Partisan Predictors for Collective Bargaining Agreement Items

Abstract

The United States is rare among nations in its highly decentralized process for negotiating collective bargaining agreements with local teachers’ unions. To determine whether partisanship can predict these highly localized decisions, I construct an original database of Pennsylvania collective bargaining agreements (CBAs) merged with publicly available voter registration records to predict the presence of high-profile contract items. Using spatial autoregression and probit regression, I reveal that the partisanship of a school district is a significant predictor for fewer seniority protections but not for lower salaries. These partisan relationships can guide both district administrators and union leaders in future negotiations.