Ideas

Ideas | Jesse Singal

How the Internet got the ‘Google memo’ wrong

Francois Lenoir/Reuters, file

Suddenly last weekend, everyone on the Internet knew what the “Google memo” said — even people who hadn’t read it.

Social media exploded after the tech website Gizmodo published a document that had been widely circulating among Google staffers. Its author, a software engineer named James Damore, argues that Google is unfriendly to conservatives, that the company’s diversity initiatives are misguided, and that the company’s internal culture limits the open discussion of difficult issues. But the part of the memo that hit Twitter and Facebook like a bomb concerns biological differences between men and women. Damore argues that such differences, not just misogyny, could partially explain hiring disparities between men and women.

Amid a heated ongoing conversation about the underrepresentation of women in tech, the online reaction to Damore’s memo was fierce. But that reaction consisted largely of people responding to claims Damore himself never explicitly made. Many people on the left, including journalists covering the controversy, treated the memo as a hateful “screed” — that was the word Gizmodo used — with zero basis in research. On the right, commentators saw the memo as a straightforward compilation of obvious truths that only an easily-triggered liberal snowflake could deny.

Advertisement

During any Internet frenzy, the object being discussed stops bearing much resemblance to what it is in real life, and this frenzy was no exception. In actuality, the sex-differences part of the memo was based mostly on sound science. Damore didn’t make up or misrepresent basic research findings. He did, however, overextrapolate from them, sometimes wildly.

Still, even problematic writing should surely be judged by what it actually says — rather than by how other people characterize it on Twitter. Here’s a quick look of some of Damone’s most controversial and widely circulated claims:

Claim: “Women generally also have a stronger interest in people rather than things, relative to men.”

Social science research backs this up. In a 2010 paper, the sex researcher Richard Lippa used multiple cross-cultural studies, with a total of more than 700,000 subjects, to examine many of the reported personality differences between women and men — including the claim that men are, on average, more interested in abstract theories and systems, while women are more interested in understanding people and their feelings and relationships. The studies reported differences that Lippa described as “large” to “very large,” in contrast to some more moderate differences found with regard to personality differences.

These studies, it should be noted, can’t conclusively reveal how much of these differences stem from biology, as opposed to socialization and culture. But Lippa does believe there is an important biological component to these differences, because they were found across cultures. “This seemed inconsistent with social role theories, to me, and compatible with the notion that there was a biological component to sex differences on this dimension,” he said in an email. Of course, “an important biological component” doesn’t mean “entirely stemming from biology,” and as Lippa himself notes in his research, biology and culture often dance in very complicated ways.

Claim: The various reported differences between men and women are “exactly what we would predict from an evolutionary psychology perspective.”

In this case, Damore makes a sweeping claim that he doesn’t fully explain, much less defend. A basic tenet of evolutionary psychology — or “evo psych,” as it’s commonly called — is that differences between male and female behavior are shaped by natural selection. Men evolved to be more aggressive to protect their women and children, goes one common story, while women evolved to be more nurturing, better equipping them to care for children.

Advertisement

Many of the stories told by evo psych have attracted immense controversy; critics claim these stories are too broad and too vaguely defined to be scientifically tested. Making matters worse, journalists — and seedy Internet subcultures such as the pickup artist community — misrepresent and overstate evo-psych findings, often in ways that reinforce misogynistic ideas.

“Much of the thinking within evolutionary psychology, but especially with how it’s taken up in popular culture, is based on overly simplistic and out-dated evolutionary biology,” said Holly Dunsworth, a University of Rhode Island anthropologist. “This memo’s author seems to be arguing that these ideas are more powerful and more valuable than numerous other scientific and scholarly explanations for the same phenomena.” There’s little reason to believe that’s the case.

Claim: “[T]he distribution of preferences and abilities of men and women differ in part due to biological causes and. . . these differences may explain why we don’t see equal representation of women in tech and leadership.”

This gets a “mixed.” To understand why, imagine if I claimed to you that biological forces “may explain” why you like hamburgers. On a trivial level, that’s true — you have certain physiological drives toward consuming protein and fat. But this explanation leaves a huge amount out. Why are you eager to eat beef, but would never eat a dog? How are your long-term habits shaped by living in the Boston area, where you’re rarely more than a mile or two from a place that can serve you a hamburger on the cheap? Maybe there’s a biological drive lurking beneath your preference, but your love of hamburgers was endlessly amplified and reinforced by many other cultural and social factors.

Back to the tech world: There is ample evidence of sexual harassment, unwelcoming workplace environments, and in some cases quantitatively measurable discrimination in scientific and technical fields. Why focus on biological differences, many of them quite small, rather than other factors that could explain certain discrepancies? That’s why some researchers, particularly those with a feminist bent, think it’s beside the point that the differences Damore mentioned are borne out by studies.

For a long time, these scholars argue, many people have used such differences to justify inaccurate and harmful arguments about women “just not being cut out for” certain types of work, ignoring the many cultural forces that stymie women’s efforts to break into traditionally male areas. “I glanced at the document but I have seen ones just like it so many times, and the underlying venom of it made me know I did not need to read yet another,” said Anne Fausto-Sterling, a Brown University professor with a lengthy career studying biology and gender development.

Get Arguable with Jeff Jacoby in your inbox:
From the Globe's must-read columnist, an extra offering each week of opinion and ideas.
Thank you for signing up! Sign up for more newsletters here

Claim: Women’s higher levels of neuroticism “may contribute to the higher levels of anxiety women report [within Google] and to the lower number of women in high stress jobs.”

Advertisement

Almost certainly false. Damore leaves himself some wiggle room with the hedge-y “may contribute to” language, but this is an example of how Damore consistently takes something with a grain of truth to it and then reads far too much into it.

What’s true is that, in studies of the “big five” personality traits, women tend to score significantly higher on neuroticism than men, just as they score higher on agreeableness (a fact that Damore also points out) in most countries where such research has been conducted. But to posit that as an explanation for why female Googlers report higher anxiety or are in fewer high-stress jobs fails a very basic statistical rule: Just because women, in studies of an entire population, display a certain trait more frequently on average doesn’t mean that this difference holds for the very small, very specific subset of people who are talented enough to be hired at Google. Why point to population-level averages rather than potentially significant institutional factors within Google?

* * *

Why even bother fact-checking a flawed document that’s caused such a negative reaction?

By the standards of scholarly research, Damore’s memo is an amateur effort — a mishmash of defensible science, questionable speculation, and unexamined assumptions reflecting the author’s own values. Of course, that’s true of lots of other things on the Internet. Damore’s memo, especially when read literally with its hedging language, expresses some fairly mainstream views that, for better or worse, many people agree with.

There are some tricky politics at work here. One reason progressives have so strongly rejected Damore’s memo is that, overall, they’re less comfortable with the idea of deep-seated sex differences than sex researchers are. “Where you will find disagreement [among researchers] is with respect to the origin and implications of gender differences,” rather than in the existence of those differences themselves, said Terri Fisher, a psychologist at Ohio State University. Just because sex differences are real doesn’t mean they explain or excuse unjust outcomes in the real world.

Unfortunately, social media is where nuance goes to die. A memo can’t just be off-base or overly simplistic; its author has to be a reactionary monster, the embodiment of gender inequity in tech. Yet it’s perfectly valid to critically evaluate claims like Damore’s that overplay the importance of gender differences without also taking the unnecessary step of denying the existence of those differences altogether. If humans are going to keep trying to use scientific research to make the world a better and fairer place, ideas — even unpopular and troubling ones — simply need to be described accurately and debated on their merits.

Jesse Singal is a contributing writer at New York Magazine. He is working on a book for Farrar, Straus & Giroux about how questionable ideas in social science go viral.