We show that intergenerational mobility changed rapidly by race and class in recent decades in the United States and study the causal mechanisms underlying those changes. Between the 1978 and 1992 birth cohorts, earnings increased for white children from high-income families relative to white children from low-income families, increasing earnings gaps by parental income (class) by 30%. Earnings increased for Black children at all parental income levels, reducing white-Black earnings gaps for children from low-income families by 30%. Class gaps grew and race gaps shrank similarly for nonmonetary outcomes such as educational attainment, standardized test scores, and mortality rates. Using a quasi-experimental design, we show that the divergent trends in economic mobility were caused by differential changes in childhood environments, as proxied by parental employment rates, in local communities defined by race, class, and childhood county. Outcomes improve across birth cohorts for children who grow up in communities with increasing parental employment rates, with larger effects for children who move to such communities at younger ages. Children’s outcomes are most strongly related to the parental employment rates of peers they are more likely to interact with, such as those in their own birth cohort, suggesting that the relationship between children’s outcomes and parental employment rates is mediated by social interaction.

Americans rarely marry outside their race or class group, a pattern with well-documented implications for inequality and intergenerational mobility. Limited exposure may partly explain these low intergroup marriage rates. We instrument for exposure using variation in childhood neighborhoods based on whether other race and class groups had more opposite-sex children of similar age. Exposure increases interclass (high- and low-parent income) marriage but has no detectable effect on interracial (White and Black) marriage. A spatial marriage model predicts that residential segregation—one of many forms of exposure—accounts for more than one-third of marital sorting by class but less than 5 percent by race.