This dissertation addresses three broad issues within the fields of labor economics and health economics. Chapter 1 studies how the skill content of college coursework contributes to early-career earnings and inequality. The continuing shift of the U.S. economy toward a high-skill base has increased demand for college-educated workers. To understand how higher education prepares students for this evolving economy, a large body of research in labor economics has focused on the causes and consequences of college enrollment, institutional selectivity, and major choice. Much less attention has been paid to a key dimension that shapes the skills students acquire in college—coursework. In this chapter, I scrape and compile a new dataset of detailed course descriptions from Texas public universities. Using a large language model (GPT-4), I extract the skills students are likely to acquire from each course, focusing on two widely taught and consistently identifiable domains: quantitative and writing skills. I then link these course-level skill measures to Texas administrative records that track students’ educational histories and quarterly earnings. To estimate the returns to coursework-based skills, I implement an instrumental-variable strategy that exploits variation in course offerings across cohorts within the same major. I find substantial early-career earnings returns to coursework-based quantitative skills, but no detectable returns to writing skills. These returns are especially large for underrepresented minority (URM) students and for students in less quantitatively intensive majors, suggesting that expanding access to quantitative coursework within majors may serve as a new lever for narrowing racial earnings gaps. Chapter 2 (with Lipeng Chen) studies whether the expansion of mobile internet contributed to the recent deterioration in mental health in the United States. The past decade witnessed a significant worsening of mental health in the United States, a trend that coincided with the expansion of mobile internet and the resulting increase in screen time. In this project, we estimate the impact of mobile internet on mental health using a two-way fixed-effects model, leveraging temporal and spatial variation in 3G internet coverage. We find that people’s mental health worsened after the arrival of 3G internet. This effect appears to be driven primarily by increased social media use. We also find that younger individuals and women were more affected by mobile internet during this mental health crisis. Chapter 3 (with Ian Lundberg) studies how researchers should define causal questions—and choose comparison groups—in staggered-adoption panel settings. A powerful data structure for causal inference is staggered adoption: many units are observed over many time periods, during which some units adopt an irreversible treatment at different times while others remain untreated. Popular methods that apply in this setting include difference-in-differences, fixed effects, matching, and synthetic control. All of these methods compare the future outcomes of treatment adopters and non-adopters to answer the question: for units that adopt treatment, what outcomes would have been realized if they had remained untreated? We show that this question may hide two distinct causal questions of interest. The first question is what would have happened if a treated unit had remained untreated throughout all periods until outcome measurement. The second question is what would have happened if a treated unit had remained untreated in the particular period when it in fact became treated. Popular methods such as synthetic control are often interpreted with respect to the former question (about a longitudinal treatment). We show that they actually answer the latter question (about a point-in-time treatment). The distinction is especially relevant when many units become treated at many time points, a common setting in applied economics and demography.