This paper uses an event study model to find the effect of a class size reduction mandate on student achievement in Florida school districts. The results show that treated districts that had to reduce their average class size from 2004 to 2006 outperformed control districts in reading and math scores for 4th-8th graders. The effects are robust to controls for funding meaning that money spent on class size reduction (CSR) had a greater benefit for treatment districts than the same money spent on other educational inputs in control districts. CSR had lasting benefits with reading and math scores around 0.1 and 0.2 standard deviations higher in treatment districts from 2004 to 2019. The benefit is largest for 4th and 5th graders in reading and 8th graders in math. The effect for schools that had to reduce their class size average from 2007 to 2010 is more difficult to estimate due to selection issues, but preliminary results show possible long term effects of CSR in PK-3rd grade on 4th-8th grade scores. Using prior estimates of the return of higher test scores to lifetime earnings, the results can help calculate the benefit to students from the CSR mandate.

Using variation in whether districts were above or below cap at the start of the policy, this paper presents event studies that show schools in districts treated with mandated CSR in 4th-8th grades hired enough teachers to close the mandate gap in the number of teachers per student between treatment and control districts. Using district-wide variables for teacher salary, level of education, and average years of experience, I estimate changes to the composition of teachers. Treatment districts hired more teachers with only a bachelor's degree. Otherwise, treatment districts had no significant difference in composition changes from control districts. Control districts also hired teachers with a bachelor's degree which coincided with a state-wide rise in minimum salaries and a decrease in average experience by 1 and then 3 years. Overall, districts (including control districts) were able to reduce class sizes by offering higher salaries to inexperienced teachers with less education. There is no evidence that control districts raised salaries with the funds they received (and didn't need to use) for CSR. While teacher quality across the state probably declined over this period, there is not enough evidence from the district-wide teacher variables to say that CSR mandates caused it since other events (i.e. the Great Recession) could have had the same state-wide effects.