Teachers Grapple with Yet Another Marketplace Reform

By Bob Peterson

photo by Jean-Claude LeJeune

When each teacher at East Chapel Hill High School in North Carolina received a $1,500 “merit pay” check this year, the teachers did a rather odd thing. They gave some or all of their checks to schools they believed were more deserving of extra resources.

“We felt that there was some inconsistency to protesting the [state standardized] test and then pocketing the money,” English teacher Martha Dill told The Chapel Hill News. “So we tried to think of what would be a meaningful response.”

The response may be unusual. But if corporate and policy leaders in state capitals across the nation have their way, similar merit pay schemes will soon be widespread. Almost half the states have passed or are considering legislation based on merit pay, according to Education Week.

The concept of “merit pay,” often referred to as “pay-for-performance,” rests on the premise that existing pay structures for teachers are not sufficiently guided by market-oriented business principles. The argument is that teachers who are more “productive” – who do a better job teaching – should receive higher pay. Further, the belief is that financial bonuses can push teachers to work harder and, in the process, can raise student academic achievement.

Currently, teachers in most districts receive the same base pay, with increases generally based on seniority and on attaining additional college credits or their equivalents.

The pay-for-performance plans are not without their critics, especially among teachers. Some complain that the proposals inappropriately impose business productivity structures onto the field of education, as though educating a human being can be as easily measured asassembling a car. Others note that merit pay plans are divisive, pitting teacher against teacher, especially teachers in affluent districts or schools against teachers in low-income districtsor schools. Others caution that most current plans are inherently flawed because they measure results based on standardized tests.

Mark Simon, president of the Montgomery County (MD) Education Association, is one of many teacher activists who is skeptical of pay-for-performance. “Merit pay is a solution in search of a problem,” he told Rethinking Schools. “The problem in public education is not that teachers don’t work hard enough or that teachers are not sufficiently motivated. Even a lot of teachers who aren’t doing a great job are still working their tails off.” The key question, according to Simon, is how can schools “attract, maintain, and develop high-quality teachers – especially given the complex nature of teaching.”

Some progressive union officials are trying to forge a third path, one that moves beyond existing pay systems but doesn’t rely on the business community’s approach to “merit pay.”

“We’ve been experimenting with ‘differentiated pay’ in Rochester for 15 years,” Adam Urbanski, president of the Rochester (NY) Teachers Association told Rethinking Schools. “Our choice right now is between two destructive proposals – the lock-step status quo and merit pay. I don’t support either. We need something different.”

Urbanski, who also heads up the Teacher Union Reform Network (TURN), a grouping of 21 progressive NEA and AFT locals, is encouraging his union colleagues to experiment with changes in compensation systems.


Momentum for merit pay plans got a boost in November at the 1999 National Education Summit of governors and business leaders. The summit called on districts to implement “pay-for-performance incentive plans … based on lessons learned from the private sector.”

On the state level, political leaders have responded to the summit’s call. Wisconsin’s governor, Tommy Thomp-son, National Education Summit co-chair, announced a typical plan in his state-of-the-state address in January. Thompson proposed an “Award for Achievement” program in which teachers at schools that test 95% of their students and show year-to-year gains could receive “financial awards of up to $3,000.”

The media have also promoted merit pay. When Denver teachers approved a new contract last fall that included provisions to experiment with three types of merit pay, the contract was top news on television stations and in newspapers across the country.

Although rarely noted in the media, merit pay proposals are not new. As educational historians Larry Cuban and David Tyack point out in the article on page 11, merit pay proposals in the United States have been around since World War I – and have invariably clashed with the efforts of teachers to ensure that “the fundamental intrinsic equality of all good teaching” is recognized.

Bob Chase, head of the National Education Association, and Sandra Feldman, head of the American Federation of Teachers, both emphasize that the key issue is ensuring that all teachers are paid commensurately with their training. In a speech this January, for instance, Chase noted that unionists are understandably wary of pay-for-performance. “We remember past experiments with merit pay – compensation schemes that pitted teacher against teacher,” he said. “We know that every one of those merit pay schemes was a counter-productive disaster.”

Chase, however, also noted that pay-for-performance, at its best, is about teachers cooperating, improving their skills, and increasing their compensation. “So my counsel regarding pay-for-performance is for all of us to keep an open mind, to listen, and to learn from local affiliates that are trying new things,” he said.

Union leaders have been most critical of plans that focus on individual teachers and that are tied to student academic performance. “Were main opposed to individual pay tied to student results,” Tom Mooney, the president of the Cincinnati Federation of Teachers and a vice president of the AFT, told Education Week. “I think that’s unsound, unprofessional, unethical, and no one has shown me a system that’s even halfway credible.”


Historically, merit pay proposals focused on individual teachers, rather than on teachers in an entire school. They also rested on the subjective evaluations of supervisors – although as testing became more widespread, bonuses were based on student gains in test scores. Both evaluation approaches have such serious drawbacks, however, that Harvard University researchers Richard Murnane and David Cohen concluded in a 1986 study that neither “is an effective strategy for motivating teachers to achieve high performance levels.” 1

Many proponents of current pay-for-performance proposals distance themselves from past practices. Allen Odden, a professor of education at the University of Wisconsin-Madison and co-author of Paying Teachers for What They Know and Do (Corwin Press, 1997), has called merit pay plans for individual teachers a “Model-T version.”

Criticisms of individual merit pay plans have led to two increasingly popular variations. One is to give bonuses to entire school staffs based on the schools’ gains in achievement, usually measured by standardized test scores. Several states, including Georgia, Kentucky, and North Carolina, have implemented such plans. In some cases, the financial incentive goes to the school in order to buy more resources; other times it goes directly to the teachers as salary bonuses.

The other variation – which is distinct enough from corporate plans that some do not even refer to them as merit pay proposals – is to differentiate pay on the basis of knowledge, skill, and responsibility. This variation attempts to deal with perhaps the most powerful criticism of most status quo teacher compensation plans: that the quickest way to move up in responsibility and pay is to leave teaching and become an administrator.

This variation, which has been adopted by several unions associated with TURN, goes by various names: “career ladders,” “lead teachers,” “professional development,” or “skill development.” It includes those programs providing additional money to teachers certified by the National Board for Professional Teacher Standards.

Some programs mix together the various approaches. The recent Denver contract, for example, calls for a two-year pilot program that experiments with schoolwide performance pay based either on the Iowa Tests of Basic Skills, or on criteria-referenced or other teacher-created measures, or on teachers increasing their knowledge of and skill with teaching and learning.


Many of the current proposals are too new to adequately evaluate. But a look at the problems with earlier incarnations of merit pay is useful, because it reveals some of the questionable assumptions behind the entire merit pay approach.

Murnane and Cohen, for instance, argue that under some conditions performance-based pay can be an efficient method of compensating workers – but that teaching is not an activity that meets such conditions. As a result, even though merit pay is an “old idea” that has been around for decades, it never really caught on. In their 1986 study, they noted that “the majority of districts that reported having tried and dropped merit pay indicated that their plans lasted less than five years.” They further note that “teacher union resistance cannot account for the demise of most merit pay plans, for most plans predated unions or failed in non-union districts.”

Merit pay plans fail to adequately address a number of key issues, including:

  • The nature of teachers’ work.
  • The nature of teacher motivation.
  • Inadequate support for teachers, especially new teachers.
  • Problems with standardized tests as a measurement.
  • How merit play plans can exacerbate inequalities.

The nature of teachers’ work. About 30% of U.S. companies have pay-for-performance plans. Some jobs lend themselves to such arrangements. In other situations, the arrangements are used to pressure workers to be more productive or to “speed up” – which can result in favoritism or in work place safety problems.

Few existing business models can be easily grafted onto education. Researchers have pointed out the stark difference between a classroom and, for instance, a laundry where shirts need to be ironed. One can easily count the number of ironed shirts in a period of time. Measuring the effectiveness of teaching is not as easy. 2

Because of the imprecise nature of teaching, Murnane and Cohen found that historically, supervisors using merit pay systems had a hard time answering the question, “Why did teacher X get merit pay and I didn’t?” They also found it difficult to answer the question, “What can I do to get merit pay?”

The nature of teacher motivation. The underlying assumption of pay-for-performance is that bonus pay incentives will be a key factor in motivating teachers.

Few would deny that teachers are underpaid. According to an Education Week study of U.S. Census Bureau figures, in 1998 teachers ages 22 to 28 earned an average of $7,894 less than other college graduates of the same age. The gap is three times greater for teachers ages 44 to 50, who earned $23,655 less than their peers in other occupations. Further, few teachers would refuse the chance to receive pay that is commensurate with their level of education. In this regard, a more logical reform would be to raise the base pay for all teachers.

Increasing the motivation of those who have decided to become teachers (and thus have already accepted a lower salary than their peers from college) rests more on working conditions and job satisfaction than on bonuses. Rather than merit pay, this calls for reforms such as smaller class sizes, more planning and collaboration time, more resources, and more social service support for students. These, in turn, potentially can lead to greater success in educating students.

“Teachers want to go home at night feeling they’ve been successful,” said Michael Charney, a teacher-union activist in Cleveland. “We won’t get that through some merit pay scheme, but rather through a political struggle that secures adequate resources for poor and African-American kids.”

Inadequate support for teachers, especially new teachers. Far too often, new teachers are forced to endure the “sink or swim” philosophy. This leads to one out of five novice teachers leaving the profession after three years, according to a recent survey by Education Week. Other studies show that one-third leave within the first five years.

One reason teachers leave is that they are asked to do far too much. Schools are increasingly seen as the main institution dealing with a host of daunting problems – everything from drugs and alcohol, to gangs, to lack of daycare and recreation facilities, to sexual and health education. This burden primarily falls on classroom teachers. There is a lack of institutional support and resources, especially in poor urban and rural communities.

Further, when budgets are tight, teacher mentoring, assessment, and staff development programs are often the first to be cut or scaled back.

Problems with standardized tests as the measurement. Many current merit pay plans peg bonuses to “measurable” outcomes such as test scores. This leads to obvious problems, however, even if such measures are tied to schoolwide results. Kent Jonesand Betty Lou Whitford have analyzed the Kentucky Education Reform Act, which includes pay-for-performance tied to schoolwide test scores. They argue that such a “high-stakes” system “forces teachers to focus on whatever is thought to raise test scores rather than on instruction aimed at addressing individual student needs.”3

As one superintendent from a rural Wisconsin district told Rethinking Schools, “I already have enough pressures on my staffs to teach to the tests. I don’t need financial incentives as well.”

As standardized tests take on increasing importance, teachers have even less control over their professional lives and less incentive to develop their skills. In Chicago, for example, the central administration provides scripted teaching lessons that tell teachers what to say every day and what page the students should be on.

Urbanski argues that if success is defined solely or principally by test scores, “it becomes a witch hunt and leads to staff deprofessionalization and will be harder for students in the long run.” He believes that student outcomes and tests – “assuming they’re decent tests”- ought to “be used as information but not automatic verdicts on assessing teaching.”

Urbanski also notes that relying on test scores “will inevitably lead to cheating.” In fact, large-scale testing scandals recently led to the resignation of an associate superintendent in Austin, TX; there was also a massive “cheating” controversy in New York City.

Moreover, when there is pressure to increase the number of students scoring well on certain tests, there are strong incentives for educators to ignore both students who are likely to fail the test, no matter what, and students who are likely to pass the test, no matter what. Such a mentality inevitably distorts teaching. In addition, schools have many goals. Focusing only on those curricular areas covered by standardized tests inevitably narrows what students will learn.

“Even if tests were developed that provided accurate measures of students’ skills in particular subject areas,” Murnane and Cohen point out, “incentives to allocate time strategically to particular students and particular subject areas and to neglect aspects of the job not measurable by standardized tests would still remain.”

Reports by parents and teachers that recess, art, and music classes have been scaled back for the sake of “test preparation” are mainly anecdotal, but portend an ominous development. Parent activist Makani Themba-Nixon, a school council member at an elementary school in Roanoke, VA, told Rethinking Schools that she has heard teachers talk about how they don’t want certain kids in their classes or school because it will “bring the test scores down.” She also has expressed concern about how narrowing the curricula affects multicultural teaching (see p. 22).

Pegging merit pay to schoolwide results sometimes exacerbates the problem of relying on standardized tests. A high school teacher from North Carolina said if it were only her bonus that was at stake, she could have withstood the pressure to change curriculum to meet the needs of the test. But under her state’s schoolwide pay-for-performance, she felt a sense of responsibility to her colleagues and didn’t want to jeopardize her standing with other staff members.

Exacerbating existing inequalities. Student achievement is affected by many factors outside the classroom. Linda Darling-Hammond, professor of teaching and teacher education at Stanford University, writes that “student characteristics such as poverty, non-English-speaking status, and minority status are negatively correlated with student outcomes, and usually significantly so.” She adds that “these student characteristics are also significantly and negatively correlated with the qualifications of teachers; that is, the less socially advantaged the students, the less likely teachers are to hold full certification and a degree in their field and the more likely they are to have entered teaching without certification.” 4

Given these current social realities, an over-emphasis on merit pay could easily lead to two problems. First, more qualified teachers might go in even greater numbers to schools where the students are more advantaged – thinking students can more easily make “gains”resulting in the “merit” pay. Second, disadvantaged schools will be under even more pressure to raise test scores, transforming them into little more than “test-prep” academies and, in the process, yet again short-changing low-income students.

An alternative reform would be equitable and adequate resources for all schools. Funding inequities are so severe, for example, that litigation and legislation are pending in over two dozenstates.


It’s safe to predict that pay-for-performance proposals will increase in the near future. At best, such plans might highlight the relatively low wages that teachers receive as professionals. At worst, they will divert attention from more important issues and intensify preoccupation with standardized test scores.

Given the momentum of pay-for-performance, stopping such initiatives will not be easy. It may be tempting to adopt a strategy of “just say no” to merit pay and similar initiatives. Rochester’s Urbanski, however, believes progressives must carve out a middle road that does not become a defense of the status quo, but he doesn’t see merit pay as the only alternative. “If you don’t have an agenda in times of turmoil, then you’ll have to work off of somebodyelse’s draft, ” he notes.

The Cincinnati Federation of Teachers, under the leadership of Tom Mooney, has shown a particular willingness to experiment with pay structures. The Cincinnati union, for example, has pioneered an alternative compensation system in which “lead” teachers get paid an additional $5,000 per year for working a longer school year and taking on additional responsibilities during the regular year (see article – page 8).

Several unions are also struggling with the related issue of evaluating and improving teacher performance, whether or not that leads to differences in pay structures. In particular, local unions are experimenting with peer review programs, such as those in Rochester or Cincinnati, or a professional development model in Minneapolis which incorporates peer review with strong collaboration with the administration. (For a description of these programs, see Transforming Teacher Unions, edited by Bob Peterson and Michael Charney, Rethinking Schools, 1999.)

Urbanski hopes it is possible to channel interest in pay-for-performance into compensation plans that reward enhanced skills and knowledge. But even this agenda has its complications.

Simon, of the Montgomery County Education Association, cautions that “any changes in pay schedules have to be developed very carefully and collaboratively so that teachers perceive the new plans as fair and are open to them. If it is viewed as punitive, discriminatory, arbitrary, or creating further insecurity, it will fail.”

In 1983, the national report, A Nation at Risk, called for public school teachers’ salaries to be made “professionally competitive, market-sensitive, and performance-based.” In the last 17 years they certainly haven’t become “professionally competitive” but it appears that there is now an increased push for them to become “performance-based.” Let’s hope that the spirit that encouraged East Chapel Hill High School teachers to give money to schools with fewer resources might spark a movement that will demand adequate resources for all schools and that will allow teachers to do their best based on a commitment to children, not a few extra dollars of bonus pay.

Bob Peterson ( is an editor of Rethinking Schools and teaches fifth grade at La Escuela Fratney in Milwaukee.


1 Richard Murnane and David Cohen, “Merit Pay and the Evaluation Problem: Why Most Merit Pay Plans Fail and a Few Survive,” Harvard Educational Review 56 (February 1986): 1-17.

2 This is not to imply that even in such a blue-collar workplace pay-for-performance is necessarily a good idea, as it could easily lead to favoritism and unsafe working conditions.

3 Ken Jones and Betty Lou Whitford, “Kentucky’s Conflicting Reform Principles: High Stake School Accountability and Student Performance Assessment,” Phi Delta Kappan (December 1997): 276-281.

4 Linda Darling-Hammond, “Teacher Quality and Student Achievement: A Review of the State Policy Evidence,” (Washington, DC: Center for the Study of Teaching and Policy, October 1999): 33