Essays in Employee Evaluation and Compensation: Evidence from Educators

This dissertation investigates three topics in employee evaluation and compensation, with a focus on educators. The first chapter investigates the use of high-stakes subjective employee evaluations in teaching and the determinants behind the way in which supervisors assign these ratings. I find that classroom observations of teaching remain relatively predictive of teacher effectiveness when the stakes are high, but supervisor assignment behavior is difficult to influence. Principals also appear to take into account the financial well-being of their teachers when assigning these ratings, and I find evidence that repeat observations by the same supervisor tends to increase subjective ratings. The second chapter investigates the efficacy of a program in which teachers were paid substantially more to teach in a set of chronically lower-performing schools in a large, urban school district in the United States. We find that more effective educators moved into these schools, and teacher effectiveness remained high for the duration of the program. Student achievement subsequently increased as well, but there is some evidence that achievement fell after the program ended. The third chapter investigates the extent to which principals influence their students' later life outcomes. We find that elementary school principals in the Chicago Public Schools and middle school principals in Texas substantially vary in their value-added towards future cognitive and non-cognitive outcomes in high school, college attendance and persistence, and employment. Principals that are more effective on concurrent and future cognitive and non-cognitive outcomes are also generally more effective in their value-added to college attendance and persistence and post-high school employment.