Women account for less than 5 percent of S&P 500 CEOs, and less than 10 percent of top earning executives. Researchers with the University of Maryland’s Robert H. Smith School of Business and Columbia Business School wanted to look into why so few women are represented among leadership roles, and the results of that study seem to further complicate the question of how to increase gender diversity in business.
After running a series of simulated samples based on the S&P database and other information, researchers found that that a woman’s chances of being hired to a top executive spot decreased by 51 percent if there was already a woman among the five highest-paid positions. And while the reasons why are complicated, Cristian Dezso, an associate professor at the University of Maryland and an author of the study, believes that unconscious bias has something to do with it.
Dezso wanted to find out what "pushes women apart" in real life firms after the study revealed more "bunching" of women -- more firms with two women -- in the simulated sample than in reality. The problem doesn't seem to be other women, he told Salon.
The probability of having another women in top management is still lower at firms with a female CEO, but "it’s not as low as the baseline result," he explained. "So instead of dropping by 50 percent, which is the baseline result, when you have a female CEO, the impact is anywhere between 30 to 40 percent drop."
Instead, Dezso suggested that subtle sexism among the men doing the hiring could be a factor. "It’s not that they changed their mind [after the first woman is hired and now think] that women are not good managers," he explained. "But they might simply relax... they might say, 'Well, we checked off the box, we have diversity, we have a woman.'"
Salon talked to Dezso about the role of unconscious bias in the lack of diversity in business, what his research might mean for efforts to address it and the possibility of fighting "implicit quotas" with explicit quotas.
Our conversation has been condensed and lightly edited for clarity.
Can you explain how you went about your research?
One way to do this is to predict the probability that a particular top management position in a particular firm and year will be occupied by a woman. That prediction can be based on various observable factors. Let’s say, [for example], firm size, firm performance. Or there may be some unobservable forces, like a firm being under pressure from shareholders which we cannot see, and so in response to that pressure they appoint a woman, and so on and so forth.
So what we tried to do is take the population of top managers and, knowing that some factors predict whether a woman will be in the top management of a particular firm, create a hundred simulated distributions. For example, let’s assume that in the population of 1,500 or so firms, there are 20,000 top managers and 2,000 of them are women. We created a hundred simulated samples of 20,000 top managers, 2,000 of them being women, and distributed them across positions and firms. But because some firms are more likely to appoint women in top management, these simulated samples are not fully random. It’s not like throwing darts. It is like throwing darts but you know that some dart boards are bigger than others, so you are more likely to hit those bigger dart boards. So in our simulated distributions we took into consideration the fact that some firms have higher propensities to appointment women to top management positions.
And what did you find?
Basically what we observed was that in the simulated samples, there’s more bunching. There are more firms with two women than in reality. So it means that in reality, there’s something that pushes women apart. If there is a woman in top management already, it decreases the likelihood that there will be a second woman in top management, compared to the simulated samples.
You actually found that the likelihood that a woman would be hired to one of the top five positions at a given company decreased by 51 percent if another woman already held a top position.
Why is that? Or what do you think might explain the difference?
There are two forces that we consider as working in this situation. There might be more, but we consider two main forces. It could be that if you have a woman in top management, that particular woman will work against promoting a second one. This is so-called “queen bee” syndrome. So she might occupy a special place and so she wants to be the only one, but we found evidence against this.
The second thing that can happen is not that the woman that’s in top management is working against another woman being promoted, but the men who are in top management might work against that.
To be more precise, if you want to call it bias or an unconscious bias, it can work in two ways. One could be that men simply say, “I’m completely against [another woman being promoted] and I will undermine the efforts of other women to be promoted to top management.” So maybe some active undermining [is happening].
But it’s not necessarily that they will work against other woman, but they might not work as hard for the second woman as they would for the first woman. Let’s assume that men agree that women contribute to corporate performance, that they are very good managers, they have skills. There might have been some bias in the past, they made some effort to promote a woman. They made the effort, and they promoted the first woman.
It’s not that they changed their mind [and now think] that women are not good managers, but they might simply relax, so to say. They’ve active efforts to find, to mentor, to provide opportunities for the first woman, and then once she’s there, they might say, “Well, we checked off the box, we have diversity, we have a woman.” Now the promotion of woman to top management becomes less of a priority, so they no longer make the same amount of effort that they made for the first one.
One is totally against and the other one [is more], “I’m not against it, but it’s not a priority anymore.”
You called this an implicit quota in another writeup of the study.
Exactly. Because, you know, if it were explicit, you would say that male top managers gather in a room and they start talking and they say “no” and that’s it. So that could be kind of...well, you cannot put that in writing, but it’s in some sort of sense a quota.
I think, whereas in the case [of unconscious bias], they don’t need talk to each other. It’s just, it’s there, because that’s the way their thinking goes. So we call this an implicit quota. The results are consistent as if there would be a [real] quota.
You mentioned earlier that your findings contradicted the “queen bee” theory.
Exactly. So this is the way it works. If there is a woman in top management, we see that there is 51 percent drop in the probability that a second woman will be appointed in top management. This is the baseline result. So now we said that one force is this...what men in top management do. There’s a second force [that] would be what the woman in the top management team does.
Does she undermine the chances of another woman or does she help another woman?
How would you find that out?
The easiest way to look at this is to look at companies that have female CEOs. Because if you think about it, a female CEO is the most powerful [person] in the corporation. If women help other women in top management, we sort of observe this helping effect that would be the strongest when the woman in top management is the CEO, because she has the most power, she has the most say about who will be promoted in top management.
And so what we find is that these companies that have a female CEO, the probability of having another women in top management is still lower, but it’s not as low as the baseline result. So instead of dropping by 50 percent, which is the baseline result, when you have a female CEO, the impact is anywhere between 30 to 40 percent drop. It’s a much smaller drop. So in that sense, having the female CEO serves to counter the negative effects coming from men, but not totally overcome them.
It reduces the disparities, but having a woman in a top position is not a silver bullet.
If your study has led you to concluded, which I think it has, that implicit quotas shape hiring and representation in top positions, what do you say to imposing explicit quotas as a kind of counterbalance?
So I think that’s a very good question, and probably you’re aware of what’s going on in Europe.
That is what I had in mind, actually.
I believe in Norway, in 2006 or 2007, they mandated quotas for women on Board of Directors, which is not the same as top management. They are not as powerful. They are an important position, but not involved in day-to-day decision-making and so-on. All the boards have supervisory consulting or counseling role and so on. It’s a first step. One thing that I would say is that we all talk about the need for equity and equal representation and there are various forces in society that might work for or against this process, or might speed it up or might slow it down.
So the question is can we or the government do something about speeding up the process, and the optimistic view would be “No, there shouldn’t be government intervention because if women contribute to corporate performance and if they want careers, then they will be promoted to top management. It’s inevitable.” But what we’ve seen, at least in the U.S. data, and probably in many other countries, is that while there has been some progress over the past twenty years, it’s been very very slow.
The question is: It took twenty years to reach 10 percent and now the progress is slowing down, so what are we to do about it, if anything?
And the answer depends on the culture of each country. For example in the U.S., where government intervention is not well-regarded or well-looked upon, [people may believe] that the government might not know what’s best for corporations and [quotas] probably wouldn’t fly. Whereas in other countries in Europe where government intervention on various aspects is much more acceptable or much more accepted by society, there it seems to be an accepted way of doing things.
Is there anything that you hope comes out of this research?
If you want the practical takeaway from our research it’s that there’s some evidence that once you have a woman in top management, it makes it harder for a second woman to be appointed to top management. If everybody’s aware of this fact, even the top managers in the firm, they might [want to change it]. Both the managers in the company and various other stakeholders, be it investors, be it NGOs that advocate for the promotion of women in top management. They have to put more pressure, they have to make more effort to reach this desired goal of equity and equal representation, if you want it in top management. So there might not be a need, necessarily, for government intervention, but I think there is a need to acknowledge that this is what’s happening.
So yes, we might have appointed a woman, but guess what, the chances of second one are decreasing and you guys have to make more effort or we have to put more pressure on those managers.
So hopefully this research will raise awareness about this subconscious phenomenon that might be happening, and hopefully that will help to some extent. We can talk in ten years from now and see whether there has been progress yeah, and if there’s none, then of course we can revisit the question.
But I guess it very much depends on the culture. As I said, one thing is Boards of Directors, and another thing is top management.
Your research is being published at a moment when a lot of people are talking about diversity in elite corporate spaces. Ellen Pao’s gender discrimination suit very much brought this back into public debate. But, more directly, Pao just eliminated salary negotiation as part of hiring at Reddit. She was explicit about why, that negotiating is a kind of minefield for women and so she just cut it out of the process. It’s related to what we’re discussing, this idea of making diversity a priority in a very explicit way. But critics tend to call this social engineering.
I think there are a lot of moving parts, and it’s very hard to give a definite answer. So yes, you could say that it’s social engineering and it’s intervention, but one could say the same thing about establishing a culture whereby only men can raise to top management, be it by actively discriminating against women or subconsciously. That’s also some sort of social engineering. The question is which one is the right one, which one is not.
It’s funny because some people bristle very hard at calling the status quo a kind of social engineering. They only see intervention when it’s a change from the way things work right now.
Exactly. And so, what do you say? That one is better than the other? And if you argue for that, then why would you argue the culture that increases the chances for men is the right one?
In terms of salary, it’s a very good point. On the surface, I think it looks good in that if it’s true that men that are able to negotiate higher salaries and that’s only due to negotiation. if that’s the only reason, then this might be a welcome change. But I’m thinking that it might desensitize people, there could be still other forms of [discrimination].
The data on women and negotiation is fairly nuanced. Women sometimes do not negotiate as aggressively when doing so for themselves, but other times, when they do negotiate aggressively, they can be penalized for it.
Which again, it’s part of the social engineering. The culture has evolved in such a way that women are put in a position where they cannot… it’s not that they cannot negotiate, but they are not well looked upon. You can call it a double-bind. It’s like the double-bind of women, how do they behave in the workplace to get the top management position, if they exhibit behavior typically associated with women, they may be treated like they don’t have the skills for leadership, which are typically associated with men.
But if they exhibit the behaviors of leaders which are typically associated with men, they will say, “What’s wrong with this person? Why is she behaving like that because it’s not natural.” And so there is this mind frame and then if women internalize that, well I cannot behave like that because this is how I will be looked upon, they will not behave like that, and then when observe that they don’t negotiate. I think it’s a very very hard position and I think this policy [of banning salary negotiation at Reddit] on the surface it sounds okay, but you have to couple it with something else that levels the playing field for men and women, in terms of chances of getting those positions. It’s not only that I get the same salary, but I need to have the same chance of getting the position.