Abstract
In the lead-up to the 2008 Financial Crisis, average household debt grew significantly. Some evidence suggests that female-headed households may have been disproportionally represented in this increase in indebtedness. This analysis uses Survey of Consumer Finance data from 1995 to 2013 to investigate whether female-headed households experienced greater growth in debt during the pre-crisis credit expansion relative to male-headed households and whether this debt persisted post-crisis. The results indicate that female-headed households account for most of the growth in mortgage debt among lower income households during the credit expansion. Younger female household heads exhibited greater growth in average educational debt than young male household heads prior to the crisis, and growth in educational debt post-crisis was larger for older female household heads. This rising indebtedness among lower income female-headed households was not accompanied by sustained wealth or income growth, leading to increases in leverage and financial fragility.
Acknowledgments
The author wishes to acknowledge the valuable feedback provided on previous versions of this paper by the attendees of the ASE Session on the 2008 Economic-Financial Crisis at the 2018 ASSA Annual Meeting as well as by Alexandra Bernasek, Steven Pressman, other faculty and graduate students at the Colorado State University Department of Economics, and the anonymous referee. All remaining errors are solely the responsibility of the author.
Notes
1 See Dymski, Hernandez, and Mohanty (Citation2013) for a survey of work relevant to race and gender in the area of mortgage debt.
2 The SCF uses a multiple imputation procedure to account for missing values. For this analysis, I do not adjust for the minor additional variance introduced by this technique and use the first of the five implicates for each observation.
3 This focus on debt-to-wealth and debt-to-income ratios neglects the role of educational borrowing on increased future income through human capital investment. The cross-sectional structure of this data-set precludes an analysis of the relationship between educational debt and income trajectories.
4 Results are qualitatively robust to use of original values.
5 Results including estimates for control covariates are available upon request.