The admissions office vs. standards

The best way for a college to improve its admissions process would be to abolish the admissions office. A simple formula involving high school grades and SAT scores would be best. If many applicants meet the minimum standards for admission, then a lottery can be used to select those to whom to offer admission.

Admissions offices have always worked to undermine standards. In the summer of 1974, when I was an undergraduate at Swarthmore, Professor Ooms had me work on a project for him to study the undergraduate admissions process. He was disappointed at the caliber of the student body at the time, and he wanted to see what was going on.

We used statistics to uncover the factors that determined admission and the factors that were correlated with getting a good rating on the applicant interview. At the time, every applicant was interviewed, either by the admissions office or by an alumnus.

We found that the interview rating was a very important determinant of admission. We also found that interviewers did not like applicants with spectacularly high SAT scores. The SAT influenced admissions in two ways. Controlling for the interview score, a higher SAT score increased the chances of admission. But when the effect of the SAT on the interview rating was included, the highest SAT scores decreased the chances for admission. If Swarthmore had abolished its admissions office, it would have admitted many more students with higher scores.

Once you have an admissions office, the last thing they will do is try to maintain the role of standards for admission. Their power goes up by including more factors, especially subjective factors. The admissions officers can help promote legacies, athletes, students who are well-rounded (i.e., not Asian), and so on.

Incidentally, I got into Swarthmore in spite of my high SAT scores, because I interviewed with a local alum whose son I had watched wrestle in the state championship. I talked about that match during the interview. When I arrived at Swarthmore, the Dean of Admissions said that the wrestling coach was looking forward to having me on the team. I never went near the coach. After all, I had never won a single wrestling match in high school.

39 thoughts on “The admissions office vs. standards

  1. I assume you’re correlating high school grades and SAT score with college GPA. But is your GPA in college the best measure of success as a student? What about;
    1. Overall impact on your field post graduation.
    2. Overall success post-graduation (measured in some way)
    3. Starting or running student organizations that contribute to your school.
    4. Amount of alumni giving you provide for the school post-graduation.
    5. Behavior as a student (aka. Propensity to not do dumb stuff that embarrass the school).
    6. How much you improve the learning environment for students around you by, for example, asking great questions in your classrooms that spark discussion.
    7. Students who improve the school’s reputation with the local community, eg. By volunteering at local community nonprofits or contributing to local politics or such.

    Some of those are easier to measure than others, but I’d imagine they’re all important – not just your GPA. I’d want a college admissions office that seeks to maximize all of these factors, not just your gpa. What factors go into selecting students to maximize these? Why should they be irrelevant?

    • Max Marty,

      Your (1) and (2) require lagged metrics because a person’s career performance or general success usually ripen over many years. This is why most rankings (e.g., USNWR) measure outcomes in terms of retention and timely graduation rates — clear, present or short-term performance.

      Your (3) sounds good, but the reality at selective colleges is there are too many student organizations, mostly with few members and fewer real initiatives.

      Your (4) is indeed a rationale for legacy admissions, if the family has contributed or clearly will contribute major resources that improve opportunities for other students. The resource effect must be compared with a potential corruption effect. BTW, many so-called legacy applicants have excellent academic stats for admission. One must distinguish the subset of legacy applicants who have sub-par stats.

      Your (5) indeed should flag for special scrutiny applicants who have serious disciplinary (or criminal) records. But it should not extend to character assessment of every applicant.

      Your (6) and (7) give high-school students too much credit. ‘Expectations creep’ is how we end up with admissions expectations that applicants should have helped to solve a persistent problem of poverty, conflict, or public health in a remote, underdeveloped country.

      • I should clarify two points.

        Lagged metrics (e.g., grads’ career success) might constitute evidence of a particular college’s selection effects (admissions) and treatment effects (career prep) before the lag. But one shouldn’t infer that a particular college today will have the same career impact that it had, say, 20 years ago.

        ‘Expectations creep’ about impact on community is an invitation for high-school students to posture, because real impact on community outside of school is too much to expect from almost all adolescents.

      • On the contrary, we need more character assessment and should consider using interviews and/or questionnaires for the purpose — mainly to weed out views such as Critical Race Theory which harm a college and its students and reduce the value of its degrees. In fact, this filtering should be done when a school hires faculty and staff — especially those with powers over hiring/firing and admission decisions — as well as its student body.

      • Love your comments about expectations creep. Strangely, some people might argue that students with low grades and test scores might be “late bloomers” and might really mature intellectually in college. No doubt *some* such academic late bloomers exist. I’m guessing, though, that it’s far more common for “book nerds” to late bloom in terms of “leadership qualities” as they enter full maturity and adulthood than it is for those with low grades and test scores as teenagers to become Einsteins as adults. Just look at all the former book nerds leading tech companies. People seem to change/grow a lot more emotionally as adults than they do academically.

        If a teenager stays out of trouble, then they’ve passed the character test, regardless of how much they have (supposedly) contributed to world peace. Presumably, Ivy League colleges should evaluate character differently from the Miss America contest.

    • My impression is that a few colleges, such as Harvard (endowment $42 billion), have more or less figured out which applicants are most likely to write big checks to the endowment development drive in 30 years, which is why Harvard admits so many minor sport jocks. My guess is that positive factors for donations include:

      – athlete
      – male
      – legacy
      – white
      – fraternity
      – conservative
      – money-oriented

      In other words, the Bad Guys.

      But I’ve never seen anything published on this central question of which 18-year-old applicant is likely to become a middle-aged donor.

  2. Re: Admissions lottery. See my letter in the NYTimes, almost twenty-one years ago:

    https://www.nytimes.com/2000/04/08/opinion/l-admission-charades-671452.html

    “Opinion
    Admission Charades
    April 8, 2000

    To the Editor:

    The admissions process at selective colleges (front page, April 3) invites posturing by applicants, involves much idiosyncratic bias and costs plenty in personnel and resources. Young people should not be asked to sing their own praises. Administrators should not pretend to judge character from a file.

    Colleges should abandon the chimera of fine-tuning the composition of the student body. After all, most applicants are admitted to more than one school, and many choose to go elsewhere. It would be more honest and efficient to set a threshold of academic competence and then use a lottery to select students from the qualified pool. Savings should be passed on to parents and students.

    JOHN ALCORN

    Hartford, April 3, 2000”

    I broached the topic when I served as chairperson of the elected Faculty advisory committee for admissions at a selective college. Zero interest.

  3. “.. and that in every political institution, a power to advance the public happiness involves a discretion which may be misapplied and abused. [Americans] will see, therefore, that in all cases where power is to be conferred, the point first to be decided is, whether such a power be necessary to the public good; as the next will be, in case of an affirmative decision, to guard as effectually as possible against a perversion of the power to the public detriment.” – Federalist 41, Madison, January 1788.

    It’s easy to see how that applies to the problem of college admissions.

    I think after 233 years enough we can finally do a post-mortem on the Federalist Papers. I’d love to see that project, heck, love to work on it. They are crammed full of wisdom and good intentions, but at this point about 95% of each defense of some Constitutional provision needs it’s own ‘commission’. “What they were hoping for, what actually happened, what went wrong, what it would have taken to achieve the original goal and avoid that undesired outcome.”

    • I was reading a book by historian Bernard Bailyn about the Constitutional Convention. He had a triumphal section about how the Federalist Papers crushed the (anonymous) Antifederalist Papers. But he didn’t seem to notice that every single criticism of the Constitution enunciated by Antifederalists eventually came true by the 1960s.

  4. I like your admission criteria. I’d add:
    Auction off the last 1% of the places to the highest bidder using an independent service so profs and classmates wouldn’t know who bought their place.

    • Immigration is admission to the nation state. I see a few choices:

      – selective admission by merit score. Canada has “point-based” immigration policy that is merit based. Trump proposed merit based immigration as well, which wasn’t received well.
      – selective admission by benefit to incumbent members. This is basically the selective university model and the traditional sovereign nation state model falling out of favor.
      – open admission or open borders like community college.
      – Charge a selectively high price. This is the opposite of financial aid. If you are wealthy, many nations will basically sell you the right to immigrate. Similarly, most selective colleges will offer admission for the right price, often this is masked as a “donation” which is really just fee for service with a tax deduction.

      • It’s not necessary to limit, say, the number of grocery stores in a region through licensing, let alone allocate those licenses through high school grades and SAT scores. When the old Soviet Union collapsed and the former Soviets tried to learn about capitalism, one of the things they had trouble grasping was that nobody was in charge of deciding how many grocery stores there would be. We have Open Grocery Stores, but that doesn’t mean unlimited grocery stores, just market-limited grocery stores.

        One of the options that Niko Davor left out was selective admission by markets, especially labor and housing markets, rather than government “merit score”. Market immigration is what we currently call Open Borders.

  5. The real problem with admissions is the supply of spots at prestigious universities. The importance of the admissions process is a byproduct of this fundamental problem, but not the problem itself. Swarthmore, or Harvard, or Berkeley, etc. simply haven’t grown with the population of applicants who want to go there.

    The population of the US in 1974 was around 220M versus 330M today. And that’s not including the massive numbers of international students. Yet Swarthmore is, what, 100-200 students bigger than it was 50 years ago?

    Ideally, colleges would have spots for everyone whom the college believed would succeed at that college. At least in my book that is how it would work.

    Perhaps this would shift the power to determine the character of the school from the admissions dean to the students themselves. I wonder how much difference that would make as a practical matter, though, given that specific school brands will attract certain types of people.

    Incidentally, when I was at Swarthmore in the early 1990s, the Dean of Admissions was a wonderful man who, in my senior year, saw me standing in the lobby of the music hall before a choir concert, and asked “why aren’t you out there?” He’d remembered from the admissions process that I sang in my high school choir. His recollection impressed me, and at least showed that he was taking the time to get to know the students he admitted to the college. I also had good SAT scores and a terrifically fun interview, for what it’s worth.

    • Heh, the thing you liked about that Dean would be impossible for anyone else if they tried to scale up.

      Size and Scale are the enemies of personalisation, and just create distance, insulation, abstraction, and alienation.

      This college size argument is popular and comes up a lot, but it doesn’t make any sense. Total supply is facilities x output per facility. Why focus on the latter as the preferred margin for adjustment, especially when the markets for physically-provided outputs clearly favor the former. As population goes up no one complains that the average size of any particular business establishment doesn’t rise because the answer to remaining proportional to demand is that there is are just more of them. Obviously that was once true for colleges in America too, once few, now many.

      The only way to salvage the argument is to impose some extra constraint limiting the number of elite facilities, and then to get into a perverse hybrid of the current system on the one hand and the implications of Caplan’s The Case Against Education on the other.

      That is to say that there is no way to fix an unfortunate feature of our social psychology which is to rank institutions by prestige, to accept as social currency the signal of that prestige being proportional to merit for all those associated with it, and that in the nature of things there is only so much room for institutional names on the A-list. (US News might agree).

      The limited A-list means that you can only have so many prestigious institutions in the relevant extent of the market, and that in order to prevent the unfair case of a meritorious individual being unable to obtain a good signal, the total output of those few institutions must grow in proportion to the total amount of merit in the general population, like what happens to the size of a district constituency when the number of Representatives is fixed.

      Lots of problems with that. For one, as I said about scale, things change with size and many valuable features are lost. The military and some corporations are only able to deal with this by means of hierarchy and modularity and with lots of circulation, but you can’t franchise a distinct community with a bunch of cookie cutter clones, especially since perceived coolness and prestige will inevitably diverge which is a self-reinforcing process.

      But mostly it surrenders to and thus avoids the root problem and settles for a merely inconsequential surface-level polishing.

    • The real problem with admissions is the supply of spots at prestigious universities.

      The real problem with admissions is the supply of spots at prestigious universities.

      The real problem is the low achieving students at high prestige schools get higher status and more opportunities than the high achieving students at low prestige schools. Which facility a person attends should matter less, and what they actually accomplish and achieve should matter more.

      • Yes, but one of the issues is the restricted range of what can be accomplished and achieved at universities – elite, average, or poor. So that standard doesn’t really get us anywhere…assuming we want to move past the status quo.

        • It would be helpful if you and Arnold articulated what exactly you’re looking to accomplish by changing the admissions criteria vs. what we have today.

          At the margin, are the elite colleges a sorting mechanism or a transforming mechanism? If the former, then none of this really matters all that much other than making it slightly more difficult for employers to find the best candidates.

          E.g. I know a lot of senior VP types that graduated from state schools (with very unselective acceptance criteria) that have outperformed those from much more elite schools…and the job market is able to identify and compensate them accordingly with the appropriate premium.

          http://www.econtalk.org/kling-on-education-and-the-internet/

          • I’m perplexed by your comment. I have not articulated a need to change the admissions criteria. Regardless, to answer your question, I suspect that colleges – elite and ordinary – serve many functions, sorting and transforming among them. My primary concern is goodness of fit. And I think our current model (college = only path to success) does harm to many.

  6. I think ASK is oversimplifying admissions.

    At all but a handful of colleges, the admissions office must consider the student’s financial situation. Granted, that doesn’t need an admissions office.

    Most colleges seek diverse admissions along multiple dimensions: gender, race maybe controversial, but diversity along religious or other cultural factors, geographical region, and academic interest and aptitude are crucial. (Diversity is a good in and of itself for student life, IMO; it is fair to question its cost.)

    There is a legend that a certain midwest college a couple of decades ago was a victim of the excellent reputation of it’s writing program. Far too many admitted students, and too many of the best, were dedicated to studying writing fiction, overcrowding that area (and making it over competitive) and starving the rest of the departments. Prospective students in other areas wrote it off as a writer’s college, fairly or not. That can hurt the ability to recruit faculty to those other departments. (And also your most talented alumni will make no money!) In short, a vicious circle. The admissions office had a big role in digging them out of this hole. According to legend.

  7. Community college is college without selective admissions.

    Admission lotteries aren’t needed. Individual course sections have enrollment caps. They scale the number of sections offered based on enrollment. They scale physical capacity such as classroom, campuses, parking, lab facilities based on forecasted demand.

    What does Kling and this crowd feel is lacking with community colleges? What would you suggest for improvement? Can community colleges grow to compete with prestigious colleges?

    Regarding difficulty and aptitude: A calculus class at community college is similar to that offered at a selective prestigious university. They use the same textbooks, they cover similar chapters, and I suspect grading is similar. The lack of admissions requirements or SAT scores isn’t a problem. Also, the math component of the SAT is very basic. A good community college calculus class is far more advanced than the SAT.

    • “What does Kling and this crowd feel is lacking with community colleges?”

      +1 hint: most of them have never attended a course at a community college so they have zero credibility on this topic. They would somehow prefer to be taught by an amateur grad student who is moonlighting as a wannabe college professor vs. the real deal at the community college level.

      • “When I went to college, there was no substitute for a good professor.” – ASK 12/20/20

        My experience: some of the best professors that I had in the 90s were at the community college level. The professors at the top tier university that I transferred to could not compete, at least not for the intro level courses.

        • I suspect this is different in non-STEM fields, but in STEM classes, I learn from the textbook, not the professor. I often prefer taking a class led by an amateur grad student; you say this like it is a bad thing. If I’m taking a textbook centric class, I want it run by someone who is humble, low ego, and is serious about administering the class well. Grad students are often great at this.

          • Would you rather have sex with a virgin or someone with more experience?

            I’m going with the latter. All of my cc professors were both great teachers and highly accessible/humble. Included both stem and non-stem.

  8. I agree that there are a lot of problems with the admissions process. And yet, I think it’s safe to say that the admissions office isn’t going to be abolished anytime soon. If I can be so bold as to bring “the possible” into this discussion, I’m curious about Arnold’s suggestions for realistic reforms to admissions. Arnold’s interview experience is an interesting example and potential starting point. Should we put more weight on interviews? Expand interviews to larger, less elite schools? Or perhaps we should do the opposite?

    • One possibility is not to get rid of it, but to have lots more of them: to radically decentralize and de-institutionalize admissions, and to grant selectors legal and institutional unqualified immunity for their decisions.

      Professors might try to pick their own students in the style of a master’s atelier studio, which is not too different from how some of the prestigious academic labs or research groups operate. “I studied X under Y” instead of “I have a degree from Z.” If Y is famous, no one cares under what institutional canopy the work was done.

      This can’t work at scale because there are two many Ys and their names don’t work as signals for the general audience.

      But one sees it used within institutions to separate the merely elite from the truly elite, on the basis of how much of an insider they are and their proximity to centers of power in the social network.

      A merely elite person will use non-relationship factors to signal, such as degree, rank, position, job, budget, division, etc. They are like special operations – they do cooler kinds of work in cooler kinds of branches.

      A truly elite person will dispense with all that, even consider it kind of gauche in the typical impulse to intensify the stratification of class fashions and norms. They will describe themselves as “I’m on Bill’s project”, or “I work for Claire”, and even will be described by others instinctively as “Oh, he’s one of Jack’s people.”

      Since we are naturally fascinated with the truly elite, this is exactly what the entertainment industry gives us with soap-opera-like narratives and feeling like being let in on the extremely personalized dramas of interacting and rivalrous true-elites.

      • Federalism for admissions. This could be promising – of course, it’s all in the marketing. I get a bit concerned about hyper-hyper-localization, as this might magnify some of the long-standing problems of the “members only” style of admissions.

        You contrast the “merely elite” and the “truly elite” and suggest that “we” are more fascinated by the later. Perhaps our definitions of these classes are different, but I suspect “we” are more fascinated by the former – what I might call the “near elite.” How do you define “merely” and “true” in this context? Can you give some representative examples?

    • Abolishing admissions at prestigious schools is not realistic. Expanding community colleges that already have open admission policies in place, is more than realistic, that is something that has happened in recent decades. I’d like to hear what are the shortcomings of community college and what can be done to make them more competitive with what prestigious schools are offering.

      • I agree with your affinity for traditional community colleges. One concern I have with their recent trajectory is the de-emphasis on trades etc and the increased emphasis on being feeders for universities. This trend continues to privilege the university system as the one true way to succeed and, thus, does harm to those who are not well served by it.

  9. Australia admits people to University based on their ‘Australian Tertiary Admission Rank’ or ATAR. It works well generally.

    There is also undergraduate entry into law and medicine.

    Doesn’t Caltech do this as well?

    • Caltech has bucked many admissions trends (race as a factor, financial need as a factor, legacy admissions), but the application still has an essay section and most applicants are interviewed by a professor or alum. So no, there is no threshold SAT score that will guarantee your admission to Caltech.

  10. Joanne Jacobs points out today how tests open doors for non-conformists:

    “Good grades go to the students with their hands up, the kids who do extra-credit and strive to please their teachers, Henderson writes. Intelligence and aptitude tests can find talented students who’d otherwise be overlooked.

    ‘Typically, the identification of gifted students has relied on the referrals of parents and teachers—a subjective personal evaluation. But a 2016 study published by the economists David Card and Laura Giuliano found that implementing a standardized testing requirement increased the number of poor and minority students. . . . Likewise, a British study found that when comparing students who earned the same results on a cognitive test, teachers judged poorer students as less capable.’

    Some see testing as a barrier to social mobility, writes Henderson. But the alternatives to testing are subjective and easily gamed.

    Educated, affluent parents hire tutors to boost their children’s grades. They pay advisors to “polish” college application essays, craft the most impressive extracurriculars and community service trips, get important people to write recommendation letters and market the product to elite colleges.”

    https://www.joannejacobs.com/2021/02/aptitude-tests-open-doors-for-non-conformists/

    Axing the admissions offices would prevent a lot of waste. Traditional “vestibular” college admissions in Brazil are a great example of how student bodies can be improved by going with objective measures. Per wiki:

    “In order to enter university in Brazil, candidates must undergo a public open examination called “Vestibular”, which lasts about 1 week and takes place once a year. Some universities may run Vestibular twice a year, for two yearly intakes instead of only one. This option is popular with private universities, while public universities usually run Vestibular only once every year (in November, December or January). Universities offer a limited number of places, and the best ranked candidates according to their overall Vestibular grade are selected for admission. Although the Vestibular format changes from university to university, it typically consists of a week-long examination on compulsory high school subjects such as Mathematics, Physics, Chemistry, Biology, History, Geography, Portuguese language and literature, and a foreign language (usually English). Private universities usually “condense” this week-long examinations into a couple of days, but some public universities still require a week-long marathon.”

    In the USA this week of testing could incorporate College Level Examination Program (CLEP) or equivalent tests so that college credit would be awarded for appropriate levels of achievement and the ridiculously exploitative four year college plan reduced to a three year program as is common in free countries in Europe and elsewhere without undoing the 120 credit hour tradition.

    • And Jacobs has more today:

      “At high-poverty high schools, 11th- and 12th-graders are taking and passing online college courses offered by Harvard, Yale, Cornell, Howard, Arizona State and the University of Connecticut, reports Erica L. Green in the New York Times.

      The National Education Equity Lab, a New York-based nonprofit, enrolled more than 300 students nationwide in a Harvard course, “Poetry in America: The City From Whitman to Hip-Hop,” in 2019.

      The college-in-high-school model includes “college advisers, mentors and high school teachers who help teach the material,” writes Green. It’s free to students.

      ‘The high schoolers met the same rigorous standards of the course created for Harvard’s admitted students — they listened to lectures, took quizzes and completed essays, and they were graded by the same standards

      Of the students who completed the course in fall 2019 — 92 percent of whom were students of color, 84 percent of whom qualified for free lunch — 89 percent passed, earning four credits from Harvard Extension School that are widely accepted by other colleges.’

      Sixty-three percent earned an A or a B.

      According to early results, 86 percent of students in the pilot have passed courses, earning college credits and building their confidence, writes Green. “This semester, the Equity Lab has grown to serve about 1,500 students from 75 of the nation’s poorest schools in 35 cities. Several school districts and universities are vying to join the consortium, which has a goal of expanding to serve 10,000 students by 2022.”

      https://www.joannejacobs.com/2021/02/meeting-the-elite-college-challenge/

  11. As long as academic standards in the first couple of years of college are rigorous and the required courses are demanding, no other factor predicts success as powerfully as does performance on the SAT…..full stop. Nobody at a serious institution with serious core courses – such as engineering schools – will deny this with a straight face. If students aren’t required to take demanding courses in math and science, if grading is gentle, and/or if a school has unserious disciplines where poor students can take refuge from demanding courses and assessment, then who needs the SAT? It isn’t telling you anything you really care to know, and it weeds out too many of the candidates who would otherwise receive a warm welcome.

  12. At Caltech, the tradition for decades had been for at least one undergraduate to read each application. The students were the fiercest to make sure that the admitted students were well qualified since they had seen the consequences of many students either flunking out or having to transfer unhappily after a year or two (not a small risk as up to a third of students in the 70s and 80s didn’t graduate in four years). However, because the students cared about standards, in recent years the admissions office has pushed out the undergrads from the admissions process. And lo and behold, the evidence is that the century long tradition of Tech’s promoting academic qualifications first above all else is slowly (perhaps rapidly) being watered down. Some on the faculty openly want to emulate MIT’s “holistic” admissions.

    • The students were the fiercest to make sure that the admitted students were well qualified since they had seen the consequences of many students either flunking out or having to transfer unhappily after a year or two (not a small risk as up to a third of students in the 70s and 80s didn’t graduate in four years).

      Having observed similar consequences in my small neck of the woods nano-Caltech – at most 1/3 of my class finished the course, the rest having transferred – I think this is a very sensible attitude. If you know of any good write-ups on this Caltech tradition, could you give them here for future reference?

Comments are closed.