This dissertation studies two classic questions in labor and education economics. Specifically, it estimates the labor market impacts of immigrants on natives in the United States and the effect of school scheduling on the students' academic performance. It is divided into four chapters, the first two dedicated to immigration and the last two to school start time. A common theme throughout is the use of cutting--edge econometric techniques with a strong focus on getting causal estimates in an attempt to improve on previous studies. In Chapter 1 (revise and resubmit at Journal of Human Resources), joint with Giovanni Peri, we apply the Synthetic Control Method to re-examine the labor market effects of the Mariel Boatlift, a large inflow of Cubans into Miami in 1980, first studied by David Card (1990). This method improves on previous studies by choosing a control group of cities that best match Miami's labor market trends between 1972 and the Boatlift. We also provide more reliable standard errors for the inference and we analyze a set of outcomes for low-skilled workers, ranging from wages to unemployment. Using the sample of non-Cuban high school dropouts in the age range 19-65 we find no significant difference in the wages and employment of workers in Miami relative to its control after 1979. The result is robust to several checks and is valid in most sub-samples. In performing a systematic comparison with Borjas (2017) we show that, especially when using the March-CPS data, focusing on specific sub-samples and matching the control group on short pre-1979 data series can produce large differences between Miami wages and control. However, when studied more systematically, those wage differences across subgroups and over time appear to be the result of classical measurement error in small samples rather than the labor market effect of the Boatlift. Chapter 2 provides a much--needed fresh look at measuring the labor market impacts of immigrants on natives under very weak assumptions. To identify the labor market impacts of immigrants on natives, prior empirical studies have relied on strict econometric assumptions. Additionally, the literature lacks a consensus on the sign and magnitude of this effects. I take an entirely different approach by weakening routinely made assumptions and provide the first non–parametric estimates. In particular, I apply conservative empirical and theoretical bounding strategies. The estimated bounds on the effects on natives’ wages under minimal assumptions rule out elasticities smaller than -0.37 or larger than 0.38. To tighten this interval, I explore mild instrumental variable and monotonicity assumptions motivated by economic theory which narrow the lower bound to -0.11. After summarizing the estimates from 63 papers on the topic, I show this effect is much smaller in magnitude than some previous studies claim it to be. Moreover, I show that the data reject the lower bound prediction of an extreme version of the canonical labor demand and supply model. Chapter 3 (published in Economics Letters, Vol.139, pp 36-39 (2016)), joint with Lester Lusher, examines the impact of a double--shift schooling system on students' performance. School scheduling systems are frequently at the forefront of policy discussions around the world. This paper provides the first causal evidence of student performance during double-shift schooling systems. We exploit a six-year quasi-experiment from a country in Eastern Europe where students alternated between morning and afternoon school blocks every month. We estimate models with student--class and month fixed effects using data on over 260,000 assignment-level grades. We find a small, precisely estimated drop in student performance during afternoon blocks. In Chapter 4 (revise and resubmit in Economic Inquiry), joint with Lester Lusher, we study to what extend differences in sleep cycles between boys and girls can explain the observed gender performance gap in middle and high schools. Sleep studies suggest that girls go to sleep earlier, are more active in the morning, and cope with sleep deprivation better than boys. We provide the first causal evidence on how gender differences in sleep cycles can help explain the gender performance gap. We exploit over 240,000 assignment-level grades from a quasi-experiment where students’ schedules alternated between morning and afternoon start times each month. Relative to girls, we find that boys achievement benefits from a later start time. For classes taught at the beginning of the school day, our estimates explain up to 16% of the gender performance gap.