Report questions upward mobility at the College

Last month, the Equality of Opportunity Project released a paper, “Mobility Report Cards: The Role of Colleges in Intergenerational Mobility,” which examined economic diversity and student outcomes at colleges and universities across the country. The report indicates that the average family income for students at the College was at the 78th percentile of all families with children born in the same year, while average individual income later in life for alumni was at the 72nd percentile.

At the College, two-thirds of students are in the top quintile of family income, while 5.3 percent are in the bottom quintile and 13.4 percent are in the fourth. The student body currently draws 18.1 percent of its makeup from the top one percent of families alone. This pattern of student composition, skewed towards the highest levels of family income, was common among the 65 institutions that the study considered “elite,” including the College. The sample of all college-goers additionally had higher family incomes than the sample of all children, but was less skewed towards the top percentiles. Among all college-goers, 28.4 percent came from the top quintile, 12.4 percent came from the bottom quintile and 1.5 percent came from the top one percent.

Provost David Love offered an explanation for the lower share of students from the lower quintiles.

“The reality is that high-achieving [those scoring above 1450 on the verbal and math sections of the SAT] students from the bottom quintile are proportionally represented among Williams students,” Love said. Nationally, about 3 percent of students scoring above 1450 have family incomes of less than $20000 (approximately the upper boundary of the bottom quintile), as estimated from College Board data.

“We, of course, continue to work to identify, recruit and enroll even more talented students from the bottom income quintile, but competition for these students is especially keen,” Love said. “We also face particular challenges related to lack of brand recognition, location and misunderstanding about a liberal arts education and affordability.”

According to Carl McPherson ’14, a pre-doctoral research fellow at the Equality of Opportunity Project, family income distribution was “fairly stable” at the College during the years analyzed in the study. However, Love indicated that the share of students from the bottom three quintiles has increased since 2013, when the study’s data collection ended. For the class of 2020, 26 percent of students came from the bottom three quintiles, compared to 20 percent of students born in 1991, most of whom belong to the class of 2013.

According to Love, the College is making an effort to attract lower-income students through programs like Windows on Williams, Previews and Questbridge, along with outreach to community-based organizations and high-school counselors.

“Our efforts to attract students from lower-income backgrounds stem from the belief that talent and potential are much less likely to be correlated with family income than grades and test scores [are correlated with income],” Love said. “In the past couple of years, we have become more intentional about using specific information about family income to identify and recruit exceptional students from less advantaged backgrounds, and that effort has begun to pay off, as seen by our rising share since 2013 [of] students from the bottom three quintiles.”

The average individual income for a student at the College from a lowest-quintile family was at the 64th percentile, while the average individual income for a highest-quintile family student was at the 73rd percentile.

“It is important to us that our students experience success, but it is really up to the student to define his or her own success,” Love said. “We want to graduate students who will go out and have a positive impact on the world and be great citizens. This is as true for our students from low-income families as it is for all of our students.”

The authors of the study obtained their data from federal tax records, meaning that any students without associated tax records, including students originally from foreign countries, do not appear in the final results and data.

All family income data is from children born in 1991, while student income data is from children born from 1980 to 1982. Individual income was measured in 2014, when the 1980 to 1982 cohorts were between the ages of 32 and 34, approximately when relative income stabilizes. However, according to McPherson, data for students at the College was inadequate in this period, so income for this time period was estimated based on 1983 and 1984 cohort data.

  • David Dudley Field

    Not a bad article, but would have benefited from quoting other people, especially those with a critical take on Williams. Or are you just Dukes’ PR agent? Start here:

  • Augustine25

    We should not be surprised at Love’s comment that talent and potential are much less likely to be correlated with family income than grades and test scores. A new study of the persistent race gap in SAT scores by the liberal Brookings Institution also found class is a less powerful predictor of SAT scores than race. See,