In late 2015, William Emmons, lead economist at the Center for Household Financial Stability (the Center) at the Federal Reserve Bank of St. Louis, and Bryan Noeth, former Center analyst, published an article documenting a troubling result. While having a college degree predicted rising levels of family wealth between 1992 and 2013 for median (or typical; those in the middle of the distribution) college-educated white and Asian families, it predicted declining levels of wealth over that period for median college-educated black and Hispanic families. Moreover, the declines were larger in percentage terms for median college-educated black and Hispanic families than for median black and Hispanic non-college-educated families, respectively.1
To be sure, a college education generally leads to more income and wealth for families of all races and ethnicities, albeit to different degrees. But the change in wealth over those two decades was a different story. To explore these puzzling findings, we organized a research symposium in May 2016. The research papers, videos of the proceedings and a summary of the symposium are available on our website.2
This article summarizes the paper prepared by Emmons and Lowell Ricketts, the Center’s lead analyst, which served as an exploration of the links between race and ethnicity, and education and wealth, as well as a framework for the symposium itself.
The key contribution of the Emmons–Ricketts paper was to challenge standard economic models that explain differences in wealth outcomes across race and ethnicity by pointing to differences in financial choices or behavior alone. These variables—such as the types and amounts of particular assets and liabilities a family has, or the number of adults or children in the household—are termed “observables” because we can see and measure them. But we know there also are “unobservable” factors—including an individual’s experiences, environment and peer groups—that the researcher cannot see or measure. These unobservable factors may, in fact, be quite important. For example, differences in the quality of education received or peer-group norms relevant for financial decision-making, or subtle forms of discrimination could be decisive in many adult outcomes. The Emmons–Ricketts research design allowed for the possibility that some of these unobservable influences could be related to one’s race or ethnicity.
This research shows that a large set of observable factors can indeed “explain” or account for much of the wealth gaps between whites and all major nonwhite groups in a statistical sense; this confirms what many other researchers have found. But they note that models of this type typically assume that all racial and ethnic groups 1) generate equal returns to education and 2) face the exact same choices and opportunity structures.
Emmons and Ricketts posit an alternative framework that assigns differences in education, family structure, financial decisions and other advantages or disadvantages (such as receiving an inheritance or experiencing poor health outcomes) not to individual choice alone but also to structural, systemic or other unobservable factors at least partially related to race or ethnicity. For example, the head of a young Hispanic family with little formal education likely makes financial decisions at least in part with reference to the choices made by other families in the community.
While differences in observable financial behaviors and choices—such as high concentrations of wealth in housing, high debt burdens or a low level of education—intuitively appear to drive diverging wealth outcomes, the underlying reasons for these choices are unobservable and often ignored. If one allows for the possibility that every family has a unique set of experiences and opportunities and that, even in 21st-century America, race and ethnicity shape many aspects of our lived experiences, it is not so obvious what the real underlying causes of different outcomes are.
The results of this alternative framing of cause and effect are strikingly different than the standard interpretation that “blames the victim” for making bad choices that lead to low wealth. Emmons and Ricketts conclude:
“[A] structural-determinants framework suggests the vast majority of black-white and Hispanic-white wealth gaps may lie beyond the scope of individual actions or marginal policy changes directed at educational attainment, family structure, financial decision-making or even wealth redistribution. Instead, the gaps appear to be deeply rooted in unobservable factors [that underlie observable financial choices] that may include discrimination or other long-lasting disadvantages.”
While it is clear that the answer to the symposium’s title question, “Does college level the playing field?,” is “No,” or at least that college alone doesn’t level the field, the symposium did make progress in explaining why the wealth outcomes of college-educated Hispanics and blacks have diverged so much from whites and Asians since 1992. A plausible explanation is that structural, systemic or other unobservable factors related to race and ethnicity may create very different choices, opportunities and peer-group reference points for families of different races and ethnicities.
Reframing the inquiry into wealth gaps opens up the possibility of new questions and alternative explanations. Rather than sticking to the conventional “post-racial” behavioral view that assumes equal returns on education and access to equal financial opportunities regardless of race or ethnicity, symposium participants were challenged to relax both assumptions.
Research alone, of course, will not reshape the lives of individuals and families who face unequal opportunities and constrained choices. But it is likely necessary to provide a road map for this long and overdue journey.
Keep up with what’s new and noteworthy at the St. Louis Fed. Sign up now to have this free monthly e-newsletter emailed to you.
FedCommunities.org is a portal to community development resources from all 12 Federal Reserve Banks and the Federal Reserve Board of Governors.