The Google memo is poorly argued and reaches the wrong conclusion

Photo: TPG/Getty Images

Imagine there were a professional field where, due to aggregate differences in aptitude and interest across genders, 35% of the good job candidates are women, and that you would expect an industry environment that truly treated men and women equally to produce an employee base in this field that is 35% female.

In the real world, employment in this field is likely to end up being a lot less than 35% female. Consider the following reasons:

  • A widespread assumption that “most” of the good job candidates will be men may lead to stereotyping in the hiring process, with hiring managers more likely to assume that men are good candidates and overlook qualified women.
  • Women may self-select out of the field because they internalize the stereotype that the field is “for men”; the stereotype may also make men overconfident in their fitness for the field and more inclined to pursue employment in it.
  • A male majority in the field is likely to be excessively self-reinforcing, as research shows hiring managers tend to use the qualitative and “culture fit” aspects of hiring to hire candidates who resemble themselves, and most of the hiring managers in a male-dominated field will be men.
  • As seen in a number of high-profile cases in Silicon Valley, male-dominated management structures may foster cultures of pervasive workplace sexism and harassment that drive women out of the field.

And this is a key problem with the now-notorious Google memo written by a now-former employee: If it is true that aggregate population differences mean that a majority of the suitable candidates in a field are men, that can make it more important for firms in the field to undertake aggressive diversity efforts to recruit and retain women. Otherwise, firms may end up with a employee base of which only a small minority is women, even when women make up a large minority of the suitable candidates.

The memo misses this entirely, jumping from a claim that gender differences in interests and aptitude “may in part explain” the strong male skew in Google’s engineering groups to a conclusion that specific efforts at Google to recruit and retain women and underrepresented minority candidates are counterproductive and should be ended.

For example, the author complains about “hiring practices which can effectively lower the bar for ‘diversity’ candidates by decreasing the false negative rate.” That is, he’s upset that women candidates get a second look when men don’t.

But this is something you would absolutely want to do in order to prevent a phenomenon described above: hiring manager biases and stereotypes leading to a lopsidedness by gender in hiring that exceeds the actual lopsidedness by gender in the qualified candidate pool. It makes sense to be extra certain that women who got screened out were rejected on the basis of qualifications and aptitude, not something else.

He also objects to Google’s various programs to mentor and develop women and members of various underrepresented minorities, an approach that he calls “unfair and divisive.” But these programs may be useful to counteract another phenomenon I describe above: qualified people in certain demographic groups self-selecting out of the field because they believe they are at a disadvantage in it.

Defenders of the memo say the firing of its author proves the memo’s point about political correctness: That certain factual questions are not allowed to be discussed in liberal-leaning organisations. Whether there are inherent, aggregate differences across gender that one would expect to lead to unequal representation in a given field is an empirical question, and it’s true that some people want to close off investigation of it.

But a reason proponents of gender equality are reluctant to discuss the question is that it’s so often raised in the way it is in this memo: As a pretext to justify whatever gender gaps exist in a field, regardless of whether the magnitude of those gaps can be explained by such differences; and to dismiss efforts to promote diversity, even when such efforts would remain justified — or even become more justified — if aggregate differences across genders are real.

The claim at the top of the memo — “discrimination to reach equal representation is unfair, divisive, and bad for business” — is a straw man. Google’s workforce in tech jobs is currently just 20% female, and so diversity efforts are obviously not coming anywhere close to imposing equal representation.

Even if you think there is a true gender skew toward men in the pool of qualified candidates for technical jobs at firms like Google, which seems more likely to be more true to you: That political correctness has led Google to choose too many female candidates for technological jobs over more-talented men? Or that a combination of the factors I describe at the top of this post has created an excessive male skew in technological employment at Google and other tech firms, even if perfect personnel practices would not eliminate the skew entirely?

The latter seems much more likely to me.

Get the latest Google stock price here.