there is little evidence that the program has had the “spillover effect” of improving the quality of teachers throughout the system (National Research Council 2008).
Research on individual district programs that reward teachers who demonstrate superior knowledge and skills shows modest and varied effects on teacher practice. A longstanding program in Douglas County, Colo., for example, provides “outstanding teacher” awards to teachers who show success in the classroom, in part using NBPTS standards. The program also provides group awards, and the district has recently instituted a “master teacher” award for teachers whose students have demonstrated improvements in achievement. An evaluation of the program found that it had a modest effect on retaining teachers, but did not affect the recruitment of new teachers at all; the qualifications of teachers entering Douglas County were no different from those in other districts (Reichardt and Van Buhler 2003).
Other studies have examined districts that use performance-evaluation systems that are based on standards for teaching practice, a form of knowledge- and skill-based pay. These studies found that the systems affected classroom practice, but these effects tended to be “broad, but shallow” (Heneman et al. 2006). For example, teachers reported engaging in more reflection and improving their classroom planning, but they did not report significant changes in pedagogy.
The results from paying group awards
Group awards are aimed at creating incentives for entire staffs to raise school performance. In contrast to individual awards, group awards are intended to foster collaboration among staff members and encourage everyone to see that he or she has a stake in improving student performance.
Although there is little research to show the effects of group awards on teachers or student performance, a study of a longstanding program in Dallas, Tex., suggests that such awards can have positive effects on student outcomes. Under a program created in 1990, the district awarded bonuses worth $1,000 each to teachers and principals in schools judged to have increased student performance the most. The study found that, in the years after the program was implemented, Dallas achievement rose faster than in other comparable Texas cities (Clotfelter and Ladd 1996).
However, the Texas study also hinted at some potential negative effects. Anecdotal reports suggested that some schools focused heavily on tested material at the expense of other areas of the curriculum, and there were some incidents of outright cheating.
In addition, studies of other group award programs suggest that the programs might contribute to high teacher turnover in schools that do not win awards.
Success lies in design
The small knowledge base about performance-based pay suggests that the policy shows some promise as a way of improving student performance. So why have so many experiments petered out?
Research on some of the experiments, along with surveys of teachers, points to the programs’ design. Based on that evidence, researchers at the Center for Policy Research in Education suggest some guidelines districts and schools should consider as they develop pay-for-performance programs (Heneman, Milanowski, and Kimball 2007). These guidelines include:
- Guarantee stable and adequate funding. Some plans have failed because funding for them dried up, and teachers grew skeptical that they would earn a bonus. By ensuring adequate funding, districts can ensure that the reward system serves as an incentive for teacher practice.
- Determine and provide competitive compensation. In some cases, the awards were relatively small and did not create enough of an incentive for teachers to change their practice or to attract new teachers to a district. Some private schools have offered bonuses of up to 11 percent of base pay, compared with about 2 percent in public school programs (Ballou and Podgursky 1997).
- Build strong measurement systems. Many pay-for-performance systems have foundered because teachers considered the ways they were evaluated arbitrary and unfair. A program should use multiple measures for teacher evaluation. Value-added measures offer promise as a way of assessing student growth, but they require sophisticated data systems to link student test scores to teachers.
In designing the pay-for-performance plans, districts should be sure to include principals and administrators, and engage teachers’ unions. Districts should start slowly with pilot programs in order to build capacity and monitor the program’s implementation and effects. If the experiments that are now under way follow these guidelines, we might have better answers in the next few years about ways to improve teacher quality.
This article was written by Robert Rothman for the Center for Public Education. Rothman is a principal associate and senior editor of Voices in Urban Education
at the Annenberg Institute for School Reform.
Posted: July 9, 2009
©2009 Center for Public Education