How Did Falling Housing Prices During the Great Recession Affect Children’s Educational Outcomes?

This study investigates how household exposure to housing price declines—specifically, negative equity—affected children’s educational outcomes during the Great Recession. Using administrative data from Florida public schools (2002–2011) and housing transactions, the authors exploit variation in home purchase timing among in-state newcomers to estimate impacts. Contrary to literature on foreclosure, they find that higher predicted negative equity is associated with increased test score growth, especially among Black students and those eligible for free or reduced-price lunch. A 0.69 increase in predicted loan-to-value (LTV) ratios led to test score gains of approximately 0.13 standard deviations, indicating sizable and counterintuitive academic improvements.

Full Citation and Link to Article

Vicki Been, Ingrid Ellen, David Figlio, Ashlyn Aiko Nelson, Stephen L. Ross, Amy Ellen Schwartz, and Leanna Stiefel. 2024. “The Effects of Housing Price Declines on Children’s Educational Outcomes.” Education Finance and Policy, 20(2): 312–360. https://doi.org/10.1162/edfp_a_00431

Extended Summary

Central Research Question

This study asks whether and how household exposure to negative equity—owing more on a mortgage than a home is worth—impacted children’s academic outcomes during the Great Recession. Specifically, it investigates whether declines in housing prices, leading to negative equity, had detrimental, neutral, or even beneficial effects on student test score growth. The paper stands in contrast to the dominant narrative that housing distress, such as foreclosure, unequivocally harms children’s educational trajectories. By focusing on Florida, a state hit particularly hard by the housing crisis, the authors assess whether housing market shocks altered children’s academic performance and, if so, through which mechanisms.

Previous Literature

Much of the existing literature has examined the effects of foreclosure and housing instability on children’s outcomes, generally finding negative associations. For instance, Bradbury et al. (2013) and Swain (2013) document lower test scores and increased absenteeism among children experiencing foreclosure-related moves. Others, such as Been et al. (2011) and Comey and Grosz (2011), highlight how foreclosure disrupts school continuity and leads to residential instability, with cascading consequences for children’s well-being.

However, the literature offers limited insights into the role of negative equity independent of foreclosure. Some research suggests that negative equity may reduce mobility (Chan 2001; Ferreira et al. 2010), potentially stabilizing school environments. Others propose that it can enable families to maintain consumption by defaulting strategically (Zhang 2017), potentially preserving household investments in children. Theoretical ambiguity persists: while negative equity could raise stress and lower investment in children, it might also create unexpected opportunities by altering constraints and incentives.

Data

The analysis draws on an extensive dataset covering Florida public school students from 2002 to 2011, linked to property-level housing transaction data from Dataquick, Inc. This unique linkage allows the authors to calculate a predicted loan-to-value (LTV) ratio for student households. The student data include standardized test scores, demographic information, attendance, and disciplinary records. Importantly, the sample is restricted to students who were new to Florida public schools in the 2003–04, 2004–05, or 2005–06 academic years and were observed in tested grades (3–8) between 2006 and 2011.

The housing data include transaction prices, lien information, and property characteristics, enabling the researchers to build price indices and assess LTV ratios across time and counties. By leveraging variation in the timing of families’ entry into Florida and subsequent county-level housing price trajectories, the authors create a proxy for exposure to negative equity.

Methods

To isolate the causal impact of negative equity on academic performance, the study implements a triple-difference (difference-in-difference-in-differences) identification strategy. The core approach compares changes in student test score growth across different arrival cohorts (2003–2005) and counties with varying levels of housing price declines, while controlling for fixed characteristics at the student, school-year, and arrival cohort-year levels.

The primary explanatory variable is predicted LTV, constructed from county-level housing price indices and distributions of mortgage LTVs. This variable proxies for the likelihood that a student’s family experienced negative equity at a given time. Because direct observation of homeownership and individual mortgage terms is unavailable for all students, the measure is aggregated at the county-cohort-year level. Importantly, the analysis focuses on out-of-state movers to reduce selection bias—these households are less likely to have made housing decisions based on intimate knowledge of local real estate trends.

The regression models control for student fixed effects (accounting for time-invariant traits), school-by-year fixed effects (capturing localized shocks and trends), and arrival quarter-by-year fixed effects (adjusting for cohort-specific trends). This rich set of controls enables a credible estimation of how exposure to housing market downturns influenced student outcomes net of confounding factors.

Findings/Size Effects

Contrary to the prevailing literature on foreclosure, the authors find that greater exposure to negative equity is associated with higher test score growth, especially for disadvantaged students. For example, a 0.69 increase in predicted LTV (the average observed for middle-cohort students during the housing crisis) is associated with an increase of approximately 0.13 standard deviations in combined reading and math test scores over four years. This magnitude is comparable to the effects of well-established education interventions such as charter school enrollment (Angrist et al. 2016).

Reading scores see slightly larger gains than math scores. The authors also document an increase in school suspensions, albeit smaller in effect size—approximately a 1.3 percentage point rise, or a 7 percent increase from the mean, associated with high predicted LTV. Effects on absenteeism are statistically insignificant.

Strikingly, the positive effects are most pronounced among Black students and those receiving free or reduced-price lunch. Among these groups, test score gains are as high as 0.18–0.19 standard deviations, nearly triple the gains for White students and students not eligible for subsidized lunch. Gains are also larger in areas with lower initial housing prices, where disadvantaged families were more likely to enter homeownership during the housing boom.

The authors propose two mechanisms to explain these counterintuitive findings. First, some families facing negative equity may have prioritized consumption for children by defaulting on mortgage payments, particularly given long foreclosure timelines in Florida. This interpretation is supported by external data from the Consumer Expenditure Survey, which shows that Florida homeowners were better able to sustain consumption during the crisis compared to renters in the same state and homeowners elsewhere. Second, falling housing prices may have temporarily lowered barriers to high-performing schools, especially for renters or buyers in neighborhoods that previously had been inaccessible. Indeed, the study finds a 7 percentage point increase in the likelihood of moving to a school with a higher state-assigned report card grade for students in high predicted LTV counties.

Notably, these mechanisms are not mutually exclusive and may have operated in tandem. The improvements in school quality access are broadly shared, but the academic gains are concentrated among disadvantaged students, possibly because they benefit more from improvements in educational environments.

Conclusion

This study offers a surprising and nuanced contribution to the literature on housing shocks and child outcomes. While the broader literature emphasizes the harms of foreclosure and housing instability, this analysis suggests that negative equity alone—when decoupled from immediate displacement—may yield temporary academic benefits under specific circumstances. These benefits are especially pronounced for economically and racially marginalized groups.

Importantly, the results do not imply that negative equity is beneficial per se, but rather that it may disrupt longstanding structural barriers, allowing some families to improve educational opportunities for their children. The findings underscore the relevance of policy measures like mortgage forbearance and eviction moratoriums that enable families to remain in their homes during economic downturns. They also highlight the role of housing markets in shaping access to quality education.

Finally, while the gains observed may be temporary, the mechanisms suggest that even short-term access to better schools and reduced consumption constraints can produce lasting educational benefits, consistent with other literature on teacher quality and school environments. The paper challenges assumptions about housing distress and calls for a more differentiated understanding of how economic shocks interact with institutional structures like education.