Thoughts on the Ellen Pao Case, and the Gender/Silicon Valley Interaction

The Pao v. Kleiner Perkins trial, decided completely in the defendant’s favor last Friday, had all the ingredients of a gripping movie: sex, gender, money and power.  What a plot! An Ivy League-educated Asian-American woman (“daughter of Chinese immigrants,” the press loved to say, as if Pao’s father had been a Chinatown dishwasher rather than an NYU math professor) sues her uber-powerful venture capitalist employer for sexual harassment, gender discrimination and retaliation.

Titillating as all this is, I’ve found the press coverage on the central issue — gender discrimination in the tech industry — to be rather misleading.  For instance, all those headlines like “Pao Case Roils Silicon Valley” are greatly exaggerated.  On the contrary, I believe that at least among engineers, most people are barely aware of the case, and more than a few are not aware of it at all.

In this post I’ll comment a bit on the case, and then discuss the gender issue in Silicon Valley as I see it.

If nothing else, Pao had courage to lay out her entire life for the world’s salacious viewing, as she pointed out somewhat defensively but quite reasonably in response to a juror query.  But as the headline “Pao vs. Pao” in one article noted, she probably was her own worst enemy.

Putting aside her personal actions that presumably put her at a disadvantage in the case right from the start, she was portrayed in company records and by numerous witnesses as having a sense of entitlement, apparently because she felt her Princeton and Harvard degrees gave her an imprimatur of top talent. Whether this charge was accurate or not in Ms. Pao’s case, my own observation is that it rings true for many in the Valley who attended the Ivies. One lesson to be drawn is that mastering the art of gaining admission to Ivy League schools does not automatically imply that one has the chops to do well in business and industry. (This lesson, of course, needs to be learned particularly by all those “Tiger Moms.”)

On the other hand, certain claims made by Pao ring true too. Based on my interaction with Valley types, Pao’s claim that a colleague said that the presence of women in meetings “kills the buzz” sounds just like what some men I know in tech would say.  (The colleague denied using the phrase.)

Another Pao claim that sounds plausible to me concerns seating arrangements at meetings.  Pao discussed her not being seated at the “main” table, and took her exclusion to be gender-based. At first I wasn’t sure what to make of this, but an incident that occurred in my class last week seemed to bolster Pao’s claim, as follows.

In any undergraduate course I teach, I have students work on programming assignments in teams of three or four. On the last day of class, I also give a team quiz, in which students work in their teams, on their laptops. Usually the students arrange the desks in circles so that team members can interact well with each other, but in the classroom last week the desks are bolted to the floor.

During that team quiz, I noticed that one team consisting of three men and one woman had the men seated in one row, and the woman seated in the row behind them.  She was constantly craning her neck to try to participate, and I never saw any of the men turn toward her or say anything to her during the entire 75 minutes of the quiz.  A second three-males, one female group was similar. (The other female students in the class formed their own all-woman team.)

My wife believes she was the subject of little or no gender discrimination during her many years as a software engineer in the Valley. Indeed, she was approached a couple of times to move into management. But not all women are as assertive and articulate as she, and though many men lack these traits too, I do believe that women are penalized more heavily for those “deficits.”

I place the latter word in quotation marks because I believe a good case can be made that the typical hiring procedures at Silicon Valley firms are misguided, placing too much emphasis during interviews on quickness and talk like, say

The next version of the gcc compiler will enable OpenACC directives, and wow, that will be SWEET! I’ve got a high-end NVIDIA card on my Linux box, and now it will be easy to port the awesome games I plan to write.

Startups, especially those in the SoMa district of San Francisco, are particularly prone to value this kind of behavior in interviews. And here is the gender aspect: I’ve only rarely seen women who talk like that. Most men don’t either, but the ones who do are almost all male. While it is understandable the employers want to hire workers who have enthusiasm for their field (though much less justified to value quickness), I believe that this view of hiring is wrongly disadvantaging the women, and is causing employers to overlook many top-notch talents who happen to be female.

A couple of months ago, I was asked to meet with a Department of Labor attorney to discuss whether the H-1B work visa program is exacerbating the gender imbalance in the tech industry. Given my criticism of the program, she was quite surprised when I answered her question in the negative. Actually, I told her that although I had not yet done a formal analysis of the data, my impression over the years has been that, at least among the foreign students, the percentage of women in that group is higher than that among their American peers.

I should mention that one does need to distinguish between Silicon Valley proper (Santa Clara County) and San Francisco.  The former is heavily immigrant — Indians of course, with a significant Chinese minority — and married with family.  The latter group has some immigrants, but more often from Europe and the like, and is typically single. And though both Indian and Chinese cultures are putatively sexist, I believe on the other hand they are less prone to insist on hiring someone because he said, “Version 4.9 of gcc will be SWEET!”

In short, I contend that there is in fact serious, albeit subtle, gender discrimination in the tech field. I’m much less sure that Pao’s case helped to bring this out.