Association for Public Policy Analysis & Management

In this paper, I identify the effect of extended family net wealth---total assets minus debts belonging to aunts, uncles, and grandparents---on college enrollment for high school graduates ages 18 to 24. The degree to which wealth contributes to determining higher education outcomes such as enrollment and degree attainment is not well understood. Within this literature, only parental wealth has received any treatment. In contrast, this paper studies the relationship between extended family wealth and college enrollment.

I start with Lovenheim (2011), which estimates the effect of exogenous wealth shocks on college enrollment, and present results for an extension that adds a measure of exogenous shocks to extended family wealth. Using this empirical methodology, preliminary results indicate that exogenous wealth shocks do not affect college enrollment. But there are significant limitations to this approach.

Lovenheim (2011) uses a narrow concept of wealth limited to residential equity. The instrument cannot identify a causal effect of wealth changes for any family member who lives in a rented home or apartment. To overcome this limitation, I present results for a multivariate regression using a broader measure of wealth for parents and extended family. In the next section, I summarize these results and propose ideas for including a broad measure of extended family wealth within a quasi-experimental research design.

In the last section, I discuss the relevance of this work for policy, including federal need-based college aid programs (FAFSA, federal student loans) and tax advantaged savings plans. Public policy aims to promote equal access to education, regardless of student ability to pay. There are many low-cost options for students, yet college enrollment and college completion remain highly unequal by wealth and income quintile. Attention to extended family wealth may reveal biases in how expected family contributions are calculated.