A potential academic contretemps between some giants in the economics profession has emerged relating to the issue of the role that colleges play in promoting intergenerational income mobility—the ability of lower-income Americans to use education as a means of achieving prosperity and sharing in the American Dream. One remarkable thing about this dispute is that it is led largely by women and nonwhite males, not the white males who dominate most of modern economics.

On the one side is Raj Chetty, a Harvard educated economist now at Stanford, and his team of colleagues located mostly at East Coast Ivy League schools. Chetty is an economics wunderkind, a recipient of the American Economic Association’s highest award, the John Bates Clark Medal, as well as the recipient of a MacArthur Foundation “genius” grant. The team’s data on incomes of students attending American colleges is widely used, recently by myself on a widely watched national news commentary show. On the other side is the remarkably prolific writer on the economics of education, Stanford’s Caroline Hoxby, arguably America’s most distinguished African-American female economist, and her coauthor, the University of Virginia’s Sarah Turner. Among other things, Professor Hoxby is known for advancing the “undermatch” hypothesis, namely that some very bright members of minority groups often end up going to schools of lesser quality than those they are capable of attending. Some able members of minority groups reduce their chances for future national distinction.

Chetty’s Opportunity Insights project uses a massive amount of federal income tax data to look at the incomes of parents of students attending college and the subsequent post-graduate earnings of their student offspring. They demonstrate, for example, that at some 38 colleges, more kids attended school from the top 1% of the income distribution than the bottom 60%. The data purportedly show the degree to which colleges promote income mobility—by low-income kids moving up the economic ladder. But the Hoxby and Turner findings show that some findings of, for example, low-income mobility, are skewed by such factors as the geographic location of the college. It is harder for the University of Connecticut to attract lots of low-income students, for example than the University of Maine because Maine is a poorer state.

The Chetty findings, heavily promoted by the New York Times, have led many colleges to alter admissions policies, but in some cases in devious and questionable ways. For example, the number of students barely eligible for Pell Grants seems to have soared (denoting greater interest in low-income students), while the number of low-income students just above the Pell Grant eligibility line has declined because colleges are gaming the system to claim they have more “Pell eligible applicants.”

I have always been a little mystified by this debate. It is unquestionably laudable to encourage kids with lower incomes to apply to good schools. But for every “undermatched” student I suspect there have been at least five that have been “overmatched” —accepted at schools above the level best fitting their academic record. This has been argued by others (e.g., Richard Sander and Stuart Taylor) to be a negative result of well-intended affirmative action programs, leading to unfortunate unintended consequences such as low bar examination passage by minority law school students.

The Chetty data show the average income of families of kids attending Ivy League schools is roughly $500,000 a year, with the median income much lower but still a hefty $200,000 or so. Our nation through tax policies and lucrative government subsidies for research favors public assistance to these schools. This is arguably inappropriate on equity grounds. The reality is that many low-income students, burdened by dysfunctional family and school attributes, have difficulty faring well in college, and some of them end up without degrees but with significant college debt. And it is true that some colleges give discriminatory treatment for children of mostly affluent alumni (legacy admissions). If colleges reach out with a wide net in recruiting students, accept them strictly based on academic potential with deep discounts for kids from poor families, they would be doing the right thing. For those seeking quotas based on income, gender, race, etc., this might seem inadequate but would be consistent with a merit-oriented system showing compassion for the less fortunate.