How Does Poverty Affect Educational Outcomes?

This study asks how the duration and timing of childhood net worth poverty (NWP)—household wealth below 25% of the federal poverty line—affect educational attainment. Using 1999–2021 Panel Study of Income Dynamics data on 2,825 children tracked from birth to age 20, the authors examine high school graduation and college attendance. Results show that 75% experienced NWP, often for long spells. A 10% increase in childhood years spent in NWP reduced high school graduation by 0.5% and college attendance by 2.5%. Children experiencing NWP in over half of observed waves faced an 11–20% lower likelihood of college attendance.

Full Citation and Link to Article

Gibson-Davis, Christina M., Keister, Lisa A., Gennetian, Lisa A., & Qiu, Shuyi. (2025). Net Worth Poverty in Childhood: How Duration and Timing Affect Educational Outcomes. NBER Working Paper No. 33959. National Bureau of Economic Research. https://doi.org/10.3386/w33959

Extended Summary

Central Research Question

The central research question asks: how do the duration and timing of childhood net worth poverty (NWP) influence educational outcomes, specifically high school graduation and college attendance? NWP is defined as household wealth less than 25% of the federal poverty line, a condition that affects far more children than income poverty. While a rich literature has documented the consequences of income poverty for child development, little is known about how persistent wealth deprivation shapes human capital attainment. The study seeks to establish whether longer or more intense spells of NWP, and the age at which they occur, have lasting consequences for children’s educational trajectories into early adulthood.

Previous Literature

Prior research demonstrates that wealth plays a central role in children’s life chances, enabling investment in housing, enrichment activities, and higher education, while also buffering households against economic shocks. Unlike income, which reflects a flow of resources, wealth reflects accumulated stock, with enduring intergenerational effects. Studies have consistently found that higher parental wealth predicts better child outcomes, even after controlling for income (Conley 2001; Pfeffer 2018; Miller et al. 2021).

In contrast, net worth poverty—defined as insufficient wealth to meet basic needs for at least three months—has only recently been conceptualized (Haveman & Wolff 2004, 2005). Gibson-Davis and colleagues (2021, 2022, 2024) show that NWP is common and racially stratified, and that it negatively predicts children’s test scores and adult health. However, no studies prior to this one directly examined NWP’s long-term association with educational attainment.

The literature on income poverty shows that its effects vary by persistence and timing, with prolonged or early-life poverty producing the strongest negative outcomes (Duncan et al. 2010; Page 2024). Researchers hypothesize that wealth deprivation may operate differently, potentially becoming more salient during adolescence when awareness of financial status and resource demands for college increase. Yet evidence on developmental timing for wealth effects has been mixed (Miller et al. 2021).

Thus, the study contributes to filling two major gaps: (1) assessing whether the cumulative duration of NWP matters for schooling, and (2) examining whether exposure at particular developmental stages (early childhood, middle childhood, adolescence) is especially consequential.

Data

The authors use data from the Panel Study of Income Dynamics (PSID), a nationally representative longitudinal survey of U.S. households with detailed wealth and income measures collected biennially since 1999. The analytic sample includes 2,825 children observed from birth or early childhood (ages 0–5) in 1999 or 2001 and followed until they turned 20, yielding up to 10 survey waves per child.

Wealth was measured as assets (e.g., real estate, checking, savings, stocks) minus debts (e.g., mortgages, loans, credit cards), excluding pensions and vehicles. NWP was defined as net worth less than 25% of the federal poverty line, adjusted for family size. For reference, in 2022 a family of four with two children was net worth poor if their assets net of debts were below $7,419.

Key dependent variables were binary indicators of (1) high school completion (by diploma or GED) and (2) college attendance (any postsecondary enrollment) by age 20. Independent variables measured NWP exposure (ever vs. never, or proportion of waves spent poor), duration (number and length of spells), and timing (proportion of NWP exposure occurring in early, middle, or late childhood).

The study also measured income poverty (IP), defined relative to federal poverty thresholds, to test whether NWP effects are independent of income. Control variables included child gender, parity, parental education, race/ethnicity, marital status, household size, and health.

Methods

The study employed descriptive analyses and linear probability models (LPMs) with clustered standard errors to estimate the probability of high school graduation and college attendance as a function of NWP. While logistic models were also tested, LPMs were chosen for ease of interpretation; results were robust across specifications.

Five hypotheses guided the analysis:

1. NWP exposure is more persistent than income poverty.

2. NWP negatively predicts high school and college outcomes.

3. Duration and intensity of NWP matter, with longer and more consecutive spells associated with worse outcomes.

4. Developmental timing matters, with adolescence hypothesized to be the most sensitive period.

5. NWP effects remain significant after controlling for income poverty.

To address missing data, the authors used multiple imputation. Robustness checks examined whether results differed when wealth definitions included vehicles or pensions, or when the NWP threshold was varied (50% or 200% of the poverty line). They also tested differences between asset and debt poverty.

Findings/Size Effects

The findings confirm that NWP is widespread and enduring. Fully 75% of children experienced at least one wave of NWP, compared to 49% who ever experienced income poverty. On average, children spent 36% of observed waves in NWP, and 31% were NWP for more than half of all waves. Moreover, 21% of children experienced at least five consecutive waves of NWP, far higher than the 5% who experienced equivalent spells of income poverty .

Educational outcomes were strongly associated with NWP exposure and duration. Key size effects include:

• Proportion of time in NWP: A 10% increase in childhood years spent in NWP reduced high school graduation by 0.46% and college attendance by 2.5% (p < .001).

• Duration of spells: The length of the longest NWP spell predicted outcomes more strongly than the number of spells. A long NWP spell was associated with a 3.2% decline in college attendance, six times larger than its effect on high school graduation.

• Intensity: Children NWP for at least half the waves faced an 11% (nonconsecutive) to 20% (consecutive) lower likelihood of college attendance compared to children never NWP.

• Timing: NWP during adolescence was particularly harmful for high school graduation, with a 0.49% reduction per 10% increase in adolescent years spent poor. For college attendance, both early childhood and adolescent exposure were detrimental; adolescence effects (–12.9%) were about twice as large as early childhood effects (–6.1%).

• Comparison with income poverty: IP was more strongly associated with high school completion, while NWP had larger and more consistent effects on college attendance. The longest spell of IP reduced high school completion by 2.3%, versus just 0.5% for NWP, but NWP reduced college attendance twice as much as IP (–3.1% vs. –1.6%).

Robustness checks confirmed that results were driven primarily by asset poverty, not debt poverty, and that findings held under alternative definitions of wealth thresholds.

Conclusion

This study makes three major contributions to the literature on poverty and education. First, it documents the prevalence of net worth poverty, showing it is more common, persistent, and “stickier” than income poverty. Second, it provides the first estimates linking NWP to human capital attainment, finding that NWP is independently and negatively associated with both high school graduation and especially college attendance. Third, it demonstrates that the effects are dose-dependent: longer and more intense spells of NWP produce larger declines in educational attainment.

The findings challenge the prevailing income-centric view of childhood poverty by showing that wealth deprivation exerts a distinct influence on children’s educational opportunities. While income poverty is more predictive of high school completion, net worth poverty is particularly salient for college attendance—a resource-intensive transition requiring accumulated assets and financial confidence. Adolescence appears to be a sensitive period, but risk persists regardless of developmental stage.

Policy implications are profound. Programs that focus solely on raising income may overlook the long-term disadvantages created by the absence of wealth. Addressing NWP may require asset-building policies, such as children’s savings accounts, wealth transfers, or targeted debt relief, that enhance households’ ability to accumulate and retain wealth. By incorporating wealth into poverty definitions and interventions, policymakers could better address the full spectrum of economic disadvantage shaping children’s educational and life chances.

In sum, the study reveals that net worth poverty is the modal form of economic hardship for American children and a significant predictor of educational inequality. Its persistent and cumulative nature underscores the need for policies that build not only income but also wealth, ensuring children have the resources to pursue both high school and college completion.