Based On: Melissa S. Kearney, Phillip Levine, and Alden Q. Barson, “The Evolving Marriage Gap in Children’s Outcomes: Family Structure and Young Adult Outcomes Revisited,” National Bureau of Economic Research Working Paper No. 35765, September 2026
Key Takeaway
Changing family circumstances in recent decades have been associated with an evolving relationship between family structure and children’s young adult outcomes. Tracking the childhood-to-adulthood trajectories of children born in the 1970s, 1980s, and 1990s, this new research shows that as the economic circumstances of unmarried-mother families have improved, the poverty disadvantage associated with being born to an unmarried mother has narrowed considerably. But this convergence in economic well-being has not extended to higher educational attainment. A large family-structure gap in four-year college completion persists, with children born to married parents substantially more likely to earn a college degree. These findings suggest that improvements in material well-being among single-mother families have narrowed family-structure gaps in poverty, but important disparities in children’s outcomes remain.
Existing evidence has firmly established that children raised by married parents have, on average, better measured outcomes than children raised by unmarried parents. In a 2017 paper, Kearney and Levine theorized and documented how this average obscures differences across contexts. They showed that the gap in outcomes for children born to married and unmarried mothers depends on a mother’s own resources, the resources her partner would add, and how much those resources matter for a given outcome. As parental resources and economic circumstances have changed over decades, this raises questions about how the gaps in children’s outcomes by family structure have evolved.
This new research investigates that question by tracking the childhood-to-adulthood outcomes of children born in the 1970s, 1980s, and 1990s. The research finds a striking divergence across outcomes. The poverty disadvantage associated with being born to an unmarried mother has diminished substantially: among children born in the 1990s, the gap in the likelihood of living above the poverty line at age 25 is small, consistent with substantial gains in the economic circumstances of unmarried-mother families. However, a large gap in four-year college completion remains: children born to married parents are much more likely to attain a four-year college degree.
This divergence in outcomes points to a more nuanced way of thinking about family structure and inequality. The relevance of having two parents in the household may not simply depend on whether a family has enough resources to avoid financial hardship, but also on whether families have the resources needed to achieve ambitious outcomes for their children.
A Marriage Gap in Household Resources
Kearney and Levine (2017) put forward a resource-based model for understanding differences in outcomes between children from married versus single mother homes. They observed that the marriage gap in children’s outcomes – defined as the difference in outcomes for children born to married and unmarried mothers – will depend on three essential elements. First, a mother brings her own resources to the household, including her time, money, networks, etc. Many single mothers have sufficient resources to keep their children out of poverty, even without the additional resources of a spouse. Second, the additional resources the household would have if the child’s other parent were also in the household depend on what that second parent would contribute. For instance, does he have a steady job and decent income? Is he helpful with the children and household chores? Because women tend to partner with men of similar economic standing – a pattern economists refer to as “assortative mating” – higher-resourced mothers tend to have higher-resourced partners. Third, how the total resources of a household translate into outcomes for children depends on the nature of the outcome being considered – namely, whether it is something that is more readily achieved with moderate resources (such as avoiding poverty) or whether it is a more ambitious outcome (such as graduating from college with a four-year degree).
Drawing on this framework, this new paper uses data from the 2023 Panel Study of Income Dynamics (PSID) to document how marriage gaps in household resources and in children’s outcomes vary along measures of maternal resources and across birth cohorts. The PSID has followed a nationally representative sample of U.S. families and their descendants since 1968. This long panel allows the authors to link children to their mothers, record the mother’s marital status and level of education at the time of the child’s birth, and observe those children when they are young adults.
The authors first document marriage gaps in household resources across decades. In these data, the share of children born to married parents fell from 75 percent in the 1970s to 66 percent in the 2000s. Throughout this period, family structure at birth is highly predictive of family structure into adolescence, meaning that children born to married parent families are much more likely to be living in a married parent household at age 14 as compared to children born to unmarried mothers. Specifically, across birth decades, between 73 and 82 percent of children born to married parents live with married parents at age 14, and less than 15 percent of children born to an unmarried mother are living with married parents at age 14.
The data also show that across all birth decades, married parent households have substantially more income. Furthermore, despite improvements in the material circumstances of single mothers, the marriage gap in household income has grown. The median household income gap at birth grew from about $39,000 for children born in the 1970s and 1980s to about $60,000 for children born in the 2010s (all in 2023 dollars). Poverty gaps shrunk, however. Children born to unmarried mothers became substantially more likely to be above the poverty line at birth, while the share for children born to married mothers held steady. This is consistent with improved material resources among unmarried mothers and reflects both changes in the composition of unmarried mothers and changes in economic and policy conditions.
Marriage Gaps in Children’s Young Adult Outcomes
Turning to longer-term outcomes, the discussion and figure below summarize the evolving marriage gap in three outcomes of interest – young adult poverty avoidance, high school completion, and four-year college degree attainment – for children born in the 1970s, 1980s, and 1990s. The high school completion gap remained roughly constant across cohorts, while the gaps in poverty and college completion moved in opposite directions (see figure).
The Marriage Gap in Young Adult Poverty Has Narrowed Substantially
The (unadjusted) marriage gap in young adult poverty avoidance has fallen over time, from a gap of about 20 percentage points for children born in the 1970s to 10 percentage points for children born in the 1990s. This reflects a decline in young adult poverty among children born to unmarried mothers, combined with a stable high rate of poverty avoidance among children of married mothers. The narrowing holds for children of mothers with and without a college degree.
After accounting for various demographic characteristics, the poverty gap for children born in the 1990s is not statistically significant at any level of maternal education. That stands in contrast to earlier decades, when the higher rate of young adult poverty among children of unmarried mothers was statistically different from that of the children of married mothers.
The Marriage Gap in College Degree Attainment Has Widened
College completion rose for children born to unmarried mothers, but by more for children born to married mothers. As a result, the college completion gap between children of married and unmarried mothers widened from 21 percentage points for the 1970s birth cohort to 29 percentage points for the 1990s birth cohort.

In terms of how this varies by maternal education, children born to unmarried mothers both without and with a college degree have seen increased college completion rates over time (4 percentage points and 13 percentage points, respectively). But again, the rise in college completion rates for children born to married mothers without and with a college degree was even larger. The result of this is an (unadjusted) marriage gap of 19 percentage points for the children of mothers without a college degree and 17 percentage points among the children of mothers with a college degree in the 1990s birth cohort. When adjusted for other changes in maternal characteristics over time, the data show that among children born in the 1990s, the adjusted marriage gap in college attainment ranges from about 5 percentage points for children of the least-educated mothers to about 20 percentage points for children of the most highly educated mothers.
Family Focus
These findings point to an asymmetry in how the link between childhood family structure and young adult outcomes has changed. With regard to poverty, the trend is encouraging. The economic circumstances of unmarried mothers and their children have generally improved in recent decades, reflecting both increased earnings and transfer income. Alongside these gains, children born to unmarried mothers in the 1990s were more likely than earlier cohorts to avoid poverty as young adults. Although differences in poverty by family structure remain, the gap has narrowed.
College completion tells a less optimistic story. Because earning a four-year college degree typically requires a substantial investment of parental resources, children are more likely to reach this milestone when they have two highly resourced parents in the home. And because women tend to partner with men of similar economic standing, marriage often adds the most resources to households that are already relatively well resourced. As a result, the college-completion gap between children of married and unmarried mothers remains wide.
These findings underscore an important distinction: meeting families’ basic economic needs and closing class gaps in access to more ambitious opportunities and outcomes are separate policy challenges.