Racial disparities in the United States are well-documented, but social scientists know little about how other aspects of race, like skin color, affect people's experiences within different institutions. A growing body of research demonstrates that racial minorities with darker skin have worse socioeconomic outcomes than their lighter-skinned counterparts. However, because this work commonly analyzes individuals from different families, scholars have difficulty extricating whether variations in the outcome of interest are due to skin color bias or to differences in unobserved family characteristics. I improve upon prior studies by using an innovative within-family approach with sibling data from the National Longitudinal Study of Adolescent to Adult Health, which I argue provides analytical leverage. I first examine the relationship between skin color and different outcomes among a nationally-representative sample. Then, to account for mutual unobserved and observed family characteristics, I use sibling fixed-effects models to consider whether outcomes vary by skin color among members of the same family. I find that skin color is a significant predictor of the life chances of siblings in arrest outcomes, educational attainment, and income. This dissertation contributes to the existing literature by providing a first step towards understanding the implications of skin color for enduring racial inequality. I improve upon existing studies of colorism by analyzing siblings who have a shared family background. In doing so, my study begins to illuminate how skin color continues to be a source of stratification in the United States both across and within families for racial and ethnic minorities.

Academic course failure is perhaps surprisingly common in the US education system. Studies in educational development suggest that course failure has universally negative effects on students’ educational outcomes. However, it is unclear the degree to which these negative results may be driven by differences between students who do and do not fail, and whether all students are equally impacted. My dissertation examines how course failure and academic labels impact young adults’ academic and labor market outcomes, focusing on differences across student characteristics, and using quasi-experimental research designs to account for differences in prior academic or demographic characteristics between students who do and do not fail. Drawing on research in life course theory, I argue that failure more broadly should not be understood as aberrant but as an important part of the developmental process. Using three large, longitudinal data sources with academic transcript information, and employing quasi-experimental methods where possible, I demonstrate that course failure can compound advantages or disadvantages to have substantial impacts on future outcomes that vary by gender and race. Chapter 2 uses students’ college transcripts from the National Education Longitudinal Study (NELS 88) to show that failing a so-called “weed out” course discourages women from pursuing STEM majors, but that men who fail major in STEM fields at similar rates to those who pass. In Chapter 3, I use transcript data from the National Longitudinal Survey of Youth (NLSY97) to show that while college remedial coursework benefits some students, the substantial number of students who fail remediation are considerably worse off (e.g., they are less likely to graduate, take longer to graduate, and earn less) than peers who were not placed in remediation. Finally, Chapter 4 uses statewide administrative data to show that students with a more negative performance label have lower test scores and worse behavioral outcomes over time. My dissertation thus underscores how students who failed in school will likely access vastly different opportunities in education, the labor market, and other institutions.

Schools often adopt restorative justice programs in efforts to mitigate racial inequality in school discipline and to address growing concerns around the school-to-prison pipeline. While both the school-to-prison pipeline and restorative justice (RJ) programs have garnered attention, little is known about both the long-term impacts of experiencing school discipline and the extent to which interventions like RJ can mitigate these processes. Thus, this project draws on theoretical understandings of racial stratification and schools as organizations to investigate the long-term impacts of school discipline and examine restorative justice implementation in two school district contexts in the U.S. Drawing on unique data sources and utilizing inference-based statistical approaches, I investigate 1) how experiencing school discipline is associated with transition to adulthood outcomes 2) the extent to which RJ practices affect school discipline trends over time, and 3) how RJ effects may vary on the logic of adoption and implementation strategies. Analyses from my first empirical chapter link school discipline with young adult outcomes, finding that this linkage tends to be stronger for Black students. Subsequent chapters rely on inference-based statistical approaches to investigate how restorative interventions impacted school disciplinary disparities in two school districts and analyzes how the logic of both interventions led to a widening of racial disproportionality in one district and a narrowing disproportionality in the other. Taken together, this work provides a nuanced view of the structural conditions that surround school disciplinary disparities and the interventions that are often adopted to ameliorate them.