Skip to content

Does Stanford under-enroll white students?

Over at National Review, Nate Hochman is annoyed at Stanford University:

The demographic profile of Stanford University’s class of 2026 is out, with 1,736 matriculated students in the freshman class of one of the world’s most prestigious universities. But as some perceptive critics were quick to notice, one key demographic is disproportionately underrepresented: While whites make up more than 50 percent of the nation’s adolescent population, per 2019 Office of Population Affairs numbers, they were only 22 percent of Stanford’s class of 2026.

I was curious about this because there's something that I bet most people are unaware of: Stanford draws nearly half its students from California, which has a very different demographic profile than the rest of the country. Here it is for public school students:

The white population is probably undercounted because these numbers are only for public schools, but not by much. Private schools make up less than a tenth of all California schools.

Stanford draws 40% of its students from California and 60% from other states. It's simple to calculate a weighted average of their applicant pool, which looks about like this:

White kids are slightly more than a third of the total pool, not half.

But this isn't enough: Stanford is an elite university that draws from about the top 5% of high school students. By chance, the "advanced" level on the NAEP test also represents approximately the top 5% of students, so this is the pool that Stanford draws from. Here's what it looks like:

Now we need to apply this to Stanford's pool of applicants. For example, white kids make up 36% of Stanford's raw pool of applicants, but only 6.5% of those are performing at an advanced level. That's 2.3% of the total raw pool. Here's the number for all groups:

  • Hispanic: 0.78%
  • White: 2.35%
  • Asian: 1.03%
  • Black: 0.08%

This adds up to 4.23%, which means, for example, that Hispanic kids make up 18% of the entire pool (0.0078 ÷ 0.0423). Here's how this looks for all groups compared to Stanford's actual enrollment for the class of 2026:¹

This is all fairly crude, but it's in the ballpark. And after all that work we're back where we started: white kids make up more than half of Stanford's total pool of applicants but only 22% of their enrollment. Hochman was right.

Oddly enough, this analysis makes it look like Stanford enrolls a disproportionate number of Asian students. If I had to guess, I'd say it's because all the other groups do poorly in math,² and Stanford might put more emphasis on math ability because of its high enrollment of STEM and computer science students. But that's just a guess.

¹Note that my percentages for Stanford enrollment don't match theirs. This is because I discarded two groups that didn't have racial IDs (international and multiple-race) and then recalculated the percentages.

²It's not because Asian kids do well in math and poorly in reading. In fact, they do as well in reading (14% advanced) as in math (15% advanced).

NOTE: California public school enrollment is here. National public school enrollment is here. NAEP scores can be extracted here. Stanford enrollment is here.

20 thoughts on “Does Stanford under-enroll white students?

  1. middleoftheroaddem

    Respectfully, if you are examining Stanford ,then the 12th grade scores you show is a pretty limited metric. Rather, if you can find it (College Board used to disclose this figure but now I cant find it) look at the demographics of people in the 98% - plus percentile of the ACT and or SAT: that is much closer to your non athlete applicant pool for Stanford.

    For example, in Kevin's analysis he suggests that 22% of Stanford's applicants are Hispanic: it would be interesting to see how many Hispanics are in the 98% plus percentile in terms of testing....

    I found this

    In 2005, 153,132 African Americans took the SAT test . . . but only 1,132 African-American college-bound students scored 700 or above on the math SAT and only 1,205 scored at least 700 on the verbal SAT. Thus, in this top-scoring category of all SAT test takers, blacks made up only 1.1 percent of the students scoring 700 or higher on the math test and only 1.5 percent of the students scoring 700 or higher on the verbal SAT

  2. Austin

    I didn't know the National Review was now in favor of racial quotas: sounds like they want one for white people, now that Asians are outperforming everyone on "merit-based" metrics. Everybody hates quotas, until the group they belong to struggles to make the cut off.

    1. MF

      They did not advocate for quotas.

      But they are asking... for those who say that the enrollment at top schools should look like America, why isn't this an issue?

  3. Perry

    Why no concern about gender? Most universities try to balance the number of male and female students. Recruiting female students is typically not a problem since their grades and scores are higher than for male applicants, but if they are emphasizing STEM that may partially explain why there may be more minority students. Female students who are good at STEM disciplines may be more likely to be Asian or other, than white female students. I don't know whether Stanford does emphasize STEM though.

  4. sonofthereturnofaptidude

    I thought conservatives hated elite universities because they're elitist. Now they have to hate them because conservatives think they don't include enough white people?
    Whatever. You can be sure the folks at NR will find a reason to hate whatever institution of higher learning for whatever reason they have at hand when it suits. That is, when they are not trying to get their kids and grandkids into those institutions.

  5. rick_jones

    The white population is probably undercounted because these numbers are only for public schools, but not by much. Private schools make up less than a tenth of all California schools.

    From a link I sent to Kevin on the occasion of his initial, aborted posting of this article the other day. A breakdown of the child population of California:

    Hispanic/Latino: 48.6%
    White: 29.8%
    Asian: 11.0%
    African American/Black: 5.4%
    Multiracial: 4.5%
    American Indian/Alaska Native: 0.4%
    Native Hawaiian/Pacific Islander: 0.3%

    which seems very much different for Hispanic/Latino and White compared to his percentages picked from the public school enrollment for the starting point of his math.

    https://www.kidsdata.org/topic/33/child-population-race/pie#fmt=144&loc=2&tf=141&ch=7,11,70,10,72,9,73&pdist=73

  6. ProgressOne

    Some commenters here are saying NR supports quotas for whites, so I read the NR piece to see if that is true. There is not even a hint of that. The writer simply opposes affirmative action, especially when it appears to discriminate against white applicants.

    1. D_Ohrk_E1

      The inherent implication of stating, "Stanford’s Class of 2026 Doesn’t ‘Look Like America'" is that there are unspoken quotas and those unspoken quotas match American demographics. Otherwise, one cannot state, "one key demographic is disproportionately underrepresented", because one would otherwise not have a metric to measure against.

      Every choice discriminates against someone else. If race-blindness ends up with a disproportionate over-representation of Whites, isn't that discrimination against those other races who are under-represented?

      Your ball.

      1. ProgressOne

        "If race-blindness ends up with a disproportionate over-representation of Whites, isn't that discrimination against those other races who are under-represented?"

        No. The goal should be a merit-based system, however imperfect that may be. Then just let the chips fall where they fall. Society is too complex for us to try to re-engineer it achieve race and gender balancing. If you really want to force race balancing, I guess we need to add a whole lot more whites to the NFL and NBA. We need to do a much better job limiting Asians going to college I guess. We need to limit the number of Jewish people getting PhDs. Gender too. We need to start limiting the number of women admitted to college. And so on.

        Giving organizations the goal of balancing racial and gender mixes according to their preferences goes against our civil rights laws and the spirit of those laws. Giving bureaucrats the power to pick who advances, based on their race or gender, is just a bad idea. If we really want that, we should re-write our civil rights laws to say it openly. Problem is, hardly any Americans would support that. Thus, the discrimination has to mostly be done quietly behind closed doors. The goal is to get around the laws.

    2. jdubs

      While it is true that many define 'quota' or 'affirmative action' as a structure/process that benefits non-whites and 'the natural state of things' as a structure/process that benefits whites (or another group at the top of the social strata)...it really distorts the discussion to use charged terms for only one group.

      Similar to the ongoing abortion issues....its an abortion when those people do it, its a medical procedure when we do it.

      1. ProgressOne

        "the natural state of things' as a structure/process that benefits whites"

        Based on that logic, the natural state of things benefits Jews and Asians the most since they do better in school and earn higher incomes than whites. College attendance is now 60% female and 40% male. So the natural state of things support women I guess. The most dangerous jobs are by far dominated by men. The NFL and NBA are dominated by blacks. Elementary school teachers and nursing are dominated by women.

        So you really want to re-engineer society to race and gender balance everywhere? Should we re-write our civil rights laws to openly allow this and even require it?

  7. interestedreader

    Something looks off in the final bar chart and the text following it. The text says: "white kids make up more than half of Stanford's total pool of applicants but only 22% of their enrollment" but the chart seems to show the enrollment figure as 29%. 22% was the figure in the NR article...

  8. MattBallAZ

    Gawd, white whiners claiming victimhood in this country just sickens me. And I say this as the father of a 99th percentile white kid who got rejected from Stanford. In favor of a less-qualified white legacy (from the same high school, so I know the kid's scores and grades and activities).

  9. zoniedude

    I don't follow this at all. Among top students of whatever race they tend to seek the top national universities. Stanford is one, but so is Harvard, Yale, Princeton, and Berkeley. So if there are five top schools then ipso facto Stanford should enroll about 20 percent of the top students. Toss in MIT, Caltech, Johns Hopkins, and other highly rated universities and the expected percentage is even lower.

  10. FishSauce

    What people should get upset about is that a school as rich as Stanford only enrolls 1,750 students per year. They have the resources to enroll 10x that many.

    1. Anandakos

      No, they don't. There aren't ten times as many Stanford-quality professors to teach the students. Quality is a thing, especially in education.

      1. MF

        Professors at a school like Stanford are chosen primarily based on their research, not their teaching skills.

        There is no reason to think that an advanced mathematics researcher will be particularly good at teaching calculus to 18 year olds and there is no evidence that skills, intelligence, or knowledge are contagious.

        It is an accident of history that we waste the time of our most advanced scientists by expecting them to teach college undergraduates. We should not do it and it is clearly not necessary.

        The true value of a school like Stanford is that when you are with a cohort at that level of intelligence all classes can move faster and dive deeper and you learn from and help your peers.

Comments are closed.