FAMILY AND INTERGENERATIONAL MOBILITY

Abstract

This dissertation consists of three chapters focused on family and intergenerational mobility. Chapter 1 examines an under-explored determinant of intergenerational mobility: racial assortative mating. Our unique identification strategy leverages two historical events - changes in the marriageable Black population during the Second Great Migration and the landmark 1967 Supreme Court ruling in Loving v. Virginia. The interplay of these events presents exogenous shocks to marriage market conditions that are conducive to interracial marriages, thereby reducing racial assortative mating. We find a pronounced, negative relationship between racial assortative mating and mobility. We further show that the negative mobility impacts associated with assortative mating might be channeled through factors such as social network formations, neighborhood characteristics, and family structure. Finally, we also uncover a significant heterogeneity in the effects across the income distribution. The negative effects are more markedly pronounced for children from low-income families than for children from more affluent families. Such a pattern is consistent with the differential influence of alternative mechanisms across the income distribution. Chapter 2 explores the important role of environmental factors, specifically sunlight exposure, in shaping intergenerational educational mobility. Our findings reveal a significant and positive relationship between sunlight exposure and educational mobility, but this effect is predominantly driven by its influence on the educational mobility of sons. In contrast, sunlight exposure appears to have a negligible or even negative effect on daughters’ educational mobility. We suggest that these heterogeneous effects may be mediated by factors such as fertility rates, lifetime maternal mortality risk, mental health, and time preference. Specifically, individuals in regions with more sufficient sunlight exposure tend to experience higher fertility rates, a lower risk of maternal mortality, better mental health, and lower levels of patience. These factors may, in turn, influence parental investment in their children’s education, contributing to the observed gender-based heterogeneity in the impact of sunlight exposure on educational mobility. Child gender could affect not only intergenerational mobility but also family stability. In Chapter 3, we explore the heterogeneous effects of having a firstborn daughter on the likelihood of divorce using the causal forest method. Our causal forest model reveals substantial heterogeneity in the effect of daughters on divorce, with the mother’s age at first marriage being the most important factor driving this heterogeneity. The daughter effect varies significantly both across different maternal characteristics and over time. For mothers who were older than 30 at their first marriage, having a firstborn daughter significantly increased divorce risk, but this effect has diminished. Conversely, mothers who married at 25 or younger are increasingly likely to divorce after having a firstborn daughter. As for middle-married mothers, child gender has no significant impact on their likelihood of divorce. These patterns demonstrate the interplay of son preference and income constraints in shaping the daughter-divorce relationship. Compared to early-married mothers, late-married mothers are more susceptible to social pressure for sons but less impacted by child-rearing costs. Over time, as son preference has declined and the cost of raising daughters has increased, the positive daughter effect on divorce has shifted from being primarily driven by late-married mothers in the 1960-1969 childbirth cohorts to early-married mothers in the post-1980 childbirth cohorts.