Notes: Coefficients reported with robust standard errors (clustered at the school level). Danielson FFT ratings for individual teachers are averaged across all components (within an observation) and then across all observations. Research evidence supports that individual component ratings should receive equal weight within an observation (Kane et al. 2011, Garrett and Steinberg 2015). The sample of schools decreases from the full sample of 44 treatment schools to 39 because teachers are missing FFT ratings either in 2008–2009 or 2009–10 at four schools, and an additional school closed in the 2009–10 school year. Change in School Leadership is an indicator that equals 1 if the school had a new principal in the 2009–10 school year, and 0 otherwise. Teacher characteristics include years of experience in CPS (experience squared), race/ethnicity, gender, teacher’s prior evaluation rating, certification status (alternative), educational attainment (bachelors, masters, or doctorate), and undergraduate major. We report the coefficient on National Board certification. To account for teachers missing prior evaluation, we include a separate indicator variable for teachers missing a prior evaluation rating. Models with school fixed effects allow for a comparison of teachers who left/switched schools at the end of the 2008–2009 school year to teachers who entered the same school for the first time in the 2009–10 school year. Coefficients statistically significant at the

↵* 10 percent,

↵** 5 percent, and

↵*** 1 percent levels.