The Toll of Tokenism

Maybe you’re a token too

In her piece — a rant, really—about Twitter’s “female problem,” Sarah Lacy says putting a woman on the board would amount to nothing more than tokenism. It would be if the appointment were totally perfunctory, and if there the woman’s appointment were for gender reasons only. But let’s get real: That’s like people telling you the only reason you’ve succeeded is because of your appearance or your gender. Do you really think, in a world of women and people of color like those you’ve named and others who must also certainly be doing, achieving, and changing the ratio, there are none who are fit to sit on Twitter’s board and offer their perspectives? None who are as qualified, perhaps even more, than some of the men who were chosen?

Way to go, Sarah. Way to not get it.


There’s a lot about tokenism that’s interesting but that doesn’t get discussed, because people—yes, even in the tech industry—don’t often like to look at academic research. A brilliant scholar, and a woman no less, named Rosabeth Moss Kanter did groundbreaking research on tokenism theory in the 1970s. We can expand tokenism from the act of simply hiring token minorities to encompass workplaces in which a minority group makes up a small percentage of the workforce, usually less than 15%.

See where I’m going with this?

Kanter’s work is decades old now, I know, but you’d be surprised at how little has changed when you read her article, which you should. It’s free, and it’s eye-opening beyond all measure.

What you’ll learn, among many other things, is how tokenism is more than about the hiring practice. It’s also about how shifting numbers can, over time, shift perspective and the “surprise” of the token.

The proportional rarity of tokens is associated with three perceptual phenomena: visibility, polarization, and assimilation. First, tokens, one by one, have higher visibility than dominants looked at alone: they capture a larger awareness share. A group member’s awareness share, averaged over shares of other individuals of the same social type, declines as the proportion of total membership occupied by the category increases, because each individual becomes less and less surprising, unique, or noteworthy; in Gestalt terms, they more easily become ‘ground’ rather than figure. But for tokens there is a ‘law of increasing returns’: as individuals of their type come to represent a smaller numerical proportion of the group, they potentially capture a larger share of the group members’ awareness.

Changing the ratio is good for women, but is it good for Sarah Lacy?

Because just as importantly, there is the effect of tokenism on tokens themselves—which a person might be without even realizing it. How many powerful women are there in Silicon Valley? Far fewer than there are men. So where some of you are are seeing “Stockholm Syndrome”, I’m seeing the effects of tokenism. After all, as Kanter writes, “For token women, the price of being ‘one of the boys’ is a willingness to turn occasionally against ‘the girls’.” Put-up-or-shut-up is a way of going-along-to-get-along because noise is trouble.


That some women have achieved at Lacy’s level, at Sandberg’s or Mayer’s or Beauchamp’s or Amoruso’s is wonderful but is not the point. Tokenism is partly a numbers game and it is partly a culture game. The culture is still one of the dominant category. That Lacy has to call out the women who have achieved at this level and identify them as such speaks volumes. There are CEOs and there are female CEOs. Why make the distinction? Because we still have to. And that’s the problem so many people want to talk about.

Yes, there has been subsequent research since Kanter’s work, such as how tokenism isn’t just about numbers, and plenty more. But start here, and you’ll see two things: First, how easy it is to concede to a dominant culture, because it can be the least resistant path to success. Second, sorry baby, we still have a long way to go.