Using a Difference-in-Differences Approach to Estimate the Effects of Teacher Merit Pay on Student Performance Mark Gius Quinnipiac University, Hamden, CT 06518, USA. The present study uses a difference-in-differences approach to estimate the effects of teacher merit pay on student graduation rates and dropout rates. Using district-level data from the Schools and Staffing Survey for the years 1999–2000, 2004–2005, and 2007–2008, results of the present study suggest that merit pay for excellence in teaching has statistically significant and positive effects on the district-level graduation rate, but no significant effect on dropout rates. Merit pay for national teacher certification, however, had no effects on either dropout rates or graduation rates. Eastern Economic Journal (2013) 39, 111–120. doi:10.1057/eej.2012.22; published online 6 August 2012 Keywords: teachers; merit pay; student performance; difference-in-differences JEL: I20; I22; I29; J33 INTRODUCTION The vast majority of public school teachers in the United States are compensated according to a single salary schedule [Podgursky 2008]. With a single salary schedule, teacher pay is based solely upon two factors: years of experience and level of educa- tion. Under a single salary schedule, the quality of a teacher’s instruction has no bearing on their pay. Instead, the more seniority and education a teacher has, the greater is their salary. It is commonly argued that such a system does not promote teacher development or student performance. It is believed that a single salary schedule protects ineffective, older teachers who, by virtue of their seniority, are among the highest paid teachers. Furthermore, under a single salary schedule, many effective and dynamic teachers are paid less than poorly performing teachers, thus demoralizing the better teachers and possibly driving them from the profession. Hence, it is believed that a merit pay system for teachers that rewards individual performance should increase student performance and do much to improve public education in the US. Many teachers, however, oppose merit pay primarily because they believe that it would undermine one of the most important aspects of teaching: collaboration. Teachers would not be willing to collaborate under such a system because it would detract from their individual performance and potentially reduce their compensa- tion. If there is a fixed pool of funds for merit pay, then helping a fellow teacher improve their teaching may result in less merit pay for everybody else; hence, collaboration would suffer if merit pay were implemented. Murname and Cohen [1986] provide supporting evidence for the teachers’ argument against merit pay and suggest that this type of compensation system is not Eastern Economic Journal, 2013, 39, (111–120) r 2013 EEA 0094-5056/13 www.palgrave-journals.com/eej/