Zuzek, S. (2017). Peer effect estimation through a covariate-adjusted regression discontinuity design [Master Thesis, Technische Universität Wien]. reposiTUm. https://doi.org/10.34726/hss.2017.46247
This thesis uses data on the Austrian labor market to obtain an estimate of a peer effect in parental leave decisions among young fathers. Identification arises through a reform of parental leave, when higher benefits led to an increase in the participation rate of 4.7 - 6.6 percentage points. I use the resulting discontinuity around the implementation date of the reform to implement a fuzzy Regression Discontinuity Design (RDD), which allows to obtain estimates of a peer effect of reform-window fathers on their coworkers. I include additional covariates on the observations in order to increase the precision of the estimator and discuss relevant theoretical results on asymptotic inference. The results indicate a quantitatively large increase of the coworker participation rate of 13 - 28 percentage points, however statistical significance of the peer effect is weak.