Revisiting Harvard and the American Dream


Like many others, we have closely followed and admired the important work of a former Harvard colleague, Raj Chetty. With access to millions of anonymous tax records, Raj and his team at the Equality of Opportunity Project have powerfully confirmed what many have long feared about declining upward mobility in America.

Chetty’s new insights about the role of US colleges and universities as engines for intergenerational mobility should be considered especially relevant for us here at Harvard, and places like it. The team’s latest working paper, released alongside a set of publicly-available mobility report cards for all 2,199 U.S. colleges and universities, shines a light on the relative success (or lack thereof) of American institutions in creating pathways to economic success for low-income students.

What one might call the “manifest inadequacy of higher education’s contribution to equality of opportunity” is unfortunately not a new phenomenon. Over a decade ago at Harvard, we were similarly troubled by the striking underrepresentation of the bottom half of the US income distribution. As President in 2004, Summers introduced the most ambitious expansion to date of financial aid, specifically targeting those families of low and moderate incomes. Along with eliminating parental contributions for parents with incomes under $40,000 (since increased to $65,000), the College made changes to applicant review and launched new outreach and recruiting efforts across the country. The idea was that the most talented students should know that Harvard was an option for them, no matter their economic background.

No doubt, this has been a persistently stubborn problem to address. And yet, the data in the Chetty paper show that Harvard’s efforts, unlike those of many peer institutions, appear to have borne some fruit.

Looking at the study’s mobility report card for Harvard, the ’04 initiatives to increase low-income student attendance are apparent in the mid-2000s upticks in bottom 60% and bottom 20% student attendance (red lines of Figure 3A). Echoing prior work by colleagues here at Harvard, the fraction of students from the lower end of income distribution increased beginning with the Harvard graduating class of 2009 (entering in 2005 after the policy was announced). These gains occurred even as real incomes of low-income families were declining during this period due to widening inequality.

Notably, Harvard has outperformed peer institutions on this front (blue lines in figure 3A). The paper notes that others in the Ivy League, plus Chicago, Stanford, MIT, Duke, did not see the same increases in the fraction of low- or middle-income students, despite similar changes to financial aid policies. For example, the share of low-income students at Stanford was essentially constant from 2000-2011.  And when it comes to enabling low-income students to make it to the top as adults, Harvard excels. If success is measured as children from families in the bottom quintile reaching the top 1% of the income distribution, Chetty’s team found that Harvard ranks in the 99th percentile nationally. Harvard also places in the top 2% nationally in its share of bottom fifth students who moved to the top fifth as adults.

Of course, much work remains to be done. On most measures of access, Harvard lags behind because of its still-underwhelming bottom quintile share – Berkeley has almost double Harvard’s percentage of students from the bottom 20%. Even when using less bottom-heavy access figures (e.g. percentage of students who move up two or more income quintiles from their parents), Harvard’s skew towards the 1% limits its mobility potential.

So we know that there are a lot of low-income high-achievers that aren’t making it to Harvard, and that the economic makeup of Harvard could certainly be more diverse. It would be an unfortunate irony if “need-blind” admission policies, intended to protect low-income applicants, had the effect of preventing institutions from giving preference to applicants from modest economic backgrounds. We suspect this requires attention.

In a day and age when Congress aims to pass a tax bill with benefits massively concentrated among those at the top of the income and wealth distribution, it is more important than ever that places like Harvard become more accessible to children of poor parents, not less.  The announcement of the Harvard Financial Aid Initiative in 2004 was framed quite directly as targeting the most serious domestic problem of our time — the widening gap between the children of the rich and the children of the poor. Over thirteen years later, the sentiments from that speech still ring true: “We need to reverse the questionable allocation of national resources that results in greater, not lesser, inequality.”

By Lawrence Summers and Rachel Lipson

Previous Three (almost) inexplicable parts of the Republican tax plan Next

Comments are closed.