Chapter 1 begins with the motivation of my study in teachers' labor market. I employ a monopolistic screening model to show that there exist multiple equilibria in the educational system; a pooling equilibrium and a separating equilibrium. The model predicts that the pooling equilibrium is optimal only when the average quality of teacher applicants is high. Using data from the OECD, I examine the relation between teachers' earnings and teacher quality of the U.S. and Korea. Chapter 2 focuses on teachers and their career dynamics, and the data is at teacher level. Using the Current Population Survey for 2001-2010, I show that public school teachers are paid less compared to other comparable college graduates in non-teaching sectors. By studying the change in earnings after career changes, I find the evidence of positive selection when teachers move into the non-teaching sectors and of negative selection when non-teachers move into the teaching sector, which results in the decrease in the average teacher quality. Chapter 3 looks at both teachers and school districts, and I use district-teacher matched dataset, based on the School and Staffing Survey (SASS) for 2007-2008. I employ a multilevel model and a propensity score matching to identify union effects in states with different legal environments for collective bargaining of teachers. I find that collective bargaining is neither necessary nor sufficient for unions to affect teachers' well-being. I show that meet-and-confer is a popular alternative to collective bargaining and that it is an important mechanism for unions to influence teachers' non-wage benefits. Chapter 4 concerns school districts, and I use SASS district level data. I reevaluate the role of teachers unions on pay structure and districts' financial status. In contrasts to previous findings, I find that the variance of teachers' earnings is higher in more unionized settings. Moreover, I show that the financial status of districts with teachers unions is stronger than that of districts without the unions. I confirm that unionism is associated with less usage of performance pay system. / Economics