History of Merit Pay

History and Issues Surrounding the Single-Salary Pay Schedule

In 1921, Denver and Des Moines became the first cities in the United States to introduce the single-salary pay schedule to teachers, named so because all classroom teachers were paid on the same scale regardless of race, gender, family status, or grade level taught.  The single-salary pay schedule was constructed from a matrix of experience in years and degree attainment.  Appropriate pay for an individual teacher was determined by locating which cell the teacher corresponded to in terms of years of experience and degree credits he or she had earned (Protsik, 1995).

For nearly sixty years, the single-salary pay schedule was seen as the best way to provide both equity and objectivity to teacher pay.  It ensured teachers would be paid for teaching in its own right and not according to discriminatory measures or grade level taught.  Furthermore, it encouraged teachers to stay in the teaching profession and rewarded them for greater levels of educational attainment.  Finally, it allowed unions to represent all members in collective bargaining agreements fairly and equally.  In 1944, the National Education Association stated that any measures of teacher merit used to determine teacher salary under the single-salary schedule were unreliable.  As a result, many cities followed the lead of Denver and Des Moines, and by 1950, 97% of cities had adopted the single-salary pay schedule (Protsik, 1995).  

There are also many problems associated with the single-salary pay schedule. According to Richard Rothstein, one of the biggest concerns of the single-salary pay schedule is that it can reward teachers with course accumulation not explicitly relevant to teacher performance (personal communication, 2009, October 28).  Furthermore, it prevents districts from using higher pay to attract teachers in shortage areas such as special education, math, and science, or lowering pay to reduce surplus in areas like general elementary education.  In other words, Rothstein argues, districts are unable to use market-based forces of salary to control teacher supply (personal communication, 2009, October 28).  Additionally, according to Professor Richard Murnane, research has shown there are no improvements in teaching quality or in student outcomes if a teacher has or has not obtained a Master’s degree, so we cannot predict which teachers are better based on this system alone (personal communication, 2009, November 2).  Finally, the single-salary pay schedule fails to provide teachers with incentives to improve their classroom skills because it treats all teachers with the same education level and experience as equals, despite potentially unequal performance and skills (Protsik, 1995). 

 Rise of Merit Pay 

There were always those few who felt exceptional teachers should be rewarded accordingly.  This opinion, coupled with the many concerns associated with the single-salary pay schedule, led to the popularity of merit pay programs starting in the 1980s. The first merit pay plan was attempted in 1908 in Newton, Massachusetts, but the 1980s represented the first time the nation as a whole called for teacher improvement through monetary incentives ((Protsik, 1995). In 1983, A Nation at Risk recommended teachers’ salaries be “professionally competitive, market-sensitive, and performance-based” in order to tie compensation more directly to classroom skill (A Nation At Risk: Recommendations, 1983).  

A wide variety of merit pay plans were tried all across the country through the 1980s and 1990s with varying degrees of success. Many awarded bonuses for excellent classroom performance, usually determined by supervisor and peer review, but some teachers felt that the awards were based on biased evaluations instead of objective assessments of performance.  Murnane and Cohen refer to such plans in their 1985 article, “Merit Pay and the Evaluation Problem: Understanding Why Most Merit Pay Plans Fail and A Few Survive,” as “old-style merit pay.” Other plans were considered “career-ladder” programs in that they created career steps for teachers based on evaluations of teaching, years of experience, and usually individual student achievement (Cornett & Gaines, 2002).  Murnane and Cohen (1985) refer to this as “new-style merit pay.”  A 1979 Educational Research Study found that most merit pay plans that were enacted during this time were discontinued within six years due to problems of administration and personnel, collective bargaining disputes, and budgetary shortfalls (as cited in Protsik, 1995).

Problems with Earlier Merit Pay Plans 

The merit pay plans implemented and experimented with in the 1980s and 1990s tried to make teachers more accountable for individual student performance.  Teachers were not genuinely convinced however, that these plans were better or more equitable than the single-salary pay schedule.  Moreover, through numerous surveys and interviews, research documented that teachers in large numbers preferred differentiated pay that provided “extra pay for extra work” rather than being graded or scored in a performance assessment.  Furthermore, an evaluation of older merit pay programs found that trying to incentivize teaching through money did little to improve student or teacher performance.  It was instead suggested that compensation plans need teachers to take on differentiated roles and pay them on that basis (Cornett & Gaines, 2002).

Merit Pay Plans Today

In today’s merit pay plans there is a shift away from individual teacher performance and student achievement to a system of school-by-school results and gains. As of 2005, twenty states across the United States have a school-based performance system in place or are in the process of implementing such a system (Johnson, 2006). According to Richard Rothstein however, there is no research-based evidence that merit pay will be successful in raising student achievement, especially if pay is tied directly to test scores.  The techniques for identifying superior teachers based on test scores still have many problems, so attempts will potentially misidentify “better” and “worse” teachers. Additionally, because test scores are only available in reading and math, creating high-stakes for teachers may increase incentives to narrow curriculum both within and across subject areas (R. Rothstein, personal communication, 2009, October 28). Similarly, Richard Murnane also expressed reservation to basing compensation on school-wide performance due to the natural variation between and within classrooms.  Since superintendents are under constant pressure from two opposing constituencies (the internal one of teachers and unions who usually oppose merit pay and the external one of tax payers who do not like to reward lousy teachers) if they insist on using performance-based compensation, Murnane argues that they must use at least three years of data for more accuracy (personal communication, 2009, November 2).