Is there a marginal propensity to borrow? Lessons from the effect of mortgage debt on borrowing for higher education

We examine the relationship between tendencies to borrow for homeownership as reflected by trends in the mortgage market, and tendencies to borrow for higher education as reflected by student loans. Over the past 15 years, we recognize two main phenomena: first, a ‘reversed U shape’ trend in the mortgage market—sharp increase until 2008, followed by a significant decline; second, a steady increase in the amount of student debt. Controlling for other relevant variables, such as tuition costs, enrollment at postsecondary institutions, and median family income, we examine the debt-profile of borrowers and the preference allocation with respect to homeownership and higher education as substitutes. While prior research addresses the effect of student loans on homeownership, we examine the reverse effect, particularly for a select group of borrowers.

For the purpose of this paper, we combine individual-level credit bureau data from the Federal Reserve Bank of New York/Equifax Consumer Credit Panel (CCP) with income information from the Bureau of Economic Analysis (BEA), and Integrated Postsecondary Education Data System (IPEDS) -derived tuition and enrollment information from DeltaCost. We use a 1% Equifax sample, which covers the period between 2002 and 2013. The sample includes approximately 40 million individuals in each quarter. Our data include not only general information about each borrower, such as age, credit score, and zip code but also more granular information about individual loans such as account opening date, current balance, and payment status. The richness of our data enables us to present an analysis based on age cohort and individual credit scores.

We further investigate the causal effect by using the most recent economic crisis and its consequences for the mortgage market as an independent variable, and perform regression discontinuity analysis based on credit score thresholds for mortgage access.

Our preliminary results show that there is a substitution between homeownership borrowing and higher education borrowing. The strongest effect is amongst the youngest age cohorts and for individuals who tend to borrow more in the first place. We conclude that the post-crisis restricted access to mortgage credit increased the access to higher education for a select group over our period of analysis.

Based on our results, we are able to develop and introduce a new theoretical concept of marginal propensity to borrow.