BEA Home Page
BEA Home Page
Rennie Center E-Forum
Report on Merit Pay

Rennie Center E-forum
Differentiated Pay for Teachers

February 2006

The Rennie Center is pleased to announce the first issue of our Rennie E-forum. The E-forum format is designed to foster constructive discourse on key topics in educational improvement. The E-forum presents three distinct viewpoints, representing policy, research and practice. Our goal is to spark informed debate, and ultimately, to find common ground.

The first issue of the E-forum centers on differentiated pay for teachers. The term differentiated pay is shorthand for several, related ideas. It is often used interchangeably with pay-for-performance and merit pay, though differentiated pay carries a broader meaning. Schools and districts might differentiate teacher pay based on a number of separate criteria and for a number of purposes. Differentiated pay might serve as a tool to reward excellence, to attract and retain high quality teachers, and/or to enhance the professionalism of teaching. Pay differentiation models can include pay for teachers who:

With Governor Romney's proposed pay incentive plan for teachers as context, the Rennie Center partnered with the Harvard Graduate School of Education (HGSE) to host an event on differentiated compensation. It explored both the opportunities and impediments associated with reforming the teacher salary schedule to be more market-driven and performance-based.

In this E-forum, we hope to engage a broad audience in this important issue by sharing the perspectives of the panelists from the Rennie Center-HGSE event. The first article details Governor Romney's plan to base teacher pay on student test scores as well as peer and principal evaluations. It is written by the Commonwealth's Chief Economist and Education Advisor, Robert Costrell. Next, Harvard Graduate School of Education Lecturer Katherine Boles and Brandeis University Lecturer Vivian Troen, who have researched and written about the teaching profession, comment on the history of proposed merit-based pay plans and their impact on the profession. Finally, Researcher Donald Gratz, who studied and provided technical assistance for the Pay for Performance Pilot in the Denver Public Schools, will discuss potential lessons to be learned from the Denver experience.

We know this is a contentious issue in the field and expect to engage in additional work on this topic in the coming months. It's our hope that you will find this inaugural E-forum issue useful in informing your thinking on differentiated compensation. We look forward to hearing your comments to help inform our future work.

S. Paul Reville

In this issue:

Governor Romney's Differential Pay Proposals
Robert Costrell

Governor Romney's proposals for differential pay are part of a comprehensive Education Reform bill aimed at closing two gaps: (1) the gap between Massachusetts student performance and our international competition, particularly in math and science; and (2) the achievement gaps within the state, between high-performing and low-performing schools.

Math and Science Bonuses for Commonwealth Teaching Corps

It is difficult to attract and retain as many highly qualified math and science teachers as we need, given the market opportunities for math/science/engineering professionals outside of teaching. Rigid union pay scales forbid differential pay for such hard-to-recruit teachers. Governor Romney's proposal for the state to pay eligible math and science teachers $5,000 bonuses at the end of each school year will enhance recruiting and retention in these key fields.

The bonuses will be available to teachers who join the new Commonwealth Teaching Corps. The purpose of the Corps is to offer an accelerated and attractive path into the classroom for highly qualified math and science teachers (including elementary math specialists), either at the entry level or mid-career. The Corps is open to candidates who have a degree (undergraduate or graduate) in math, science, or engineering, pass the teachers' test in their subject area, and agree to forgo tenure. Veteran teachers can also opt into the Corps, if they meet the same criteria. Entry-level teachers will not be required to undergo teacher preparation programs, but districts must offer them mentoring by veteran teachers " with the state providing $3,000 for mentoring stipends or release time. Corps members are also eligible for an optional 401(k)-type retirement plan, with state match (modeled after the plan available in higher education), an additional attraction both for new teachers and mid-career entrants, many of whom would not serve long enough to benefit from the ordinary pension plan. Corps members will not be required to accumulate professional development points for recertification. To qualify for the $5,000 year-end bonus, Corps teachers must receive a satisfactory teacher evaluation.

This carefully-crafted package led HGSE Professor (and former Dean) Ellen Condliffe Lagemann to submit the following testimony to the Massachusetts Legislature:

"I endorse the Governor's new education bill and am especially enthusiastic about the Commonwealth Teacher Corps. That has a very great potential to enhance teaching throughout the Commonwealth."

Differential pay for math and science is commonplace, not only in the private sector, but also in higher education. It is only in K-12 education that market realities are ignored. The usual objection that morale would suffer by paying different salaries to teachers of different subjects is not compelling: humanities professors serve collegially with their higher-paid counterparts in science and engineering, and they do so on unionized public campuses here in Massachusetts. Since market forces are impersonal, there is no reason for personal affront to be taken.

The imperative to attract more highly qualified math and science teachers led a host of education reformers to endorse Governor Romney's proposal for differential pay at the recent legislative hearings. This included the strong support of the Massachusetts Business Alliance for Education (a long-time key player in ed reform, founded by the late Jack Rennie), the Massachusetts Business Roundtable, and the Massachusetts High Technology Council. In addition, the scientific establishment also endorsed the Governor's proposal, including President Jack Wilson of the University of Massachusetts (co-chair of the Great Schools Blue Ribbon Commission on math and science education), and President Ioannis Miaoulis of the Museum of Science. The Massachusetts Urban Superintendents' Network also testified in favor of "additional financial incentives to recruit teachers in high-need areas," such as math and science.

Performance Pay Bonuses

Performance pay is a well-established principle in most fields. It is a tool both to attract high-caliber individuals and also to motivate higher performance. It is commonplace in higher education, including unionized campuses, but not in K-12, where pay is governed solely by seniority and credit hours " criteria that have no established relationship to performance.

Governor Romney's proposal is to offer $5,000 end-of-year bonuses for teachers in any subject who receive an exemplary evaluation. Up to one-third of a district's teachers would be eligible. The local school district would be required to pay 50% of the bonus, except in turnaround schools " our schools with the greatest need for talented teachers -- where the state will pay 100%.

Most of the controversy around K-12 performance pay surrounds the difficulty of evaluation. The National Council on Teacher Quality (NCTQ) and others recommend multiple measures, including student performance together with evaluations by principals and senior faculty. Value-added measures are the gold standard for data-based evaluation, so they need to be developed here in Massachusetts, but will never be applicable in some subjects and some grades.

Accordingly, Governor Romney's proposal rests on a nuanced, but still crisp standard for evaluation. Under the proposed language, the primary performance standard will be the teacher's contribution to student learning. Evaluation will be based on data (state and/or local) on growth in student achievement " value added " when and where this is possible, as well as peer and principal review. Details are left to each school district, but to enable the district to craft an effective instrument, state law must be changed to remove evaluation from collective bargaining " as is the case in virtually every other field. Evaluation is the linchpin of teacher quality " not only performance pay, but to determine necessary training, appropriate school assignment, and much else " so it is critical to reform evaluation from its current ineffectual status.

Bonuses for Teaching in Turnaround Schools

The most recent contract of the Boston Teachers Union allows the superintendent enhanced teacher assignment powers in five underperforming schools. However, the contract explicitly bars the district from using monetary incentives to bring its best teachers to these most needy schools. It is hard to think of a clearer example where the interests of children take a back seat to the prerogatives of collective bargaining. Governor Romney's proposal will restore the districts' flexibility to use pay incentives for this purpose " a measure vigorously supported by the urban superintendents. In addition, as mentioned above, the state will pay 100% of performance bonuses in these schools (vs. 50% in other schools). These measures (along with a set of other "turnaround" proposals in the bill), will provide additional tools to close the achievement gap.


Differential pay and performance pay are coming to K-12 education. Promising results are emerging from a variety of programs -- the Milken Teacher Advancement Program (TAP), Chattanooga's teacher reward program, an expanding program in Little Rock, as well as programs abroad. (See for a recent survey.) The first plan to emerge in Massachusetts may be in Springfield, due to the work of Superintendent Joseph Burke, designed with the help and advice of progressive union leaders from Ohio, among others.

Governor Romney's statewide proposal was based on the recommendations of the NCTQ (which endorsed the bill) and also the Teaching Commission, a blue ribbon entity consisting of 18 national leaders, including governors from both parties, big city superintendents, the late AFT President Sandra Feldman, and others. The Chairman, Louis V. Gerstner (former CEO of IBM) submitted testimony to the Legislature in support of Governor Romney's proposal. He wrote,

"I am … impressed by the performance pay bonuses included in the bill…when our very best teachers are making exactly the same amount as our very worst, something is profoundly wrong…I don't believe at this point we know of any single perfect way to pay teachers based on performance…But that does not for a moment mean that reform itself is ill-considered…I commend the fact that the bill would re-envision teacher evaluation…It's a sensible, balanced approach."

Robert M. Costrell is Education Advisor and Chief Economist in the Commonwealth's Executive Office for Administration and Finance. He is also Professor of Economics (on leave) at the University of Massachusetts in Amherst.

Merit Pay for Teachers? Not Then, Not Now, Not Ever
Katherine Boles & Vivian Troen

Critics of the uniform salary schedule for American teachers argue that the system both rewards poor performance and fails to motivate outstanding achievement. Merit pay " a scheme that has had a perfect record of failure for nearly 300 years " is often proposed as a remedy.

The first known instance of merit pay, sometimes called pay for performance, surfaced in England around 1710. Teachers' salaries were based on their students' test scores on examinations in reading, writing, and arithmetic. The result was that teachers and administrators became obsessed with financial rewards and punishments, and curricula were narrowed to include only the testable basics. So drawing, science, and music disappeared. Teaching became more mechanical as teachers found that drill and rote repetition produced the "best" results. Both teachers and administrators were tempted to falsify results, and many did. The plan was ultimately dropped, signaling the fate of every merit plan initiative ever since.

Today's merit pay proposals arrive in many forms and flavors and give teachers bonuses for such things as higher student test scores, obtaining satisfactory evaluations by principals or committees, acquiring additional duties, gaining new skills and knowledge, and serving in hard-to-staff schools. None of them, past and present, has ever had a successful track record. None has ever produced its intended results. Any gains have been minimal, short-lived, and expensive to achieve.

The idea of merit pay is simple in its appeal. It promises to raise student achievement by providing financial rewards for those teachers who meet specified goals and objectives. That is why, despite the fact that merit pay has never worked, politicians love it. Merit pay, while fiendishly " even impossibly " difficult to organize and manage, is easily explained to voters in simplistic sound bites. It gives the appearance of being "strong on education." A recent issue of Education Week observes that support for merit pay among politicians is always strongest around election time. At this moment, in about a dozen states around the country, there are proposals for merit pay schemes of one kind or another on the governor's desk.

It doesn't matter that these proposals are highly unlikely to be put into effect. It doesn't even matter that should a proposal be put into effect it will waste tens of millions of taxpayer dollars and will ultimately be scrapped. Because there will always be someone to blame for its failure: administrators who couldn't possibly manage such a bloated and complicated program; teachers who once they fully understand its negative impact on their work will balk and resist; and the unions, who never wanted merit pay in the first place.

Most merit pay plans are based on individual teacher performance, creating competition among teachers. Everyone wants the best students and the limited number of dollars available for bonuses. This undermines, even destroys, the staff teamwork needed in schools. But some schemes, like the one now being put forward by Nevada's Governor Kenney Guinn, would base merit increases on a school's overall performance or the achievement of students in a specific grade. "We need to use it in a team approach," Governor Guinn says. "If just certain people are successful, a school will still end up on the failure list."

We don't need a crystal ball to see where this is headed. Anyone who has ever participated in a group project consisting of people unaccustomed to working in teams knows how those turn out. A few hard workers carry the load while everyone enjoys the same reward, if any. A good way to foster resentment.

In addition to the desire to raise student performance, however, merit pay is often proposed to make the job of teaching more attractive to qualified candidates, and act as an incentive to prevent good teachers from leaving. Proponents of merit pay also argue that other professions " like law and medicine " provide financial rewards for increased professional achievement.

The truth is, we need both better teachers and better teaching practices in order to achieve better student performance. And that can only happen when teaching becomes a "true" profession. We need to institute a comprehensive system of education reforms that includes a career ladder in which teachers can, by acquiring the requisite skills, knowledge, responsibilities and certification, climb from one career level to the next -- for example, by advancing from associate teacher to teacher, then professional teacher, and finally chief instructor. And by further professionalizing the practice of teaching so that teachers teach in teams instead of in isolation, increasing collaboration and accountability. And by including ongoing professional development in the career path of all teachers, just like in other rigorous professional fields such as medicine and law. These steps must all be taken together in order for any of them to succeed, and for society to achieve the ultimate goal of improved student performance.

Until that happens, no education reform can possibly be effective over the long run. Not now, and not ever.

Katherine C. Boles is a lecturer at the Harvard Graduate School of Education. Vivian Troen is a lecturer at Brandeis University. They are the authors of "Who's Teaching Your Children?: Why the teacher crisis is worse than you think and what can be done about it," (Yale University Press).

Tying Teacher Compensation to Results: What Works and What Doesn't
Donald Gratz

Governor Romney recently proposed improving schools by paying bonuses to teachers based on student test scores, joining some 8-10 other states implementing or considering such plans. This sounds appealing " why not reward teachers who do the best job? Unfortunately, as Denver's complex four-year pay for performance pilot shows, basing teacher pay on student test scores is neither feasible nor desirable. What might work, the pilot suggests, is revising teacher compensation based on shared goals.

Educational performance pay plans stretch back hundreds of years. In the mid-1800s, for example, Britain's schools and teachers were paid based on tests of student performance. After 30+ years, however, the bureaucracy of testers had burgeoned, cheating and cramming flourished, and the practice was abandoned as a failure. Public reaction to testing had become so strong that the 1905 education handbook granted teachers "unlimited autonomy," each to "think for himself" how best to "assist" children to learn.

In 1918, 48% of U.S. public school districts described their pay systems as "merit based." But "merit" was subjective: high school teachers and men were paid more, minorities and women less, giving rise to today's uniform pay scale. By the 1950's, only 4% of the nation's districts were merit-based. Since then, performance pay was tried again in the early 1960's after Sputnik and during the Nixon and Reagan presidencies, but each attempt was abandoned after a few years. When Nixon launched a system of "performance contracting," for example, the experiment started with great claims of initial success, but ended in cheating scandals and failure.

In 1999, Denver's school board and teachers union jointly sponsored a pay for performance pilot based largely on student achievement. As head of the outside research team for the first half of the project, I can attest to the honesty and energy with which they approached the task. After four years and a substantial effort, however, the new pay plan is only 25% achievement-based. Denver's ground-breaking new compensation plan replaces the current steps and lanes approach with a four-part package including student academic growth, teacher knowledge and skill, professional evaluation, and teacher market value (hard to fill positions, hard to serve schools).

Denver abandoned test-based performance pay for the same reason it has been dropped historically " it doesn't work. But Denver's pilot, carefully implemented and studied, provides information on critical reasons for this failure, a few of which I note here. First, while there's no comprehensive definition of teacher performance, parents want more than test scores. They want their kids to come home happy and excited about learning, to grow socially and physically, to build complex thinking skills and to develop their special talents and interests. Achievement is primary, but teaching also includes hard-to-measure activities like fostering student confidence, involving parents, helping younger teachers, and creating a positive school culture.

Second, even when achievement is defined strictly as test scores (a dubious definition), valid measurement is extremely difficult. Do we calculate the number of children who meet a certain benchmark, as state tests do? That's a sharp disincentive to teach the neediest children " those with special needs or in urban areas. Do we measure student growth? That's better, but requires expensive data analysis and specific tests for every subject. Further, one year's change in test scores in a class of 20-25, with students coming and going, often does not reach statistical significance. Can districts pay teachers for results that may simply reflect good (or bad) luck?

Third, supposing we accept tests as a surrogate for good teaching and solve the statistical problem, major issues remain. In many districts, well over half the teachers teach gym, art, or music; they teach kindergarten or students with special needs; they are librarians, counselors, nurses, and other specialists. High school teachers offer many courses for which no tests exist. Should they all be excluded?

Finally, performance pay is built on the belief that higher pay leads to greater effort, which produces better results. Research does not support this belief. Even business, which uses incentives when outputs are clearly connected to effort, has largely dropped such pay schemes. Without clear and direct links between activity and result, research shows that they alienate workers, lower morale, and reduce output.

So how might we improve teacher compensation? One key finding in Denver mirrors the lessons of a century of research in industry: incentives can work when employer and employee agree both on their goals and the means of assessing whether those goals have been reached. Halfway through its pilot, having determined that the test-based system it was piloting wouldn't work, Denver formed a labor-management committee to study other alternatives. The plan they arrived at recognizes the complexity of teaching and need for professional growth with the four part plan mentioned above.

It is notable that Denver's school board, teachers, business leaders and voters have all endorsed this new plan, at an additional cost of some $25 million. Given the resistance in other locations, how could that happen? In the end, it's no mystery: when serious people work together to craft goals and measures they can all endorse, it's easier to win the support of their larger constituencies. The lesson from Denver, from industry, and from centuries of history, is that imposed performance pay doesn't work, but people working together around a common goal can accomplish great things. In Massachusetts, encouraging and funding such discussions might produce the best results.

Donald Gratz is Director of Graduate Education Programs at Curry College. He is also a Senior Associate at the Community Training and Assistance Center (CTAC). In addition, Don is a school committee member and a member of two unions.

Send us your comments:
To respond to any of the commentaries, please email us at We will post selected responses in the "E-forums" section of our website.

Recent Research from the Rennie Center New Book: Win-Win Labor-Management Collaboration in Education This new book highlights innovative "best practices" for improving labor-management relations in public education. Edited by the Rennie Center, written by Linda Kaboolian and Paul Sutherland, and published by Education Week Press, Win-Win Labor-Management Collaboration in Education is a handbook for union leaders, teachers and managers offering innovative best practices on how to reform the collective bargaining process for the benefit of students. Covering topics like "peer review," "pay for performance," and "school intervention processes," this book provides a unique national review of path-breaking collective-bargaining agreements and illustrates how districts and unions are putting their shared interests in students and learning at the forefront of their work together. Strides made by districts throughout the nation are highlighted, as well as best practices implemented in major urban regions. Read an exerpt at:

New Report: Scaling Up: Reform Lessons for Urban Comprehensive High Schools This report lays out an action agenda for large comprehensive high schools and clarifies what needs to happen at the school, district and state levels in order for sustainable change to take effect. While the movement toward small schools as a high school reform model is benefiting some students, the majority of urban high school students are still served in large comprehensive high schools. The majority of high schoolers have yet to benefit from their new high profile among politicians and reform funders. This report highlights the lessons that comprehensive high schools must heed in enacting improvement efforts. Scaling Up provides promising examples of urban high schools that are making it possible for all students to achieve at high levels. It focuses on three interrelated pieces of the reform puzzle, each of which is an essential component of whole school improvement: personalizing the learning environment, building teacher capacity, and setting and meeting high expectations for all students. See the complete report at:

About the Rennie Center The Rennie Center's mission is to develop a public agenda that informs and promotes significant improvement of public education in Massachusetts. Our work is motivated by a vision of an education system that creates the opportunity to educate every child to be successful in life, citizenship, employment and life-long learning. Applying nonpartisan, independent research, journalism and civic engagement, the Rennie Center is creating a civil space to foster thoughtful public discourse to inform and shape effective policy. For more information, please visit

Top of Page