In this dissertation, I study the interplay of familial and geographic factors in influencing human capital development and economic mobility in the United States. The first chapter extends a canonical model of intergenerational human capital investment to a geographic context in order to study the role of migration in determining optimal human capital accumulation and income mobility in the United States. The main result is that migration is considerably influential in shaping the high rates of economic mobility observed among children from low-wage areas, with human capital investment behavioral responses being important to consider. Equalizing school quality across locations does more to reduce interstate inequality in income mobility than equalizing skill prices, and policies that attempt to decrease human capital flight from low-wage areas via cash transfers are unlikely to be cost-effective. The second chapter, joint with Joanna Venator, studies how childcare costs, the location of extended family, and fertility events influence both the labor force attachment and labor mobility of women in the United States. We begin by empirically documenting strong patterns of women returning to their home locations in anticipation of fertility events, indicating that the desire for intergenerational time transfers is an important motivator of home migration. Moreover, women who reside in their parent's location experience a substantial long-run reduction in their child earnings penalty. Next, we build a dynamic model of labor force participation and migration to assess the incidence of counterfactual scenarios and childcare policies. We find that childcare subsidies increase lifetime earnings and labor mobility for women, with particularly strong effects for women who are ever single mothers and Blacks. Ignoring migration understates these benefits by a meaningful extent. The third chapter, joint with Owen Thompson and Jason Fletcher, studies the long-run impacts of court-ordered desegregation. Court ordered desegregation plans were implemented in hundreds of US school districts nationwide from the 1960s through the 1980s, and were arguably the most substantive national attempt to improve educational access for African American children in modern American history. Using large Census samples that are linked to Social Security records containing county of birth, we implement event studies that estimate the long run effects of exposure to desegregation orders on human capital and labor market outcomes. We find that African Americans who were relatively young when a desegregation order was implemented in their county of birth, and therefore had more exposure to integrated schools, experienced large improvements in adult human capital and labor market outcomes relative to Blacks who were older when a court order was locally implemented. There are no comparable changes in outcomes among whites in counties undergoing an order, or among Blacks who were beyond school ages when a local order was implemented. These effects are strongly concentrated in the South, with largely null findings in other regions. Our data and methodology provide the most comprehensive national assessment to date on the impacts of court ordered desegregation, and strongly indicate that these policies were in fact highly effective at improving the long run socioeconomic outcomes of many Black students.