Academia does discriminate against white men

Why Jacob Savage is right.

Written by Noah Carl.

Jacob Savage has penned a brilliant piece about the discrimination faced by young white men in certain US industries, namely academia, journalism and creative arts. The article deservedly went viral. However, one weakness is that Savage left himself open to the charge of having cherry picked some of his figures.

He writes, “Since 2022, Brown has hired forty-five tenure track professors in the humanities and social sciences. Just three were white American men.” And he cites similar numbers for a few other schools. Which seems to suggest that universities are discriminating against white men. However, you can’t infer that much from a handful of schools in a country with literally hundreds. Perhaps some universities discriminate against white men and others discriminate in favour, and Savage just happened to pick ones that discriminate against. (I don’t seriously believe this. I’m just imagining what a critic might say.)

As a matter of fact, Matt Bruenig ran the numbers and found that the percentage of white men employed in “arts, design, entertainment, sports and media” has remained relatively stable over time. He also observed barely any change in the percentage of white men aged 30–39 in the top 10% of the income distribution. Though as some commenters noted, Savage’s argument doesn’t necessarily predict that white men’s income should have declined because the alleged discrimination is happening in jobs that confer prestige and influence without offering particularly high pay.

Is Savage wrong? While he might have slightly overstated his case, I’m pretty sure he’s onto something. To begin with, some of his figures clearly weren’t cherry picked.

For example, he points out that the Disney Writing Program “has awarded 107 writing fellowships and 17 directing fellowships over the past decade—none to white men”. Zero out of 124 seems unlikely to be a fluke.¹ And while there might be hundreds of universities in the US, how many companies like Disney are there? Five?

What is more, we have experiments showing that when you hold applicants’ research quality and teaching experience constant, American academics prefer to hire women and non-whites. And not just a little bit — they strongly prefer applicants who aren’t white men. Yes, these preferences were elicited in hypothetical scenarios, but it would be surprising if they didn’t have some impact on actual behaviour.

To investigate, I examined the distribution of assistant professors by gender and race in 1991 and 2023 — the earliest and latest years with available data. Note that assistant professor is the most junior tenure-track position in US academia. Most such faculty are in their late 30s.

For each of the two years, I compared the observed distribution to an expected distribution based on population size and average IQ, assuming that academic faculty are selected from among those with IQs greater than 110.² Data on the number of people in each group aged 35–44 in the preceding year were taken from the US Census.³ And I used Richard Lynn’s estimates for the average IQ in each racial group. However, I assumed that men score 2 points higher than women and that men’s standard deviation is 4% higher. (As discussed later, these are likely underestimates of the true sex differences.)

Results for 1991 are shown below. White males are neither over nor underrepresented, whereas black males, black females and Asian males are all substantially overrepresented (in percentage terms). By contrast, white females, Hispanic females and Asian females are all slightly underrepresented.

Author’s analysis.

While lingering discrimination against women might account for some of their underrepresentation, the more important factor is surely that fewer women at the time pursued employment in general. For example, the male labour force participation rate was still about 19 points higher. You can chalk this up to sexist norms in the past, but that’s different from blaming academia itself. The overrepresentation of blacks is presumably down to affirmative action.

Results for 2023 are shown below. Now white males are underrepresented — to the tune of more than 25%. Hispanic males are also underrepresented. Meanwhile, white females and Asian females are overrepresented, along with black males, black females and Asian males.

Author’s analysis.

In fact, white males are the only group that saw their representation among assistant professors fall between 1991 and 2023. Every other group saw at least a small increase, with white, black and Asian females seeing the largest rises. The continued underrepresentation of Hispanic males may be partly due to their outward similarity to white males.

What about my assumptions? You can quibble with Lynn’s estimates, but they must be approximately right. As for my assumptions about sex differences in IQ, they are probably conservative. The most comprehensive study put the male standard deviation at 6% higher. And some evidence suggests that men score 3–4 points higher than women. In addition, I ignored the fact that men score higher on other traits conducive to scholarly achievement, such as systemising and general knowledge (which helps to explain why they make up 85% of Wikipedia editors).

The findings clearly support Savage’s argument. And even if you don’t buy my analysis, just consider this simple fact: in 2023, the absolute number of white female assistant professors was almost 30% higher than the absolute number of white male assistant professors. Discounting psychological sex differences, you’d at least expect parity. But no, there were 30% more white females.

America is still reasonably meritocratic.⁴ Recent discrimination against white men doesn’t show up in the overall labour market statistics. Yet in certain industries — notably those dominated by ideological leftists — they do face tangible barriers to advancement. “White male privilege” isn’t merely wrong; it’s the inverse of what’s true.

Noah Carl is an Editor of Aporia.

---

Support Aporia with a paid subscription:

You can also follow us on Twitter.

1. If white men are 25% of the working-age population, the probability of picking no white men in 124 trials is approximately zero.

2. I had AI project the population size of each group based on the assumption of normally distributed IQ scores. Three different models were tried and they all gave the same results.

3. The preceding year was used because the only usable figures I could find for the earlier time point were for 1990, rather than 1991, so to keep the comparison constant I used figures for 2022, rather than 2023, for the later time point. This is unlikely to have had any impact on the results.

4. The big exception is affirmative action for blacks, as Charles Murray and others have documented.