This dissertation is comprised of three essays, each examining a different topic in labor economics. In Chapter 1, I examine the productivity and selection effects of K-12 teacher tenure by leveraging variation from New Jersey’s TEACHNJ Act. This law extended the pretenure period from three to four years and allowed districts to dismiss consistently low-performing teachers. I use multiple identification strategies to estimate the productivity effects of tenure across a teacher’s career. I evaluate the productivity effects at tenure receipt using a difference-in-differences design, which compares fourth-year tenured and pretenured teachers. At tenure receipt, math value-added declines but English language arts value-added and summative ratings remain unchanged. To estimate the productivity effects later in the career, I use a regression discontinuity design relying on discontinuities in job security around summative rating thresholds. Later in the career, tenure has no impact on productivity. Thus, tenure induces a transitory decline in math value-added without impacting other dimensions of teacher performance. Focusing on the labor market effects, I compare teachers hired before and after TEACHNJ within the same district and experience level. The TEACHNJ Act disproportionately increased male and Black teacher turnover rates. TEACHNJ did not impact the quality of the teacher labor market as measured by value-added, though higher rated teachers often filled new vacancies. Since the TEACHNJ Act only relies on summative ratings to make personnel decisions, this result aligns with a multitask principal-agent model where only one of several measures of performance is used to evaluate the employee. In Chapter 2, I evaluate techniques to identify high-quality teachers. Since tenure restricts dismissals of experienced teachers, schools must predict productivity and dismiss those expected to perform ineffectively prior to tenure receipt. Many states solely rely on evaluation scores to guide these personnel decisions without considering other dimensions of teacher performance. For example, New Jersey uses summative ratings, which primarily rely on supervisor evaluations. I use various predictive models to rank teachers based on predicted value-added and summative ratings. I then simulate revised personnel decisions and test for changes in average retained teacher performance. In this exercise, I adjust two factors that impact the quality of the predictions: the number of predictors and the length of the pretenure period. Both factors impact the precision of the predictions, though extended pretenure periods also may negatively impact selection into teaching. I find that revised algorithms using both value-added and summative ratings increase the average value-added of retained teachers by 0.01 standard deviations without decreasing summative ratings or diversity. This Pareto improvement equates to a present value gain of $2,240 per student. These returns are a product of using additional information rather than advanced algorithms, as I generate similar gains when using simple ordinary least squares regressions or advanced machine learning techniques. In comparison, algorithms that extend the pretenure period beyond one year do not provide enough additional information to significantly improve average retained teacher performance unless dismissal rates increase dramatically. In Chapter 3, my coauthor and I examine how returns to enrolling in science, technology, engineering, and mathematics (STEM) programs vary with students' academic preparation. We match data on STEM admissions at a Colombian flagship university to nationwide college and earnings records. Our identification strategy combines a regression discontinuity design with variation in admission quotas. We find that less-prepared students were less likely to complete a STEM degree than their more able peers, but they had larger earnings returns to enrolling. Our results suggest that policies that encourage less-prepared students to enroll in STEM programs can yield large but unevenly distributed earnings gains.