Shaping the Future of Finance: Recent Gender Diversity Research

CFA Institute Alpha and Gender Diversity Conference
18 September 2017
Toronto

ROGER URWIN: Good morning again. What a great pleasure to be moderating this event. Before we kickoff, it really is worthwhile, I think, emphasizing that link. So what we’re talking about here is shaping the future of finance. And new research and thinking on gender diversity in that context.

Now, the future of finance depends critically on meeting the challenge of gender diversity going forward. And that’s an assertion, but it’s so important I’m going to say it again. The future of finance depends critically on meeting the challenge of gender diversity going forward. And this session has its purpose really in doing a rigorous job on some of the research and thinking that lies behind that statement.

So I’m delighted to be turning, in a minute, to Heather. Heather is, again, someone who doesn’t really need too much of an introduction here, I think. So until recently, senior in the Morningstar organization. Now, First State Investment head of the Americas. A very interesting job she has in both respects there. But as importantly, I think, vice chairman of the Board of Governors of the CFA.

And Brad Barber. So Brad is a professor in finance, a social science application of finance, at UC Davis. Author of a number of studies about the subject of gender diversity in its widest context in finance. And his co-authors here, Adams, and Odean, and Barber make up, I think, the most respected source of research in this particular field.

So my job is a very short framing, but let me just get some participation here. So a question, no clickers. We’ll do it in the traditional way. So the question is, and I’ll just read it out so everyone gets it clear. I think gender diversity and inclusiveness is progressing well or badly in the investment industry. So I’m not giving you the middle course. So, I think the investment industry is progressing well or badly in the investment industry.

So those that say it’s progressing well, hands up, please. A little bit of a pause, you see. So that’s a group dynamic, pretty interesting, but a few people. But very few. Just for those looking in on the ether, I could count that in about seven or eight people. And who thinks it’s progressing badly in the industry? And we have a 90% show.

And let me ask the question again in the context of your organization. So your organization. I think gender diversity and inclusiveness is progressing well. Who thinks that at your organization? Hands up, please. And we have probably a 20% show on that, might be a little bit more. And then the number of people who think it’s progressing badly at your organization. And we have — that’s interesting. Not quite as many there. And a lot of people in the middle on that particular question. Or a bit hesitant to declare. It could be either.

So one of the things that this session is trying to help with is trying to put some context against those particular points. Let me just position this subject in framing. So, if you will — four points, really. Firstly, as Paul identified, what we have with gender diversity is a subject that needs to be put in the context of cognitive diversity.

So gender diversity and cognitive diversity have a relationship. And they both have a business case. That business case is here to be discussed. But those — that business case is really, really critical to what we’re doing here.

Point two, there is a sweet spot for diversity. Now, I think many people would think about it as being a maximizing. But really, it’s an optimizing. Insofar as there are activities where being tribal, and not necessarily having maximum diversity, works out well. But in the complex area of many investment subjects, we need more diversity.

And in the area where I identify it at, roughly speaking, as a 30%, 20%, 10%, [INAUDIBLE] some pipeline of 30%, a lot of core of 20%, leadership at 10%, we are way short of the diversity we need for the investment industry. It’s not even close. So that’s a very key issue, which we’re obviously going to go deeper into.

And then, the strategies for diversity and inclusion. And really that starts with evidence and beliefs. Witness this session. It majors on aspirations at the soft end, quotas at the hard end. It also has to major on empowerment and engagement. And by the way, one of the things we’re going to look into quite a lot, in this session, very interestingly, is mentoring. One of those key facets of

[INAUDIBLE].

Now, the last point, really, to make is that it’s through the business case for diversity that progress is made. So it’s really a financially laden subject that we’re dealing with here. And really, that is an effective strategy without having to worry so much about what Paul referred to as that social fairness. So there’s no question we have social fairness here.

But really, the emphasis in our financial world, with its values of finance, is that that becomes a collateral benefit. It’s not necessarily read out as the issue that we’re trying to battle with. Our battle is through a financial lens. And that is, I think, well worn as a strategy success factor.

So I wanted to just put those points together because they will feed into both Heather’s talk and Brad’s talk in a minute. So Heather, delighted to have you come up. Heather’s got around about 25 minutes of material. Brad has a bit longer, deeper research that we’ll be taking straight afterwards. Please welcome Heather to the podium.

[APPLAUSE]

HEATHER BRILLIANT: Thanks, Roger. Thank you all for having me here today. I’m a little shorter than Roger. Brad will have to move those back up if he decides to stand here.

I wanted to tell you a little story about my presentation before I get started. And actually, before I do, I just want to say there’s this amazing new invention. I’m still miked. Can you see it? This is a new — this is the first time I’ve ever seen this. It’s a necklace mike. So, way to go AV guys.

[LAUGHING]

Very cool for our conference focused on gender diversity, especially. So the story I wanted to share with you is just really about my recent career change. And so, I was supposed to present today on the latest Morningstar data on gender diversity. So the idea was that I would present whatever Morningstar was coming out with, and Brad would present his latest data. And it would be a lot of original research, which I was really excited about. But a couple of things happened in the meantime.

First, I decided to leave Morningstar. And second, I left on very good terms and still have, obviously, communicated with the Morningstar team. And they don’t have any new data yet. So — so as a result, what you’ll find is that, in my presentation, I’m going to talk a little bit more about the data that is out there in the open. So it’s nothing that you may — you may have come across some of this data before. But I think, hopefully, we’ll put it together in an interesting way.

And so, just anecdotally, while we’re on the topic of me switching jobs. So when I announced I was leaving Morningstar, there were a lot of women at Morningstar that were devastated. And not because I am necessarily so special or unique or anything like that, but the problem is, I am a little special and unique. And especially, that I was the only country manager at Morningstar that was a woman. And I was one of very few business leaders at Morningstar that was female.

And so I got this really overwhelming amount of emails from women at Morningstar saying, you know, we really looked up to you, we really valued having you as a role model. And that was wonderful to see and great to know you have an impact.

But the email that surprised me the most, and the one I wanted to just share with you briefly, is that — I actually got an email from a male colleague who said, I never knew that having a woman leader could make a difference until I had you running our business. And I was so stunned by that because I think all of us in this room can agree very quickly that having a woman leader can make a difference. And that there can be a lot of wonderful impact from having a diverse team.

So anyway, so I’m really happy to be able to talk a little bit about that as my topic today. And I also want to say that, while I wanted to share that anecdote with you, it’s an anecdote. And we can’t really draw any conclusions from anecdotes. And so, what I’m really going to try to do today is to present the data on the topic of what works when it comes to diversity programs. So, oh, I have to click my slides.

So this is the Morningstar data that I was able to come up with. And this, actually, I’m not sure has actually been published yet anywhere. But it’s just really a time series of how many female portfolio managers there are in US portfolios. And it’s separated by international stocks and taxable bond funds and US stocks. But it’s all US portfolios, specifically.

And the reason why I wanted to share this with you is because you can see that the data is pretty abysmal. Specifically, that it’s going in the wrong direction. So, over the last 15 years or so, we’ve actually seen a material decline in the number of women portfolio managers in the industry. Well, why is that?

So I tried to really start thinking about, well, what is it about what we’re doing as an industry that could be preventing women from excelling, or moving forward, or continuing to advance their careers? And I found this stat, 66% versus 22%, in the CFA Institute Gender Diversity report, which is what Paul was referring to in his opening remarks.

And does anyone know what this might refer to? 66% of female CFA Institute members carry most of the dependent care responsibilities in their house. 22% of male members carry most of the dependent care responsibilities in their house. And I thought that was just a really interesting data point to share with you because I think it does help highlight that we are not necessarily starting from an even playing field.

And another interesting point from that study that I wanted to share with you all is that a woman is far more likely to have a spouse with the full-time occupation. 79% of women have a spouse who works full-time, compared to their male counterparts, where it’s 51%.

So, either way, we all know that if you have children in your house, if you have pets, if you have any dependents, elderly dependents, there’s a lot of care required in all of those situations. And what we find is that falls disproportionately on women. In our industry, specifically. But I think we can certainly talk about that more broadly too.

So then I thought, well, let’s take a look at what our industry thinks about the importance of diversity. And wow, I was so disappointed, actually, to see that in a recent CFA Institute Research Foundation report, 45% of investors believe gender diversity does not matter. And that’s pretty sobering when you think about it, because it really says that it’s going to be an uphill battle to convince the industry that gender diversity matters.

But there’s a couple of points I want to make on this, in particular. First of all, the good news is that 48% of CFA charterholders believe diversity — the diversity of viewpoints that comes from mixed gender teams is valuable and does add to performance. So there is some belief that, yes, the data that supports that gender diversity does matter is useful, and is something we should use.

But another thing I wanted to point out — and I actually think this relates very closely to some of Paul’s remarks — is that there’s actually a lot of different ways to look at diversity and the benefits of it. And so that’s really what I’m going to talk about in the next couple of slides.

So, how many of you were at the conference last year? That’s wonderful. I actually couldn’t make it last year. As many of you know, I was living in Sydney. And it was just one of the trips I was unable to make. But I did recommend a speaker, who, hopefully, you all heard. His name is Scott Page.

So, does everybody remember Scott Page? Yes. I’m seeing lots of nods. I think he is just amazing. And when I read his book, The Difference, it made a huge impact on how I thought about diversity on my own teams, as I’ve been going through my career

And that book came out about 11 years ago. But Scott Page actually just published a brand new book that literally came out, I don’t know, a week ago, if it’s even published yet. And Julie was kind enough to get me a copy. So I was able to read a good chunk of it on my way over here. It’s called The Diversity Bonus. And it is an absolute must-read. And I’ll tell you why.

The way Scott Page really looks at diversity in this book is to focus on the importance of diversity when it adds to the cognitive analysis of a problem. And so, he basically says that if we just advocate for diversity for diversity’s sake, or for social benefit, that we’re actually more likely to cause backlash, and more likely to have issues that will prevent us from achieving our goal.

But, in fact, if we can focus on ensuring that we have diverse teams, where diversity can really make a difference, specifically on very complex, more cognitively oriented problems that need solving by multiple different perspectives, we can prove the benefits of diversity.

And, interestingly, I think that also advocates for the fact that we need diverse teams. We don’t need, necessarily, teams where everyone is the best in their field. And Scott Page goes into this in great detail in his book. It’s so fascinating to think that, if you have a team of, let’s say, eight portfolio managers, who are all the very best portfolio managers they can be, actually, adding another fabulous portfolio manager to that team could have zero positive impact from a diversity perspective. Whether that is — that portfolio manager is a man, woman, of a different race, of a different sexual orientation, anything.

But, in fact, if you add someone to that team who, perhaps, is a mediocre portfolio manager, but has some very different perspective or way of looking at — way of looking at things, that’s where you can achieve the diversity bonus. According to Scott. So I’m going to talk a little bit more about some of the things he brings up in his book as we go through this.

So I think a lot of people think that we can improve diversity in the industry by focusing on diversity training. And I wanted to talk a little bit about why I don’t think that is necessarily the solution. And some of the research on this actually came from a Harvard Business Review report that I think I cited on the next slide, but it should really be on this one too, called, “Why Diversity Programs Fail.”

And so, this study, actually, really showed that mandatory diversity training can be very challenging for a number of reasons. Not the least of which is that it can cause backlash, because people feel like they’re being forced to accommodate diversity. And they don’t see it as something that they want to do, or that there’s a real benefit to doing. And the data is really fascinating on this.

And I think that this Harvard Business Review study really goes into it in pretty good detail. Specifically, five years after instituting mandatory diversity training, companies saw no improvement in the proportion of women or minorities in management. No improvement. Also, performance ratings are seen as a way — we can look at performance ratings, make sure we’re being fair and equal in how we distribute our performance ratings.

But actually, the data from this Harvard study shows that that doesn’t work either, because managers are very effective at working around performance ratings. And they know how to game the system. So, if they know that that’s what’s being measured, then that ends up not being effective either.

And finally, of course, I think we can all agree that bias can be unconscious. And so it can be really hard to train around it or to force people to get outside of their unconscious biases. So what can we do instead? Well, in this particular Harvard Business Review article, there’s really a few things that are specifically cited. And the reason why I think these are so interesting is because there’s some great data backing up what has happened five years after these types of programs were instituted.

So let me talk a little bit about each of them. Engagement, exposure, and accountability. So first, on engagement. What does it mean when we even say engagement? And I think the idea, really, is to think about voluntary diversity-related programs. And actually, the Harvard Business Review article makes the point too, that you might not even want to call them diversity programs, but really just advocate for different types of engagement among employees.

And a couple of things that they mentioned specifically, are voluntary training, which allows employees to feel like they’re taking initiative. Even college recruiting efforts can be a great way to engage people in the workforce. In helping bring in the next generation of leaders, trying to make sure that you’re bringing in a diverse group of future leaders that people — the data actually shows they really engage on that.

So specifically, for companies that adopt college recruitment programs, five years after they’re adopted, the share of women in their management team rises by 10% on average. So that’s pretty powerful data.

There’s actually, in this particular article, if you’re interested in all the specific data, they really go into great detail about women by gender, men by gender. I mean —

[LAUGHING]

Sorry, women by race —

[LAUGHING]

— and men by race. I did not sleep last night.

[LAUGHING]

So anyway, it is — there’s really fascinating detail, I think, on how the different types of engagement can have an impact on how many of the senior managers of a company end up being women, or being minorities of different groups, five years after some of these programs are implemented.

The second area I wanted to talk a little bit about is exposure. So exposure is really focused on the idea that people working together to solve a problem realize very quickly that there are benefits to diversity, and that there’s actually no reason in judging someone based on whether they’re a woman or a minority. But rather to judge them based on what they’re bringing to the team, and the contributions that they’re bringing to the team.

And so, a couple of ways to do this specifically, to increase exposure within your firms, are to A, to start a task force around diversity. So task forces specifically, were found to lead to an 11% to 24% improvement in the proportion of managers who are women after five years. And the difference in that 11% and 24% is really whether it led to an improvement for Asian women versus African-American women versus white women, et cetera. So the diverse — the data behind this is really very rich.

And so the idea behind a task force, really, is that you get people working together to solve a common goal. And they’re working shoulder to shoulder and really, I think, collaborating. Which leads to a great level of openmindedness and gets the debate away from just being about — just being about diversity, but also to expand the problems that you’re trying to solve as a firm.

One way that I’ve actually seen this done very effectively in my career is actually using Agile, and trying to come up with cross-functional teams to solve problems that you need to solve as a business. So, you might have the research team over here and the trading team here and the operations team here, and they’re all — they all report up through their functional lines, and that’s great.

But getting together to solve a problem, like, oh, well, how can we recruit more women? Or, how can we make sure we are implementing the right CRM system, which really affects everybody, can be great opportunities to bring teams together across functional lines and across diversity lines, as well.

Another thing that this Harvard Business Review article cited about exposure that works is rotation programs. So, if you have a formal rotation program — where you take people from one area of a business to another, move them around effectively — that has been shown, by this Harvard Study, to boost the number of women and minorities in management by 3% to 7%, which is pretty impressive.

And actually, those cross-functional teams with diverse groups working side by side, lead to a 3% to 6% increase in the number of women and minorities in management roles over five years. So, one other thing I wanted to mention — well, actually I’m going to save that. Let’s talk about social accountability.

So this is the third area that I mentioned. And really, the idea behind social accountability, which you can probably get a sense of from this image, is transparency. And so, this Harvard article talked about how — in fact, if people know that their behavior is going to be watched or observed, and that the data behind how diverse their team is, how fairly they’re treating their team, how fairly they’re compensating their team — if those data points are shared broadly within a company, then the data in this article basically demonstrates that you’ll see an 8% to 30% improvement in the number of women and minorities in leadership roles.

And actually, there’s been some really interesting examples of this. Deloitte is a really interesting example. So, I think about a decade ago, Deloitte’s head of the consulting area tried to implement some task forces to improve diversity and to make sure that women were getting the same opportunities that men were getting. With regard to some of — being put on the best projects, and being able to, essentially, have equal opportunity.

And so, all he did was that he implemented a task force and he published the data. So he said, oh, in this — within this practice of our consulting business, this many women have been given the lead partner role on — with these client engagements. And over here, the women are making 87% of what the men are making.

And so, essentially, the transparency of data has led to a material increase in the number of women who are getting those roles. In fact, it’s gone from — I could be a little bit off on this — but something like 17% to 35% of the engagements at Deloitte are now being led by women. And they now also have a woman CEO, and they have a woman Australian CEO, as well.

So those are just a couple of ideas that, I think, came out of this Harvard article that I wanted to share with you guys. But then I wanted to also dive into a couple of other things that I think we as an industry can do in a little bit more depth. In order to really be able to walk out of this room and say, I feel like I have the data to understand why I should be considering some of these courses of action at my firm, and also have some ideas for how I might be able to implement it.

So, I think one of the topics that comes up a lot with regard to diversity is, how do we improve the pipeline? And I think we can all appreciate that we do not get as many women applying for our roles as we would like. And it feels like an excuse sometimes, it feels there must be women out there, how can we find them?

So I tried to dive into that a little bit. And one of the greatest sources on this, I will say, that I definitely recommend, if you haven’t come across her, is Iris Bohnet. And so, she recently published a book, I think it came out in 2016, called, What Works: Gender Equality by Design. And she’s a very prolific speaker. So you can see some of her talks. She did some Google talks and things like that, that are pretty widely available.

But there’s a couple of things that I think are worth really diving into on this topic of improving the pipeline, that she has really highlighted, more broadly than even within our industry. The first is to consider having a more structured process around talent acquisition. And I have to tell you, when I was doing the research for this presentation, this was a real eye-opener for me. Because I have not questioned the validity of the interview process in my history.

I’ve always interviewed people normally. I just ask them questions, we talk about different things. I end up trying to convince them at the end why they should join the firm that I am with, et cetera. And wow, good thing I did this research, because that has been completely debunked.

So if you are doing unstructured interviews to hire talent, I could tell you two things, really importantly. First, is that it doesn’t work. It doesn’t lead to you finding the best talent for your roles. And second, it can actually increase bias. Because people do like to hire people who are like them. And that’s part of unconscious bias, that’s something that we are all — we all suffer from. Men and women alike, whoever is in the hiring position, essentially.

And so, there’s a couple of things that Iris suggested. And then she’s got some great data in her book that really supports this course of action, I wanted to mention to you guys. So specifically, she said, we should not be doing panel interviews. You know, when you have two or three people from your office who are in the room together, having the same interaction with a candidate.

Instead, what we should implement are structured interviews, where you have a list of questions, you ask the same questions to each candidate. And before you go on from one question to the next, you actually jot down, one out of five, how do you think they did?

Because if you don’t do that, then their answer to question three, four, or five ends up coloring what you remember about how they answer question one. And question one might have been the most important, in terms of really predicting job performance. So in order to really force yourself to be honest with how you felt about each candidate, you have to really implement a very structured process.

The second that I thought was fascinating, especially as it could apply to our industry, is that the most efficacious way to see if a candidate would be good at a job is to try to give them some kind of written test that estimates how they would do at the job. And then evaluate those tests without names.

So, without knowing what gender they are, without knowing their background, or anything at all. Just simply give them some kind of a written — writing test, exam, analysis, whatever is appropriate for the role, to really be able to get a sense of what you think they could do. Or whether you think they could do the role.

So these were two things that she specifically called out. There’s a couple of others mentioned here, but those, I think, the data on them is very supportive. And in fact, once I started digging deeper into this — and I think Brad might touch on this as well — essentially, all the social scientists have completely debunked unstructured interviews. So I was like, wow, I’m really behind the times. So if you are as behind the times as me, we can all catch up together.

[INAUDIBLE]

OK. All right. That definitely means Roger must be liking what I’m saying, or he’d be like, OK,

[INAUDIBLE].

So, the other thing I just wanted to mention on here is mentoring. And the reason why I wanted to mention mentoring is because I am someone who has not ever really had a formal mentor. I would say I’ve certainly sought input and advice from lots of people in the industry at different points in my career. But I never feel like I have participated in a formal mentoring program.

And so, as a result, I personally haven’t been a big advocate for them because I see the challenges. I don’t see the opportunities. I think it’s really hard to match people together properly. It’s really difficult to make sure you’re getting the right chemistry, and lead to, I would say, pairings that necessarily lead to good results.

However, in doing the research on this project, there are multiple studies that corroborate that women don’t ask for mentoring. And actually, the senior white men in our industry would actually love to be mentors, but they aren’t going to reach out to the younger women and minorities in your firm.

So I was actually very surprised to learn, also, that mentoring in a very establishing, structured mentorship program is critically important to helping advance women in our industry, as well. In fact, according to that Harvard Business Review report I mentioned before, formal mentoring programs boost representation of women and minorities by 9% to 24% on average. And this is over a five-year period, as all of the data in that study were.

The other thing I’d mentioned too, I just want to reiterate what I said about — from Scott Page’s book. That you cannot hire the quote, “best people if you want the best team, you have to hire the people who will clearly add to the cognitive diversity of the group.” And I think that is another way to really help us when we’re thinking about improving the pipeline.

So the three pivot points — actually, the idea of talking about the three pivot points in a woman’s career, specifically, but really, in anybody’s career — came from Sally Blount. And Sally is pretty much the only woman dean of a business school — of a top 10 business school right now. She’s the dean of Kellogg’s. And unfortunately, she just, a couple of weeks ago, announced that she’s stepping down.

But before she did that, in March, she wrote this fascinating article — if you haven’t read it, I strongly recommend it — called, “Getting More Women into the C-Suite Means Keeping Them in the Talent Pipeline.” Now, I was a little hesitant to include this topic, only because Sally did not do a formal academic study. So all the other data points and ideas and thoughts I’m mentioning really came from evidence-based research that I think was conducted very thoroughly.

But I thought what Sally had to say resonated with me personally so much that it was worth, at least, discussing briefly with all of you. So now you’re going to tell me when three minutes are up. Because now the timer is not going anywhere.

So the first is launch. So, essentially, there are multiple studies that show that women start their career making 80% of what men make. And that’s partially because we take different roles. It’s partially because we get paid less for the roles that we do take. And starting your career on the right foot makes a huge difference for your future career trajectory.

The second is mid-career. So this is really the point at which those women who choose to have children are going through that phase of their life. And how do we, as an industry — specifically, business since Sally was talking from the Kellogg business school perspective — but how do we make sure that we’re supporting women to come back to work? How do we make sure we’re making it inviting and really supporting that very difficult transitional period?

So she actually cited a study that — in 2004, there was a study done of women who left work to have children. 93% wanted to return to work, but less than 75% managed to do so. And only 40% returned full-time. So that shows that there are some material failings in the world of business that we could be doing — that we could be changing, in order to attract more women.

And specifically, she cited we could look to countries that have done a better job supporting women with childcare, with elder care. And also, making sure that we’re offering flexible work hours and career paths. I think we can all agree we have the huge benefit of being in an industry where we can structure career paths that can be more flexible than in some industries. And then she also specifically cited mentoring as well. So another data point in favor of mentoring.

And then finally, executive transition. When women are moving from a leadership role into the C-suite, or into a strong executive role, Sally cited a couple of studies where, essentially, women are opting out. And this is partially because women are less likely — according to these studies — to be driven by money, as opposed to be driven by wanting to make a difference in the world. Or work for organizations where they think their time is better spent, in terms of contributing to society.

And then there’s also a mention though, that I think is an important one, about women being overlooked for C-suite roles. And how many times do you really want to be overlooked before you are — before you say, you know what, I’m done trying this, time to move on to my non-executive director career. Or something like that.

So the solution, with regard to that executive transition phase really, which I think is a great one, is strategic talent reviews. And I actually, I think this is a critical thing to insist that your companies are doing. Not just on people who are already in the C-suite, but two or three layers down. How are we helping make sure that the leaders in our company, and the future leaders in our company, know what their potential career paths look like in our firms? So that they stay with our firms, so we can nurture them, and help them grow into those executive roles.

All right, now I’m out of time on my most controversial slide, but this is it. So I think there’s a lot of debate on quotas. And, like mentoring, quotas are something that I have had a very diverse set of thoughts on, even within my own mind. And I just want to quickly share with you all that there’s a fascinating study that I came across in this research about quotas in India.

And I actually, I think, is this one you’re going to touch on? OK, so I’m going to stop here because Brad’s going to talk about it. But, suffice it to say that there is mixed evidence on quotas. But I did want to make the case for considering, at least, some kind of target or quota in order to be provocative, so we can have a good debate. But I’m going to leave all the details around that to Brad, because I’m out of time. And —

[LAUGHING]

— and I’m sure he’ll do a great job.

So in conclusion, I just — I do think there’s a lot of things we can do. I mentioned a few here. Specifically, really just recapping some of the things that these studies talked about. Engagement, exposure, improving transparency, improving our pipeline, and making sure we’re supporting women as they go through all of the transitions in their career. So, with that, I’ll stop.

ROGER: Heather, great job.

[APPLAUSE]

I thought that was a really well-crafted piece. So, delighted to welcome Brad to segue in. And I think you’ve given him a number of openings, and looking forward to his session. And a little bit of time for questions right at the end.

BRAD BARBER: Great. Thanks, Roger. And Heather, that was really good. You tossed me the hot potato, right?

HEATHER BRILLIANT: Indeed.

BRAD BARBER: So I am really thrilled to be here. I started work with the CFA Institute about a year ago with my longtime co-author Terry Odean and Renee Adams. Terry’s at UC Berkeley. And Terry and I have done a lot of work in behavioral finance.

This I would call behavioral research, as well as trying to understand why we have so few women in the finance profession. I’m an academic, and I can tell you, even in the academic arena, when I go to conferences and asset pricing or investment management, the number of women in the room is usually the reverse of what we see here. The women are sprinkled throughout the room. And it really is a situation that we’re thinking about carefully in academia, as well.

And so, I consider myself a scientist, a social scientist. But still a scientist. And so, it really is a provocative question. Why do we not see more women in this profession? And I think there’s lots of mechanisms at play. And when I was asked to talk today, I thought, well, what I really want to lay out to you is just the basic facts.

I’ll give you some facts about where we stand today. And some facts about what we’ve learned from experimental and empirical research in this field. So I’m not going to be able to solve this problem for you. It’ll be a problem that I think will take a long time to figure out the mechanisms at play, and how we can address it. But I hope to give you some facts that will arm you going forward.

I also want to give a particular shout-out to Rebecca Fender, who has really been great to work with. And she’s really advanced this initiative for the CFA Institute, through my eyes, at least. So, Rebecca, thank you very much for all you’ve done.

[APPLAUSE]

So when we talk about women, in general, I’m going to structure the talk around four basic things. I’m going to give you basic facts about women in the workforce. Because finance is just one part of the problem. And I think we need to understand broadly, why women are participating in certain occupations and why they’re participating in the workforce at all.

The second point I’m going to make is something that dovetails with what Heather was talking about. There’s lots of research now that suggest role models are particularly important for girls and young women as they enter the workforce.

The third point I’m going to make is there’s a funny thing going on with math. And this relates to the participation of women in STEM, in general. So finance, I think, you can kind of put into the STEM problem. It’s a broader problem in our society, that we do not have a lot of women in STEM-related fields, and finance is one of those STEM-related fields. And finally, I’ll throw out a proposal to you about some of the things that I think the CFA Institute and the industry can be doing to try to advance some of these issues.

So the basic facts — this is a graph from a really excellent review paper that just came out in 2017, in the Journal of Economic Literature, by Francine Blau and Lawrence Kahn. And they just survey the entire literature on gender pay inequality. And a couple of facts that come out here, just so you’re aware, one of the biggest shifts we’ve had in labor markets over the last 50 years is the participation of women in the labor market.

And so this graph goes from 1947 to 2012. Back in 1947, 30% of women participated in the labor force. It was nearly 90% of men. You can see men have declined their labor force participation over the last 50 years. Women have increased it. But one of the interesting points, which social scientists don’t quite understand, is women participation stalled about 1990, and hasn’t increased since then. And so, we really don’t understand why that stall has occurred.

The other thing you’ll notice is, since 2008, the financial crisis, participation rates have gone down. So I’m sure you heard about that in the past. That’s true for both men and women. So we’ve kind of stalled here, in terms of female labor force participation. Even though we’ve come from dismally low points to a much better place.

The other big point that comes out of this is the pay gap. I’m sure you’ve heard the $0.60 on the dollar figure for women’s pay disparity. And that comes out in the data back in the 1970s and 1980s. And it’s true that there has been some improvement in that over time. But we’re still at a point where women are paid about $0.80 on the dollar, relative to men.

Now, that’s with no adjustment at all for occupation, education, or experience. And so, you can sort of think about, well, is this because women are going into different fields? Is it because they don’t get a college degree, or graduate degree, that might afford them better earnings? Is it because they have less experience? You can adjust for those things using standard econometric methods.

And what Lawrence and Francine do is, basically, look at four snapshots. So this graph is showing you 1980, 1989, 1998, and 2010. The blue bar is the unadjusted gap that I showed you before. $0.62 on the dollar back in 1980 goes to $0.74, 77.2, and 79.3 in 2010. Now, first point to make there is all of the improvement came from 1980 to 1989. And we stalled again, in terms of the improvement in the pay gap.

Second point to make is the red bars adjust for what they call human capital, which is just education and experience. So you can adjust for education and experience. Now you could argue that that’s also endogenous, that women don’t get as good of educational opportunities or experience opportunities. But that adjustment, if you say it’s leveling the playing field in the extreme, only gets you to $0.71 back in 1980. And it goes to $0.82 — $0.82 from ’89, ’98, and 2010.

The green bars do the full adjustment. In other words, take out industries that you work in, experience, occupation, education. And so that’s a full adjustment. Even with that full adjustment model, you still don’t explain all of the gender pay gap. It’s $0.92 on the dollar. And so this is the basic facts of the gender pay gap.

And the real thing that I want to emphasize here is, we haven’t made much progress over the last 20 years. Those bars look very similar, 1989 to 2010. All of the advancement, if you will, came from 1980 to 1989.

This is just a graph that shows you women in STEM. And so if you look at the y-axis here, you can see that most of these professions are in the 20% to 30% to 40% range, in terms of the proportion of women in STEM. Just to put this in context, about 45% of the total labor force is women in the United States.

And so, in STEM-related fields, particularly engineers, computer workers, even life in physical sciences, we’re at 41%. So below average. You get more social scientists, which is in the STEM field. But that is all psychologists and sociologists, which fall in the social science bin.

Unfortunately, if you break this down — and this graph, I’ll let you look at it. It’s in your slide deck, or in your conference packs, rather. You’ll notice economists down there at the bottom, which is probably closest on this list to finance, they’re at about 30%, 35%. So in those more quantitative social science fields, again, you see women under-represented.

So another interesting thing is, there’s a gender gap in college. And the gender gap in college is the observation that women are graduating at far higher rates than men from college these days. Now, that didn’t used to be the case. If you look from 1870 to 1980 — and this is a work by Claudia Goldin and her co-authors — they show that men were graduating at much higher rates than women from college.

But if you look in the more recent years, this is by birth cohort, by the way. So somebody born in 1980 had nearly a 35% to 40% chance of becoming college educated, for a woman. But you can see, for men, it’s slightly less than 30%. To give you more recent data on this, I pulled the data from the 2015 iPad survey. This is the integrated, post-secondary education data that I pulled.

57% of all majors in 2015 were women, all graduates. This is undergraduates. 47% of business administration majors were women. And 30%, roughly, were finance majors. So you can see that this is part of the pipeline problem here in finances. We’re just not getting women in finance sort of fields. But it’s not all about the pipeline.

If you look at CFA’s now — this is the proportion of women who are CFA members by age — you can see the y-axis peaks out at about 20%. And we’ve got bars for the CFA members in their 20s, 30s, 40s, 50s, and 60s. Yes, you do see some improvements there by age cohort. But not a whole lot.

And more importantly, remember I just said 30% of finance majors were women? We’re down to 20% of CFA members are women. This is consistent with what Paul was saying before, is you get candidates who are women at about the same level as we have finance major. But for some reason, they don’t end up as CFA members or in the profession.

And finally, Heather mentioned the Morningstar stuff, where she is, or was, working. Was. And this dovetails — the CFA experience dovetails with mutual fund managers. This is from the Morningstar survey. And basically, what they find is fund managers by gender — 90% of funds are managed by men, 9.4% by women. So this is even more dismal than the 20% that we’re seeing among CFA members.

And then, if you look at teens, you’ll see that 78% of funds are managed by men only teens. And so there’s really not a lot of women in the asset management business, even when you consider teens. Which is really quite remarkable. Among the occupations within the CFA — just to describe, again, the problem — the red bar here is the 18%, that Paul was alluding to before, of CFA members are women.

And this is just among the different occupations, were sorted from low representation of women to higher representation by occupation. And it’s really quite stunning that anything that has a chief next to its name is not — is less likely to be a woman. We have chief investment officer, chief executive officer, and chief financial officer, all hovering around a dismal 10% of women.

So these are just the basic facts about representation of women in the labor force, in general, in STEM, and how that relates to finance. I think those things are all related.

And so the second point that I wanted to make is about the importance of the role models. Because I think there’s one way in which we can think about trying to shift these numbers over time. It’s not going to happen overnight. It’s by for providing role models for young girls — or girls and young women.

And so, there’s a couple of questions here that social scientists have been grappling with. Which is, do role models differentially affect men and women? And does gender matching of role models matter? In other words, does having a female mentor matter to a young girl, as opposed to a male mentor? And so, I’m going to talk about some studies related to this.

And this is the study that Heather was alluding to before. And actually, there’s a series of study by this co-author team. And it’s Beaman, Duflo, Pande, and Topalova. The last page of my slides, by the way, has full references to all this stuff. So you can look these papers up later. But what this author team, co-author team, studied was a 1993 law in India. And what the law did is it, basically, reserved leadership positions, called pradhans. For women in randomly selected village councils in a part of India.

So imagine that you’re situated in a village, that village has chosen to be — have at least one female leader, or pradhan, in that particular village. Next door, there’s a village that is not assigned that female leadership position. So that random assignment is key. It’s the gold standard in social science to identify causality. It’s the equivalent of a drug test that we do for drug safety and efficacy.

Now this was first implemented in 1998. And the councils were elected every five years. And the authors collected data in 2007, following 1998 in 2003 election. So there’s two election cycles. And so what you get is survey responses from 8,400 adolescents and parents in almost 500 villages.
And the random selection of villages created three conditions. There were villages where, in both the ’98 and 2003 elections, there were no women. They were not randomly assigned to get a female leader. There were some villages where one election got a woman, but the other did not. And there were some villages where, in both the ’98 and 2003 election, they were assigned a female leader. So this is the quotas that Heather was talking about, in terms of what happens.

Now the really exciting part for social scientists is we can now study whether having exogenous female leadership, randomized female leadership, affects what people do. And the answer is, it does. And here was the evidence that came out of that. Let me just talk about the metrics that were measured in the survey, briefly, before giving you the results.

So the parents and adolescents were asked about aspirations. Parents were asked about aspirations for their children. Adolescents were asked about themselves. And they were asked, if they — what’s the highest grade of educational attainment you want for your child or yourself? This is coded as a one if you wanted more than a grade 12 education. What age do you want to be married? Or, do you want your child to be married? It was coded as a one if it was greater than the age of 18.

What’s your preferred occupation at the age of 25? Doctor, engineer, scientist, teacher, or a lawyer was coded as one. And, what’s your desire for a child to become a pradhan, a leader themselves? Coded as one, if yes. And then they asked about educational outcomes. Did you attend school? Can you read and write? And what was the last grade completed? So these are actual educational outcomes for their children.

And finally, they did a time use survey to see how much time was spent on household chores. So think about this girls versus boys. This could be a big indicator of whether there’s changing role models, as a result of this female leadership. And here’s what comes out of that.

So, if you look at the left bar here, just to focus on that to see what’s coming out of this graph, you see parents’ aspirations. And each bar represents the gender gap. So a positive gender gap means that the aspirations for boys exceed those of girls. So basically, when you look at parents’ aspirations for the never reserved councils, which is the gray bar, you can see that they have better — higher aspirations for their male children than their female children.

If you look at the reserved once, you can see that they have very little difference. But when you look at the councils that are reserved twice to have a female leader, all of a sudden, you start to get some movement. The parents have higher aspirations for their female children, relative to their male children. That gender gap starts to close. And that’s the really provocative result.

If you go to the teenagers’ aspirations, this is what they thought of themselves. You can see the same basic result. No reserved councils, the gray bar, versus once reserved, not much difference. But councils where there were two times female leaders, you start to see a movement. Same thing for teachers educate — teenagers’ educational outcomes. Very little difference, or movement, in the gender gap between never and once reserved. But you start to see movement for the twice reserved.

And so, this is very provocative evidence, that with female role models in very prominent positions in the village, you start to see, actually, the people in that village changing how they view women. It takes time, it’s gradual. But you see the movement in the data.

In a separate study, they actually analyzed the same phenomena. And asked the question, well, does this lead to greater participation of women in politics? And what they found was the answer there is yes. So here’s a separate study that they did. And they look at, in the horizontal hash, this is the number of pradhans that you observe in the never reserved, reserved once, and reserved twice.
So the horizontal hash bar is no surprise, because those are assigned. That’s the random assignment. Never reserved, you do see some women pradhans. Reserved once, you see a lot more. And reserved twice, you see even more. But what’s really interesting is, again, this reserve — never reserved, reserved once, and reserved twice — the dotted bar and the solid bar, these are candidates who were not for the pradhan, but for other positions.

These are — are we bringing out women to be leaders in other positions? And you see in the never reserved versus reserved once, not much movements. But in those villages, where they were reserved twice to have a female leader, all of a sudden, you get more female candidates and female winners of other elections. So there is this network effect that seems to be occurring, that draws more women into these positions once you have that random assignment. And I think that’s a really provocative and powerful result.

Another random assignment — I’m going to really harp on random assignment, because that’s really how we learn about these sorts of issues. Another random assignment is the random assignment of female professors to students. And this occurs in the US Air Force. And so, Scott Carrell and Marianne Page, who are colleagues of mine at UC Davis, and their co-author, West, look at the question of, does the gender of professors affect the performance of students?

And they analyze 2001 to 2008 graduates of the US Air Force. And again, the key thing here is that students are randomly assigned to classes in the Air Force. And so, you can look at whether, with that random assignment, does a young woman assigned to a female professor do better than a young woman assigned to a male professor?

And what they find is, among all students, if you look at the female students on the left, who are assigned to female professors, you can see that there is — this as grade point average. So it’s very small differences, but they’re statistically significant, if you will. So the female students are doing slightly worse than the average student in both female and male professors. But the gap is narrower, if you will, when they’re assigned to the female professors.

When you look at, interestingly, the high SAT students — so think of this as about skimming the cream of the students who have high aptitude — very interestingly, what you see is among the female and male students, very similar. Except one assignment.

When women are assigned to a male professor, all of a sudden, those high achieving women are not doing as well. That’s a very interesting and provocative result, again. Because this is about random assignment to classes. And so we’re able to causally identify, is it about them gender matching of these students to professors that matters?

And finally, another random assignment, mentoring and engineering. Remember, I told you engineering is one of the more dismal representation of women as an occupation. And what this study did is it took 150 female students at a public university and randomly assigned them to a female peer mentor, a male peer mentor, or no mentor at all.

And then it asks the question as, does this improve retention of women as engineering majors and their identification as engineers? And they basically conduct a survey of these students on a seven-point Likert scale. One if not at all true, and seven if very true.

They ask questions about belonging, like, I felt connected to my careers in engineering. Or self-efficacy. Do you think you have a talent for in engineering? And then they measure retention. Did they stay as an engineering major after the year of study?

And so, what you see here, interestingly, is this is the change in their measure of belonging. The black bar is female mentoring. The dashed — large dashed bar is male mentoring. And the dotted bar is control. And what you see is that, with female mentoring, there was very steady sense of belonging, sense of self-efficacy.

But with a male mentor, or no mentor, you see a decline in that. So imagine that you’re a woman coming into an engineering major, where 80% of the student body is male. It would be very easy to feel displaced in that major. And I would posit to you, it probably feels very similar in a finance class at many undergraduate institutions. That’s where this female mentoring, I think, is really powerful, and can help improve women — young women’s sense of belonging in the profession.

They also looked at retention. And their retention is, are you likely to stay in the major? Now, most people stick with their major, despite adversity. And so, you can see here that, even in the control and the male mentoring conditions, 80% to 90% stayed in their major.

But in the female mentoring condition, you do get reliably higher retention rates at 95%. And so, there’s evidence here that, both your sense of belonging, your sense of self-efficacy, and your willingness to stick with the major even though you’re among the minority, is improved with that mentoring condition.

And finally, Terry, Renee, and I have this work where we looked at — along with the CFA Institute via survey — whether women CFAs and male CFAs are more or less likely to have a STEM parent. And this is an interesting way, because parents are probably our most provocative role models.

And the questions we were interested in is, do STEM parents close the gender gap in finance? And if so, what’s the mechanism? Is it a role model or math training mechanism? And what we found is that this is the proportion of CFA members who have a STEM dad or STEM mom by gender.

So if you look at the proportion of male CFA members, 34.6% have a STEM dad. 44.5% of CFA women had a STEM dad. You look at the STEM mom, you get a similar pattern. Although, the bars are lower because there aren’t a lot of STEM moms out there for our current generation. But 10.8% of the men had a STEM mom. 16% of CFA women had a STEM mom.

So this clearly shows that women in the profession, and CFAs, are more likely to have had some sort of parental role model who worked in science technology-oriented degree. We can actually do a little bit of Bayesian math on this, and think about, well, if I pull somebody out of the random population, what’s the probability — how much does it increase the probability that they will become a CFA member if they have a STEM dad or mom?

And it turns out, it’s just the ratio of those two probabilities, or bars, that I showed you. We, in the paper, refer to it as the probability impact, if you will. And so, this is the relative impact that having a STEM father has on a young girl versus a boy. And it’s a 28.6% greater effect on a young girl versus a young boy.

If you look at STEM mothers, it’s even bigger. It’s almost double, 47.6%. So even though having a STEM mother is a rare occurrence, it has a much bigger effect on a young girl than it does on a young boy. And actually, we see very similar things among sisters and brothers.

So having a STEM sister is much more common for somebody who is a CFA member and a woman, as opposed to a man. And having a STEM brother is much more common for CFA women versus men. And so, you see a very similar pattern there, as well. I think this is probably closer to home and suggestive of the role model evidence that I presented to you before. This is not quite as causal, but is certainly provocative evidence that role models matter, as well.

So the last thing I want to talk about, just to give you some other basic facts, is what’s math got to do with it. I think we all are aware that finance, in general, is a fairly math intensive training program. And also a fairly math intensive profession. And so one of the things that comes up in this is, how is math related to it?

And just to give you the basic facts, this is from the Program for International Student Assessment, the PISA study, which is an ongoing assessment done in many countries. It’s about 80 countries now. And it’s a standardized test provided to students in all of the countries within their study set. And we have all the countries across the horizontal axis here.

And what you’re looking at on the dots is the math gender gap by country. And so, one observation that social scientists have made, of course, is that their boys tend to do better on these math scores than the majority of countries. If you look across there, it’s about 80% of countries have a math gender gap, where boys are doing better than girls. Actually, in verbal reasoning, it’s the opposite.

But this math gender gap has been one of the — sort of, focus of trying to understand what it is.

Let me be very clear. I think boys and girls have equal ability at math. This is actually evidence that there’s a cultural aspect to this. Because you would not expect to observe cross-geography or cross-country variation in the math gender gap, unless it was culturally driven. What we do not understand is, what are those cultural drivers of the difference? And that’s where the social science is really focusing on.

By the way, the other dots here are — the triangle, if you will, is what happens at the 90th percentile in the gender gap. So another aspect of the math gender gap that I want to emphasize is, it’s not all about the mean. But if you look at the 90th percentile, the math gender gap grows.

You’ll notice the triangles tend to be higher than the circles. And that’s because, at the 90th percentile, we have a bigger math gender gap. And when you translate that into people who will be available for the best and the brightest jobs, that means there’s a lot of men at the top of that math gender gap. Or boys at the gender gap pool, if you will.

So again, just to emphasize, geographic variation in the math gender gap indicates culture, and not biology, as the driver of the variation. And the questions that that arises is, what are the cultural drivers? And does the math gap predict occupational outcomes? And if yes, is the correlation because of a causal effect?

In other words, a direct result of the math training, or is it because of a common cultural factor? In other words, do gender stereotypes jointly drive the differences in the math gender gap in labor outcomes? Or, is this about attitudes about competition, for example, that might drive differences in math participation, or the math gender gap in occupational outcomes?

Let me just say that gender stereotypes — there’s really compelling evidence that gender stereotypes are part of the explanation for why we observe some of these differences in boys and girls. And so, one study that I really like to highlight is by Reuben, Sapienza, and Zingales. And they asked subjects to sum sets of four two-digit numbers over four minutes.

So just be — you’re just summing numbers over four minutes. Your score is how many you do over four minutes. Candidates, one man and one woman, are randomly selected as, basically, candidates for a job. And then the remaining subjects are employers asked to hire one of the two candidates for a math test.

And there’s three treatments. No information about the candidates, that’s kind of a control treatment. Cheap talk, candidates can talk up their performance without any verification. And then past performances revealed. The employer, prospective employer, gets your score from the four minute test. And this is what we observe in that very controlled environment.

First of all, with no information, about a third of the candidates who are hired are women. When we get cheap talk, fortunately, it doesn’t move in the wrong direction. But it doesn’t move at all. Still, about a third of the candidates that are hired are women.

If you go to the far right panel, when past performance is revealed, you would think that would eliminate the gender bias, right? This is a simple task. But still, men are more likely to be hired than women. So we’re hiring 43% women in this controlled experiment. But when past performance is revealed, it doesn’t eliminate the bias. Helps, but doesn’t eliminate it.

Now, what’s going on in the bottom two bars? This is the percentage of low performers that are hired. These are people who are really bad at adding numbers. Interestingly, only 45% of low performers are hired in the no information set. And the authors conclude that we can actually get some cues from appearance of presentation when we interview — when we just look at people.

There’s no talking going on. They just — they just see the subjects. So there is evidence that there’s some information in appearance.

But when we look at the bottom bar, conditional on being a low performer, you’re much more likely to have been a man that was hired. Let me say that again. If you’re a low performer, you’re much more likely to be a man that was hired. So this is, again, evidence that men are being viewed as having better ability for whatever reason, despite the fact that they don’t.

When we go to the cheap talk condition, look what happens to the conditional probability of hiring a man. Men are really good —  low performing men are really good at talking up a good story. 92% of the low performers that were hired were men. And then if we go the past performance condition, low performers are much less likely to be hired.

But still conditional on hiring a low performer, it’s a man that’s hired. And so, this is evidence of gender stereotypes, with respect to math, that I think is very provocative. And suggests that stereotypes are part of the issue here.

I encourage all of you to go to Project Implicit, it’s a website. You can take these things online. It’s a great way to have some of your — you can think about incorporating them in your workplace to see what implicit bias are. It’s very revealing. Women have this bias just as much as men have this bias. And that’s a really startling thing, I think, to a lot of people is, even though you’re part of a class where you feel that you are discriminated against, or there’s implicit bias against you, you also have that bias.

What this study did, is it used that implicit bias test back to the experiment, where we’re asking people to hire people for a simple math test, adding up the two digit numbers. They gave the implicit bias test to all of their subjects. And what they found was, people who registered a higher implicit bias were, of course, more likely to hire men than women. And so, this is all consistent with implicit bias being related to the patterns that we saw before.

Along the horizontal axis here, we’re getting the employers’ implicited bias score in this experiment. And the red line is showing you there’s a positive relationship between your expectations for men versus women. The y-axis, and what your implicit bias was in the simple computer game. And so again, it’s showing these implicit biases, or implicit stereotypes, are feeding into our choice decisions in a hiring setting, in this experimental hiring setting.

This is a graph from the work that Renee and Terry and I are doing. And it’s a scatterplot of countries. And the y-axis here is the percent of female CFA members. And the x-axis is the math gender gap at the country. And you can see here, most of the math gender gaps at the 75th percentile, are boys doing better than girls.

And you can see here that, in the data, occupational outcomes and the math gender gap are quite correlated. And so, this suggests that there’s something about the math gender gap that’s modulating career outcomes. We do, in a regression framework, we regress the proportion of women who are CFA members on the math gender gap, the percentage of women who are in the labor force in a particular country, and the gender inequality index.

And we still find that there’s a very large negative effect of the math gender gap. These are standardized variables. So one standard deviation move in the math gender gap means that there’s seven percentage points less women in that country. So that’s a fairly large effect.

And more importantly, gender inequality is negatively related to women’s representation in finance. But it’s not explaining the math gender gap channel. And total labor force participation is also related. But it’s also not explaining, if you will, or taking out this math gender gap example. Which is, economically, the biggest effect that we’re finding in the data.

ROGER: Brad, two minutes if you can.

BRAD BARBER: Yeah. Sounds good.

So this is just showing math gender — finance in the math gender gap across states. That should read across states. So one thing we did to test the robustness of this is look across states within the United States. And we see the same correlation between the math gender gap in the proportion of women who are working in finance within the state.

It’s really quite a stunning and remarkable result, because you wouldn’t expect that much variation across states within the US. But in fact, you observe much the same result. So, in deference to time, I’m going to leave some of this stuff on competition out. And I’m going to go to the la —

[INAUDIBLE]

— which is the proposal. And my proposal to you is, remember what I’m show — what we know. Big differences in how men and women sort into occupations. Finance is just one of those occupations. We know that role models matter greatly for outcomes. And we know there’s this math channel going on, as well. We don’t fully understand the math channel. Whether it’s a cultural factor that’s correlated with math, or it’s a direct effect of math.

But the proposal is, threefold, if you will. Mentorship programs for women. I think the CFA and industry representatives, in particular, are in a really unique position to start thoughtful mentorship programs. Structured mentorship programs would be really helpful. And so, think about ways to develop, coordinate, and most importantly, test the effectiveness of them.

As a social scientist, what I would really like to see is partnerships between industry and academia. To understand, what are the key features that lead to effective outcomes? In other words, we need to have clear goals. What is that we’re trying to do? Are we trying to increase retention rates in majors? Are we trying to increase the proportion of finance majors who take the CFA? What is the goal that the mentorship program has?

And then we could, in fact, have random assignment to some of these mentorship programs. And actually measure outcomes, see what features of the mentorship programs actually lead to greater participation by girls versus boys, or young women versus young men. And I think CFA and industry organizations can really help coordinate that effort, with both resources and commitment.

And I know I’m just one scholar. There’s many scholars working in this area that would really welcome the opportunity to partner with both the CFA Institute or industry, to think about ways to design these studies, and see if they’re effective. I think one of the things that I tried to get across in my talk is, to learn more about this, we need random assignment, and ways in which we can identify the causal factors driving how people are making their choices.

ROGER: Brad, we’re going to have to finish.

BRAD BARBER: OK. Last thing I want to say is about affectfinance.org. And this is financial female — female financial economist organization. And they provide a directory of women financial economists. They encourage women as discussants, invited speakers, and panelists. And they conduct research on women financial economists.

I think those directories, and having those available, and encouraging the industry to have female panelists, discussants, and speakers, is a great way of just integrating role models into the profession. So, thanks very much.

ROGER: Brad, thanks so much.

[APPLAUSE]

So we have managed to reach the perfect time of the break. Don’t how we did that. But I’m going to renegotiate. [INAUDIBLE] Definitely ran over time. So can we have five minutes? I’m looking six, seven?

SPEAKER 1: We have five.

ROGER: We have five. Great, OK. Well, I had three topics. We’re probably going to crunch them to two. As ever, a ton of really good questions. You can’t have better answers without better questions. The ones that I wanted to first put to Heather were to do with quotas. So, several questions were really challenging. How you do quotas effectively, and what you mean by succeeding with quotas.

HEATHER BRILLIANT: So, I think — look, I personally, believe that if you say, quota, you mean you will actually require a certain number of roles to be filled by women. As opposed to a target or trying to more — casually move towards a specific number. And I actually think the study that Brad very eloquently described in India shows that quotas can work.

And the thing that I don’t think he necessarily mentioned too explicitly, but is covered in those studies, is that, in some cases, women were randomly assigned to leave the village. They had no experience. They had no knowledge of what they were doing. And they did just fine to great. In fact, the ones who made it to a second term were considered to be as competent as men in similar villages, or similar situations.

And the other thing I just want to mention about this, because I’m actually not meaning to advocate for quotas, but given the way your question came out, I just feel the need to, at least, put it out there for food for thought. That in fact, the likelihood of a woman speaking at a village meeting increased by 25% in the villages where there was a woman leader.

And so, it just goes to show, I think, how the impact can reverberate through society by getting people over the hump of thinking that women are somehow less capable of taking on roles where, I think, we can prove over time that that isn’t the case.

ROGER: So, subject number two, mentors and role models. So, Brad, a ton of interesting stats. I loved your Bayesian stuff. But on the role model, any kind of final thoughts on what that might look like? And same question to Heather.

BRAD BARBER: Yes, so I mean, there’s a lot of research on mentoring young women in college. And so, I think the industry CFA members — if they could develop some sort of formal mentoring program of women as they enter college, freshmans who have decided they want to pursue a degree in finance. For example, the engineering study that I cited speaks directly to that.

But I also think the smaller things, which would be difficult to measure in a scientific way, but I suspect are also impactful — which is having women, who are professionals, talking in middle school, or in younger ages, about their careers in finance, would also be encouraging to young girls.

That’s a much harder thing to measure, because you’d have to measure outcomes 10 to 20 years later. But again, I’m willing to go out on a limb and say that I suspect that those things would be impactful as well.

ROGER: And is the structured program feasible, is quite interesting. What — maybe, Heather, you could pick that up. Formal or informal is kind of the interesting dimension of this subject.

HEATHER BRILLIANT: Well, I think — my own personal opinion on mentoring now is that perfect is the enemy of the good. And so, it’s better to give mentoring programs a try, even if we can’t necessarily perfectly match everyone together. And even if not, all of those pairings will work out. The ones that do work out can be hugely impactful. And so, I think different CFA societies, different organizations, can come up with, I think, a variety of different ways to solve this problem. But I think giving it a try is the most important aspect.

ROGER: So, last quick question, and it’s a three liner from each of you, if you will. Cognitive diversity in teams. How do we organize that? Heather.

HEATHER BRILLIANT: OK, well, first of all, I’m only halfway through Scott Page’s latest book. But he actually has really good recommendations on it. And I think we should all read his book. Because it’s so empowering, in terms of really thinking about cognitive diversity, in terms of, literally, what types of problems are you trying to solve.

So, not even necessarily from a permanently assembled team, but to say, OK, we have this group of people in our organization. This is the problem we’re trying to solve. Let’s pull people from different parts of the organization in order to compile a team, in order to address that exact problem or challenge.

And so, I think, trying to be really specific about — we need somebody — he gave a great example, actually, of thinking through solving the issue of obesity. And you might think, oh, well we need a dietitian. Yes, that’s great. But what about the social impact of people who don’t walk to school anymore? And how they are driven to school? So that — there’s all different angles on the challenge of obesity. I mean, literally, he probably listed 15 of them in a matter of two sentences.

So I think, if you really start thinking about some of the more complex problems we’re trying to solve, you can certainly break them down into the types of skills that you need to be able to solve those. And I think, compile teams that bring that.

ROGER: Scott Page, “The Diversity Bonus” is the book. One liner from you, Brad.

BRAD BARBER: Celebrate the difference.

ROGER: Celebrate difference. OK. Three last thoughts. Best teams, not best people. Interesting stretch. Research and fact, definitely ahead of opinions and the feeling in this area. And it’s really research-driven strategies for making progress with this gender diversity thing, and first-class look at what those strategies might look like. Thank you very much to everyone.

[APPLAUSE]