Necessary but Not Sufficient

Research has consistently documented unequal outcomes for men and women in the workplace, but to truly fix the problem, we need to begin to uncover the drivers of gender inequality.

Transcript

Today, we’re going to be talking a bit about my research on gender inequality and understanding the drivers of this inequality. But I want to get us started with a quick image. Something that most of you have probably seen, right? We’ve seen a lot of headlines around the SpaceX Falcon Heavy launch this week. Most of these headlines have been focused on how monumental this is, right? It’s a big deal, a big step forward for science. We have this car-looking object in space that could potentially be in orbit forever.

Another category of headlines that stood out to me, though, is about how this isn’t all great. Right? There’s maybe a negative element related to this launch. And it usually ties back to an image like this.

This a room full of engineers who were celebrating the launch at the time that it happened. So, here, we’re talking about gender today. So, if you look at this picture, you’ll be hard-pressed to find many women, if any at all. I’ll also comment that it’s a lack of diversity overall. It’s not just that there aren’t many women, there is a room full of white men.

So, what does this actually mean? Is this a unique experience? By definition, the nature of the panel, we know that, no. Right? It’s actually the fact that, very often, we’re talking about this type of gender inequality. We’re talking about the absence of women from science and technology.

Arguments like this, why are women missing from STEM? We see women enter into these programs, but then, through the career pipeline, they spill out at some point, and we don’t actually see them persisting through. The reference to the glass ceiling is another familiar one. It’s not about STEM and technology, per se, but that women are just missing from the upper ranks. They’re not present in senior management. They’re not in the C suite. They start out at some point, but what happens? Why aren’t they there? A different flavor of gender inequality, it’s not just all about where women are missing, but when they’re there, they’re making less. We often hear statistics like the one up here, “For every dollar a man earns, women are earning about 80 cents on the dollar.” It’s closing, it’s better than what it used to be. It used to be 60-some cents on the dollar. But it’s still not parity.

We now have equal pay day, to represent the day in the following year that a woman has to work through to earn as much as a man earns in one calendar year. It’s sometime in April, it sort of fluctuates from year to year.

Thinking about technology, in particular, there are sectors that are dominated by men, where we see that even when women do enter in, they’re earning less within those sectors, often doing the very same job. And there’s call to action, right. Companies constantly responding, saying, “We’re going to fix it. We’re going to do something to make it better.” BBC recently promised to close the gender pay gap by 2020. We’ll see what actually happens.

A lot of the background on why we care about this is social. We want there to be social justice. It seems unfair that if I’m doing the exact same work as one of my colleagues, I’m making less money for doing it. But beyond the social objective, we’ve heard a lot about the business case for equality, rather, for diversity, right? We started this morning with Kathy. We heard our panelists talking about how greater diversity means better outcomes. It means more success for businesses.

This quote from Melinda Gates sums it up, “If we don’t have women in the tech space, we won’t even be asking ourselves some of the right questions.” It suggests that, by not having women in the game, we’re not thinking about our consumers in the same way. It means that if we don’t have women as VCs, we’re not funding companies that could be the next success story, thinking about Pinterest, from an earlier example, as well.

All of this leads us to wanting to essentially redress inequality. Whether you’re motivated because you want social justice, or you’re motivated because you’re a capitalist who wants to make money. Ideally, some combination of both of these is at play. In order to redress inequality, we really do need to move beyond this notion of documenting that it’s there. I don’t think anyone in this room would be surprised by some of the figures and headlines that I shared. We know inequality exists. But the problem is, we keep fixing it by identifying that it exists and trying to fix that imbalance. We say, “Women are making less, so we need to pay women more.”

But that’s sort of a Band-Aid approach. Right? It doesn’t actually get at the root causes, what I’m calling the drivers of inequality. It’s only when we understand where the leak is happening, why women are exiting, where the imbalance is happening in pay, how these processes are actually unfolding, that we can begin to correct the problem fundamentally.

So, I want to share two parts of my research today that touch on very specific concrete conditions under which we see inequality, what these drivers actually are. So, part of this looks at evaluations. And I do research on evaluations, because evaluations, in a lot of ways, are everything. When you’re getting hired, the evaluation begins from the very point that you submit that application, whether or not you’re getting attention. Are you getting looked at, moved into the next pile?

I then talk a bit about networks. And again, we hear all the time how important it is to develop your network. But, are networks contributing to imbalances in ways that we aren’t readily seeing?

I’ll start with the evaluations piece. As I already began to tell you, evaluations are a key way by which women and men experience outcomes. I mentioned hiring, we think about setting pay and getting promotions within the firm. When we think about winning awards, being evaluated for a work product, these evaluations are at play and govern who actually gets selected, who gets recognized.

Very often, gender is thought to factor into evaluations when we don’t have perfect information. So, for example, if I’m comparing two candidates and I know that math skills are really important, but I can’t perfectly observe the man and the woman’s math capacity, how good they are at it, I might think, I know men tend to be better. So, I fall back on these heuristics, these biases, and go with what I think is the safe choice. I choose the man because I expect that men, on average, are better at this skill that I can’t observe.

So, in some of my research, what I try to do is compare men and women where you can very clearly and directly observe objective quality and performance. When we have this quality information, do we still see bias? This is what we often refer to as double standards. Do women essentially need to outperform men to get the same recognition, the same evaluations?

So, I’m looking at investment professionals, so, hedge fund and mutual fund managers, who are posting stock recommendations on an online platform. They’re posting these recommendations so that their peers can then go on, look at what they’re suggesting, and decide whether or not to pay attention to them. Do they click on the tagline for the recommendation and read the couple thousand-word model backup supporting that actual recommendation? The important thing here, is that, at the moment that they’re deciding whether or not to click on a given recommendation, they see a person’s name. So, they know that the person is named Michael, or named Mary. And they see their performance to date.

When we interview these people, they say, “We don’t care about gender, of course, we don’t. We’re capitalists. We’re here to make money. If I see a strong performing recommendation, that’s what I’m going to go for.”

Unfortunately, the results suggest something different. What we actually see is, compared to a person named Matthew, who posts the same exact recommendation, when someone named Mary posts that recommendation, she’s 14-percent less likely to be clicked on, to get attention from her peers, to be reviewed. The interesting thing we were able to do here is, we were able to limit the analysis to just men.

So, what that allows us to do is basically say unless we fundamentally think that a man named Kelly, whom someone might interpret as being a female, fundamentally differs from a man named Matthew, we shouldn’t see an effect. But, in fact, we see the exact same effect. So, even if you have a female-typed name, you might experience these same penalties on a platform such as this.

So, I told you I was going to focus on drivers. The good news here is that this effect isn’t always present. The driver is really related to a sorting heuristic. We see this gender penalty when a person is deciding among a large pool of options. So, basically, if I’m looking at a hundred different things that I could look at, that I could spend time reading, I used gender as a filtering tool to decide which ones I’m going to focus my attention on.

Moving on to networks and thinking about network’s social capital or relationships. We heard a bit about how a lot of the deals that get sourced really come through these networks. We know they’re important. What I wanted to understand, is when we compare men and women who have access to the same networks, so they basically have the same opportunities to create benefits, the same opportunities to access resources, do they get the same value? A common explanation for why women might get less from their networks is that they just have worse networks.

So, what I did was, I found a setting of entrepreneurs where all the men and women that you see represented in that red circle, essentially have the same access to the same set of social ties. They’re looking to exchange business connections. They want to grow their businesses and they want to do so by leveraging their social connections. “Do you know someone who’s looking to hire an attorney? I’m looking to grow my business in this area.”

So, what I was able to do here is look at the number of business referrals the men get relative to the women—again, all things being equal, people in the same industry, et cetera. What I basically find can be summarized here. Imagine that the green person is the person deciding who to connect to. If I’m deciding to hire that lawyer myself, there’s no difference in the likelihood that I go to a man versus a woman. So, I’m not exhibiting bias in my personal decisions. But on the flip side, if I’m thinking about my outside contact, the yellow dot represented outside of the circle, I’m thinking about my client. Do I connect my client to Jane or to John? I’m much more likely to connect them to John than to Jane. And again, the driver that I’m identifying here has to do with anticipating the gender preferences of others.

So, even if I myself am not biased, I’m not sure if my outside contact, if that client has a preference for men. So, it’s safer to recommend the man and to avoid the potential threat that would come about, had I recommended Jane.

So where does this leave us? I started off with this idea of redressing inequality. It’s not for lack of effort on companies and legislation to try to redress this inequality. There are tons of lists of the different things you can do. How can we actually correct this? Maybe we can throw more money at the problem, right? In Silicon Valley, millions of dollars are being spent to improve outcomes. Yet, the situation isn’t really getting better.

We see the argument that maybe we need to get men involved. And I’m happy to see the men in the room today. Roping the men in, making it more than just a women’s issue, getting everybody on board. Policy and legislation changes. We recently saw states starting to adopt this, “You’re not allowed to ask previous salary questions.” Because, by asking previous salary questions, you’re anchoring women who are already earning less historically, to continue to earn less.

None of these solutions have been effective in and of themselves. So, what I’m trying to highlight with my research, with the two things I shared with you today, are really the conditions under which we see this inequality, and how can we interpret those conditions to really get at actionable items, steps that we can take to help improve the situation.

So, linking back to the first piece I shared with you, related to evaluations. I told you that the bias was really there when there was a large volume of options. So, one thing companies can actually do is think about when an individual is sorting through applications, for example, how do you minimize the number of candidates that they’re considering? Is it really necessary to have one person looking at 500 applications? Knowing that bias is more likely to be introduced in that situation, you might just think about reining that in, having fewer options to look at any given time.

Thinking about networks, I told you that this bias really came from this third party, right? When I’m thinking about someone else’s preferences. So, for women, the advice would be, make the connections directly to your primary resource holders. If you know someone would be really valuable to you, valuable to your business, you don’t want to rely on someone else to speak on your behalf. You want to develop that tie directly. You want to overcome the potential that somebody could be hesitant to connect you simply because they anticipate the partner being biased.

So, all of this is to say that these two steps are two of many steps. But we really need to take strides to move beyond showing that inequality exists. I think that’s the easy part. The difficult part is moving beyond these baseline correlations and just showing that we see differences in patterns to start to really get at underlying mechanisms and understand how to unpack these.

---

Ivy Exec is proud to announce its partnership with Columbia Business School, to bring an insightful collection of thought leadership pieces for the modern-thinking strategist in finance, leadership and more to its platform.