Manuel S. González CanchéEdgerton, Adam Kirk2023-05-222021-06-102021-08-312020-01-012021-08-31https://repository.upenn.edu/handle/20.500.14332/31129Education reform rhetoric frequently pits the vested interests of teachers’ unions against those of students and families. To test whether union restrictions are related to student learning, I analyze a unique database of contractual items for the 2016-2017 school year across all 499 Pennsylvania school districts in order to examine a) variation, b) partisan political predictors, and c) relationships to student achievement and graduation rates. I also examine changes in 105 contracts that occurred during the 2015-2016 school year. I depict variation among items using GIS mapping. I use OLS regression, probit regression, and spatial autoregression to examine relationships between contract features and student proficiency and graduation rates. I also use propensity score weighting with generalized boosted models (GBM). After controlling for spatial dependence and district demographics, I find a significant negative relationship between the percentage of registered Republicans in a district and bonuses for teacher graduate credentials. I find a significant and positive relationship between Republican registered voters and math and science proficiency. This relationship diminishes in magnitude for ELA proficiency. I also find a significant positive relationship between average years of teaching experience and ELA proficiency in grades 3-8. Using GBM, I find significant positive estimates (+2%) of teacher qualification indicators on students’ math achievement in grades 3-8, and a significant positive estimate (+2%) between harsh consequences for ELA teachers and student proficiency. I also find a significant positive estimate between higher teacher pay and biology proficiency (+4% for historically disadvantaged students), as well as a significant negative estimate of graduate credential bonuses on graduation rates (-6%). These correlational results suggest that subject-area and grade-level differentiation in contracts – such as higher wages for STEM teachers – might be beneficial. The most effective STEM teachers might be seeking out positions in the best-paying districts with the strongest contracts.179 p.application/pdfAdam Kirk Edgertonaccountabilitycollective bargainingstudent achievementteacher evaluationteachers unionstestingEducation PolicyDissertation/Thesis