Re-Examining the Role of Elite Higher Ed in American Society

Princeton. Ivied walls or moss-backed walls?

by Reed Fawell III

“Going to Yale Could Make You Rich, or Lonely,” by Lyman Stone, published in The Federalist on Dec. 19, 2018, exposed some surprising findings regarding the costs and benefits of college attendance. Stone is worth quoting at length:

There’s a long-standing economic consensus that, for high schoolers smart enough to get admitted into the University of Kentucky (average SAT score of about 1000-1100) and Yale (average SAT score of about 1400-1600), it really doesn’t matter which college they attend … [Researchers have identified] how much money students earn 10 or 20 years later. It turns out going to Yale doesn’t add one penny to how much money a Yale admit earns.”

At least, that’s what economists used to think. But that research was based on a sample that basically dropped all part-time workers and excluded almost all women for various reasons. Brand-new economic research released just last week, however, overturns this long consensus. Economists have now found … that, for a male student admitted at a highly selective school like Yale, it doesn’t matter where he goes. If he goes to Yale, he’ll do fine surrounded by brilliant peers. If he goes to the University of Kentucky, he’ll do fine as a big fish on campus.

But for women, attending a highly selective school has massive effects. A young woman admitted to both UK and Yale faces a resounding choice about her future life. If she chooses Yale, odds are that her annual income when she is 40 will be about 40-70 percent more. However, her odds of ever getting married are about 25 percentage points, or about one third, lower. Crucially, her odds of having a higher income rise only if she gets married! So really, there are three outcomes:

  1. Outcome one: our hypothetical female student goes to UK, gets a degree, has about a 75% chance of being married by age 40, and probably makes about $50,000 a year.
  2. Outcome two: our hypothetical student goes to Yale, gets a degree, is still unmarried at age 40, and still only makes about $50,000 a year. If she goes to Yale, she has about 50% odds of this outcome.
  3. Outcome three: our hypothetical student goes to Yale, gets a degree, gets married, and ends up making about $100,000 a year. There’s about 50% odds of this outcome.

In other words, Yale does not make women better off: It makes women who win the meritocracy tournament better off, and leaves the rest no better off, but with plenty of debt and no spouse. Crucially, the study authors can’t say why this is happening, but a substantial driver seems to be about spousal characteristics. Going to Yale seems to make women marry much higher-earning men than going to UK does, even for women of similar backgrounds and SAT scores. Either Yale women find these men, get married, and end up as a wealthy power couple, or they hold out for the perfect guy … forever.

… Nobody tells young women this when they are 18. Nobody lays it out for them what kind of choices they’re facing, and what the odds actually are. … Those odds won’t look so great to a lot of women who get into highly selective schools.

The basic finding that it does not matter where you go to college if you are a highly qualified student was first published 16 years ago. As explained this past Dec. 11 by Derek Thompson, staff writer at Atlantic, in “Does It Matter Where You Go to College?

“In November 2002, the Quarterly Journal of Economics published a landmark paper by the economists Stacy Dale and Alan Krueger that reached a startling conclusion. For most students, the salary boost from going to a super-selective school is “generally indistinguishable from zero” after adjusting for student characteristics, such as test scores … Dale and Krueger even found that the average SAT scores of all the schools a student applies to is a more powerful predictor of success than the school that student actually attends.

These findings raise collateral issues, such as impacts on minorities and the opting out of the trained profession by women in elite universities, some of them yet resolved, but all important and worthy of attention. For example, again quoting for the same Atlantic Magazine article:

This month, economists from Virginia Tech, Tulane, and the University of Virginia published a new study that reexamines the data in the Dale-Krueger study. Among men, the new study found no relationship between college selectivity and long-term earnings. But for women, “attending a school with a 100-point higher average SAT score” … (can) have is a huge effect. Has one of the most famous papers in education economics been debunked? Not quite, says Amalia Miller, a co-author and an economist at the University of Virginia. “The difference we found is that college selectivity does seem to matter, especially for married women … For the vast majority of women, the benefit of going to an elite college isn’t higher per-hour wages. It’s more hours of work. Women who graduate from elite schools delay marriage, delay having kids, and stay in the workforce longer than similar women who graduate from less-selective schools.

A few obvious questions arise from these findings.

  1. Should we rethink how college rankings impact an applicant’s post-education outcomes?
  2. Might college rankings imply a benefit to students where there is none and omit other significant impacts students should know before selecting a college?
  3. Is the great variation between schools’ tuition justified?
  4. How can we ensure that these findings, known to the educational establishment since the 2002, are known to students (and their parents) when they apply to colleges?
  5. Do these findings justify substantial changes in the entire college application process?

The findings also touch upon broader economic and social issues such as soaring college costs, ballooning student debt, falling marriage and birth rates, campus unrest, and increasing polarization of the citizenry.

Might these demographic trends be driven, in part, by the growing divide between students who attend elite universities and every one else, whether they go to college or not? Is it good social and economic policy to perpetuate this divide? Is it good education policy?

Indeed, should we reconsider the role that the elite higher ed institutions — not just the Harvard, Yale, and the other Ivies but public elites such as the University of Virginia and the College of William and Mary — play in our society?

As stated by Professor Patrick J. Deneen, in “The Ignoble Lie, How the New Aristocracy Masks its Privilege, in the April 2018 edition of First Things Magazine:

Our ruling class is more blinkered than that of the ancien régime. Unlike the aristocrats of old, they insist that there are only egalitarians at their exclusive institutions. They loudly proclaim their virtue and redouble their commitment to diversity and inclusion. They cast bigoted rednecks as the great impediment to perfect equality — not the elite institutions from which they benefit. The institutions responsible for winnowing the social and economic winners from the losers are largely immune from questioning, and busy themselves with extensive public displays of their unceasing commitment to equality. Meritocratic ideology disguises the ruling class’s own role in perpetuating inequality from itself, and even fosters a broader social ecology in which those who are not among the ruling class suffer an array of social and economic pathologies that are increasingly the defining feature of America’s underclass. Facing up to reality would require hard questions about the agenda underlying commitments to “diversity and inclusion.” Our stated commitment to “critical thinking” demands no less, but such questions are likely to be put down—at times violently—on contemporary campuses. …

This helps explain the strange and often hysterical insistence upon equality emanating from our nation’s most elite and exclusive institutions. The most absurd recent instance was Harvard University’s official effort to eliminate social clubs due to their role in “enacting forms of privilege and exclusion at odds with our deepest values,” in the words of its president. Harvard’s opposition to exclusion sits comfortably with its admissions rate of 5 percent (2,056 out of 40,000 applicants in 2017). The denial of privilege and exclusion seems to increase in proportion to an institution’s exclusivity.”

Share this article


(comments below)


(comments below)


6 responses to “Re-Examining the Role of Elite Higher Ed in American Society”

  1. Dick Hall-Sizemore Avatar
    Dick Hall-Sizemore

    The conclusions implied by the comments and questions in the post are counter-intuitive. But it is hard to argue with data–until you dig into it a little bit. First of all, from my limited research, Dale and Kruger did not really compare the effects of getting into “elite” schools with the earning potential of getting into any other school. The authors conducted two surveys, one in 1976 and one in 1989. Here are the schools included in those surveys:

    1976 survey
    Barnard College, Bryn Mawr College, Columbia University, Duke University, Emory University, Georgetown University, Miami University, Morehouse College, Northwestern University, Oberlin College, Penn State University, Princeton University, Smith College, Stanford University, Swarthmore College, Tufts University, Tulane University, University of Michigan, University of Notre Dame, University of Pennsylvania, Vanderbilt University, Washington University, Wellesley College, Wesleyan University, Williams College, Xavier University, Yale University,

    1989 survey
    Bryn Mawr College, Duke University, Georgetown University, Miami University of Ohio, Morehouse College, Oberlin College Penn State University, Princeton University, Stanford University, University of Michigan, University of Pennsylvania, Vanderbilt University, Washington University, Wellesley College, Wesleyan University, Williams College, Xavier University, and Yale University.

    No JMUs or VCUs in that list. No William and Mary or UVa. either. And in response to Mr. Stone of The Federalist article, UK was not in the list to be compared with Yale, either. So, the list seems to include “super elites” and “sort of elite”.

    Even Alan Krueger pointed out that the results are not applicable to all categories. The pattern of no difference in earning potential did not apply to blacks, Latino students, low-income students, or students whose parents did not graduate from college.

    Finally, Derek Thompson, whose Atlantic article is quoted extensively, has this summary:
    “The simplest answer to the question “Do elite colleges matter?” is: It depends on who you are. In the big picture, elite colleges don’t seem to do much extra for rich white guys. But if you’re not rich, not white, or not a guy, the elite-college effect is huge. It increases earnings for minorities and low-income students, and it encourages women to delay marriage and work more, even though it doesn’t raise their per-hour wages.”

    1. smoretva Avatar

      Good point, Dick re: the relatively narrow selection of institutions included in the survey. It would be nice to see such surveys and analysis completed with a broader array of institutions, including in particular some that are considerably less selective than those in the sample.

  2. LarrytheG Avatar

    re: ” it depends on who you are”.

    indeed. Some of the most successful people come from little-known colleges and others were dropouts.

    I think someone who comes from a family of successful people – someone who is raised by successful people has a huge advantage over those who do not – and the choice of college is almost secondary.

    Is there a Good-old-boy- network? Yep. it’s an added advantage to those that need it but those who have already made their own way – less so.

    The folks who have entrepreneurship in their families are the most fortunate because they learn that no single thing – including education – is a plus or a minus… it’s what you make of the things you do have – and how you overcome the things that are challenges.

    The most successful folks are folks who have learned from failures…

    make the most of every single experience you have – learn from it and benefit from it.

    Zuckerberg – a dropout – was/is a genius in terms of seeing an opportunity in the market. All these fancy educated folks ahead of him – failed to see what he saw … and you can say that about Bill Gates, and dozens of others… where “education” was not the deciding factor – elite or not.

    1. Reed Fawell 3rd Avatar
      Reed Fawell 3rd

      Right you are, Larry.

  3. smoretva Avatar

    Reed, thanks for mentioning that interesting new paper by Ge, Isaac, and Miller. I previously had read the important papers on this topic by Dale and Krueger. I have not yet carefully studied the new paper, but I reviewed it closely enough to provide some initial reactions to your questions:

    1. Should we rethink how college rankings impact an applicant’s post-education outcomes?

    Yes, definitely. That has been overdue for some time.

    2a. Might college rankings imply a benefit to students where there is none…?

    Here I think the main question is what benefits are being considered. I attended a big state school for undergrad, which at the time had nearly open admissions, and then two highly selective, elite graduate schools. While the findings in the referenced papers largely resonate with me, at least relative to full-time employed college graduates, I do think I received a lot of benefits from attending the elite institutions. While I worked very hard at all three places, the two elite schools offered a level of access of several kinds (very talented and ambitious peers; well-connected alumni in leadership positions across every major industry sector and every major country; on-campus interviews with top firms, etc.) that was simply amazing. I learned a great deal at all three places. The elite schools made it easier to gain access to other educational and professional opportunities, and they also offered far greater institutional support due to their far greater institutional resources (e.g., well-staffed, comprehensive career services; ample counseling and academic support; relatively easily accessed field projects with top firms and countries around the world).

    2b. Might college rankings…omit other significant impacts students should know before selecting a college?

    Yes, but not the negative ones implied by Mr. Stone.

    I have been impressed by some of Mr. Stone’s writing in other contexts, but in the Federalist article I think he has oversimplified the findings in the new paper, at least based on my cursory review of that paper. First and most important, the new paper clearly affirms the previous findings of Dale and Krueger for full-time employed college graduates, i.e., that there is no statistically significant earnings benefit of attending an elite institution (on average), once certain individual characteristics are taken into account — a finding that is true for both men and women. Only when considering part-time or non-working individuals do the new findings emerge relative to women. I’m just not sure we should jump to conclusions there. People who don’t work or work part-time sometimes do so for economic reasons (e.g., they prefer full-time work but can’t secure it for some reason), while other times they do so for non-economic reasons (e.g., they need to care for children or a family member). I have not yet seen a rationale for why we should ascribe those choices or outcomes to whether someone attended an elite university.

    The new paper suggests that women who attend elite universities may set a higher bar for who they would be willing to marry, causing them in some cases to not be married at higher rates than other college graduates their age. If true, is that the result of the college they attended, or is that more a reflection of the individuals involved?

    One important technical note here: In discussing the new paper, Mr. Stone implies that its authors knew whether women in the sample were married by age 40. Unless I misread the paper, the referenced survey doesn’t include such data. Rather, I think it conveys whether the surveyed women were married at the point they were surveyed, typically around their late 30s. Accordingly, many of those who were unmarried at that time actually were divorced, not never married. Data I have seen elsewhere ( suggests that nearly a third of women in the U.S. with a bachelor’s degree or higher who were born in the time range of those in the sample used in the new paper were divorced at least once by age 46. If Mr. Stone has otherwise correctly summarized the results of the new paper, it could be that more than half of those he characterized as “still unmarried at age 40” actually were divorced. Presumably many of those divorced women were left with childcare responsibilities that limited their career options and/or working hours, resulting in lower earnings.

    3. Is the great variation between schools’ tuition justified?

    With the large discount rates in place at many institutions, tuition really just represents a sticker price, not the cost of attendance. I think I recall correctly that some more selective institutions often offer a total cost of attendance similar to that of less selective public schools, at least for well qualified, lower-income students. Having said that, I suspect they generally are more expensive to attend. I think what is really happening here is that selective institutions often will charge what the market will bear. They are always interested in increasing their quality and standing, aims that generally can be facilitated with additional financial outlays for faculty, staff, facilities, and financial aid (to attract more exceptional students and/or to enable more qualified lower-income students to attend), among other things.

    4. How can we ensure that these findings, known to the educational establishment since the 2002, are known to students (and their parents) when they apply to colleges?

    As I’ve written previously, I think we need to improve the availability and ease of use of data illustrating linkages between individual characteristics (including academic preparation, interests, and ambition), higher education, and the labor market. A great deal of literature suggests that one of the most broadly embraced opportunities to enhance future labor market outcomes for employers, current high school students, and/or college students would be to expand the availability of high-quality data illustrating linkages between academic qualifications, degree programs (and completion rates), higher education institutions, occupations, earnings, and locations, or some subset of those, ideally including both short- and long-term outcomes. Such an initiative largely could be accomplished by linking samples of student records of high school and college graduates with wage records, de-identifying the results to protect the confidentiality of individuals. Some of this work already has been done, but I’ve not yet seen it done via a comprehensive, robust, and easy-to-use platform. I’m hoping Virginia will be one of the first states to do so.

    With improved understanding of such linkages, students and their families could make more informed educational and career decisions, and employers would benefit from market signals being more clearly conveyed to potential future employees. Likewise, counselors in high schools and colleges would be better positioned to provide more robust, fact-based advice to students.

    5. Do these findings justify substantial changes in the entire college application process?

    Interesting question. What do you have in mind?

    1. Re: “The new paper suggests that women who attend elite universities may set a higher bar for who they would be willing to marry, causing them in some cases to not be married at higher rates than other college graduates their age. If true, is that the result of the college they attended, or is that more a reflection of the individuals involved?”

      Yet these same women do double their income if they marry? That implies, a marriage which overcomes that bar somehow benefits the woman extraordinarily well. I’ve met a few such women over the years and, anecdotally at least, I’ll venture to say many of the ones who didn’t marry and whose salaries didn’t double, deliberately chose to commit passionately to a lower-paying public-interest or academic career, even to the extent of limiting social life. For those who did marry, their connections through their spouses to that ‘elite’ world they had entered might have made a big difference in the job they landed. And, your point about divorcees with single-parent responsibilities included in the “unmarried” category is well-taken. An ‘elite’ college is an entry ticket of sorts to a different lifestyle, but not necessarily to a remunerative lifestyle.

      So is it “worth” the cost to go to an elite college? Our young people are voting with their feet.

Leave a Reply