Iris Bohnet, Professor and Academic Dean of Harvard Kennedy School

MS. ROSENWORCEL: Welcome to another episode of Broadband Conversations, a podcast where you'll hear from women working across the technology, innovation, and media sectors. We'll talk about what they're working on, what's on their minds, and what they think is next for the future.

I'm Jessica Rosenworcel, a Commissioner at the Federal Communications Commission. And if you are new to this podcast, welcome. If you are an avid listener or anything like it, welcome back.

My guest today is Harvard professor Iris Bohnet. She is the academic dean of the Harvard Kennedy School and the co-director of the Women and Public Policy Program where she is leading a research portfolio on gender and technology.

She also has written an award-winning book, "What Works: Gender Equality by Design." I'll let her tell you a little more about her work. And, in the meantime, welcome, Iris, and thank you for joining me.

MS. BOHNET: Thank you very much for having me, Commissioner.

MS. ROSENWORCEL: All right. Well, let's get started with a little bit of back story. So, tell us just a bit about how you got to where you are today.

MS. BOHNET: That's a big question. But I started out studying economics and have always been interested in learning about tools that helped me understand how the world works, how society works, understanding some of the big challenges in the world. And I've always been interested in gender, but I did not start out as a gender scholar.

So, I studied economics and then I became interested in combining economics and psychology, which is now called behavioral economics. It's a bit of newer field, although now it's maybe 20-years-old, that is combining tools from both of these disciplines.

And I happened to have been an economics then for maybe the first 10 or 15 years of my career, and in the last ten years I became really interested in questions of gender equality. And I think I became interested for two reasons.

One was that I've always been in a feminist at heart that really cared about questions of equality. But maybe equally as importantly, I felt that I had some tools available that weren't currently used in the diversity, inclusion, and equality space and I wanted to employ those tools. And that's the book that you just mentioned, using insights from behavioral science to help us close gender gaps.

MS. ROSENWORCEL: That sounds good. I like the idea that you're bringing the economics discipline to think about gender and equality. So, this might seem like a completely basic question, but it's so fundamental I want to start with it.

Why does having more women matter to technology and innovation?

MS. BOHNET: I think there are many good reasons for that. Let me maybe start first just with the talent pool. So, the tech sector is growing rapidly, obviously, as we all know, influencing almost every if not every aspect of our lives, society, organizations.

And, clearly, organizations, whether that is the public sector, or the private sector or civil society want to attract the very best minds to work on these projects. And so, we just need to be able to draw from one hundred percent of the talent pool. So, therefore, I think it is no question there that we need and want to have more women and more people, just more generally, more women then included.

But let me give you maybe two more thoughts. A second one is related to the diversity of groups. And there is the research, for example, coming out of MIT, the collective intelligence of teams, which shows very clearly that diverse teams, including -- and specifically gender diverse teams, perform better than homogenous teams.

And whether that is all women, or all men doesn't actually matter. It really is the diversity that makes a difference, and come up with more innovative solutions, for example, -- and are more productive and more creative. So, I think there is this, maybe, a more narrow definition of the business case as well.

And then, thirdly, many organizations do point at our customers and how an increasing number of consumers is, in fact, female and that women are responsible for a lot of our consumer spending.

I think some sources say it's like three-quarters. I don't know whether that number is the exact, right number, but certainly women do a lot of the shopping and make a lot of those household decisions. And technology is an increasingly important part of those types of products that we might want to use.

And so, we need to design products that also appeal to women and we, therefore, want to have more women on our design teams. And it has, just to give you maybe a more dire example, we, of course, have had historical examples where, for example, we ran clinical trials, so that's medicine, only on male subject pools to only realize that male and female bodies are not, you know, created equally.

And yet it can be as simple but also as impactful as just having more women in our subject pools, in our focus groups to better understand what kinds of technologies appeal to women, what kind of technology women need for their lives.

So, I think all three of those are important. So, it is the consumer, it is the workforce, and it also is the employee, the talent pool.

MS. ROSENWORCEL: Yes. And I like something that you said early on, which is we need to think about technology as an input to every aspect of modern life, civic and commercial, and not just think about it as a sector or a silo on the side with its own unique set of products or workers. It is really throughout the economy.

MS. BOHNET: Oh, I could not agree more. I mean, I have recently been working a bit more with the financial sector and it almost feels as if we have more technologies working in banks than bankers these days. And that might actually be a true statement. This is not empirical sounding, but it certainly feels that way.

So, I completely agree, technology is everywhere.

MS. ROSENWORCEL: So, when I travel, because I get to do some of that in my work, and I meet with technology companies that are working on everything from spectrum to software, and some are big, and some are small, but something strikes me everywhere I go. There aren't that many women in the rooms that I travel to.

So that's anecdotal and that's my experience. But I know you've done research on gender and technology, so you can tell me a little bit more about that. And, also, what are the goals of that research, what do you hope to achieve?

MS. BOHNET: So that research actually came about because Melinda Gates had heard about my book, "What Works," that you mentioned before, and was intrigued by the evidence-based approach in the book. So, what I'm trying to do in the book is (inaudible) and form our decisions in our workplaces.

And the book is not specifically focused on technology but more generally on kind of the workplace, wherever that might be. But what it does is it focuses on how we can see bias, our practice and procedures in hiring, promotion, performance appraisals.

And that I think appeals to her, thinking that we should move beyond diversity training and, in fact, fix our systems rather than trying to fix mindsets, and that that would work in the finance sector, in the public sector, and in the tech sector, wherever we might go.

And that's really what we have been doing, so that's the research that we are leading, that we work both with tech companies, so we'll start with the big tech companies, to help them level the playing field for men and women. De-biasing everything that we kind of commonly think of as power management, but also beyond formal procedures and getting into questions of culture and inclusion and how we organize our meetings, and who gets which work assignments, which opportunities, et cetera.

MS. ROSENWORCEL: So, what --

MS. BOHNET: So that's one aspect.

MS. ROSENWORCEL: -- you're talking about --

MS. BOHNET: Go ahead.

MS. ROSENWORCEL: What you're really talking about is how do we take this data that we know exists and then implement it in the real world and in the workplace. You know, once you get past strictly making everyone sit down for some coffee and have diversity training, what does it look like, what's effective, and does your research have some data that reflect what really and truly works?

MS. BOHNET: Yes. That's exactly what we're trying to do. And I might also tell our listeners how we do that. We typically run experiments, and that is that we learn from the natural sciences, and run clinical trials in organizations where, for example, we usually -- a company asks me whether they should, in fact, work with a tech startup which has developed an algorithm, a hiring algorithm, which they are using which would increase diversity of hiring.

And so, I get a call from this company and they said, should we use them? And I said, you know, I know the founders, I don't know the scientists -- it all sounds great -- but you will not know unless you measure. So why don't you run your traditional hiring scheme, whatever you currently do, and maybe you do a CV evaluation, maybe have some other tests that you apply.

Do what you always do, but in parallel also have all of your applicants go through this other tool and then you see what the outcome would have been if you had just used your traditional mechanism as compared to what the outcome would be if you used the new mechanism.

And that's the work that we do, so that's also my research but this --- my particular example was a real application, I didn't actually study this. But that's how we measure what the outcome of a particular tool is. And that is maybe my biggest message here, that we have to do a much better job, not just, you know, dreaming about the kinds of things that could work but, in fact, measure their impact.

And so, what does work is a much more optimistic about things that focus on systems rather than mindsets. So, for example, many organizations, not all, and I have to be very clear here, but many organizations have seen future benefits from blinding themselves to the name on resumes, so making sure that they're not influenced by, for example, racial or gender stereotypes that go along with -- for example, thinking of an engineer, we're more likely to think of a man rather than a woman. The blinding technique can be effective.

MS. ROSENWORCEL: So, it's kind of like when they talk about auditions for an orchestra and --

MS. BOHNET: Yes.

MS. ROSENWORCEL: -- they strictly make you listen and not actually look at the performer so that you just truly evaluate their competence for the job without letting those inherent biases we all have filter into the evaluation.

MS. BOHNET: Yes. That's exactly right. And -- but you can do many more things. So, the good news about technology is it is a sector where we have a particularly challenging entitlement for women's advancement, but on the other hand technology is also a part of the solution.

So, for example, people have those algorithms which help us de-box the language in our job advertisements, but also in our performance appraisals, so anything that we right. And many, many companies now use these algorithms to understand how their job descriptions will appeal to men and women, respectively. So that's another tool for which we have pretty good evidence that it's actually working well.

MS. ROSENWORCEL: So, what has really surprised you in your research? Have you had any ah-ha moments with all that you've studied and done?

MS. BOHNET: It's a very good question. Maybe the biggest surprise was how complex it is to work with big organizations on those types of interventions. So, one would think we just need, for example, the promotion data and then we can see whether there are gender gaps in promotions, controlling for people's performance, and maybe we could introduce a fix for this.

What surprised me is how -- how shall I say this now, how unequal data collection, the quality of the data is in many, many big organizations. That for many companies it is impossible to recreate performance appraisal data for the last five years because they just haven't collected it in any systematic way or haven't collected the promotion data in any systematic way.

So, we can't actually tell them whether given the pool of available talent that they have and could have promoted, whether they have a gender gap or not. That was, as I say, I'm an economist by training, that was a big surprise for me.

MS. ROSENWORCEL: So, is that because so much of those promotions are about, you know, vague notions of potential or corporate fit and things like that, and it doesn't get recorded in some data-centric way, or is that just a practice that they don't spend a lot of time on?

MS. BOHNET: Both. So, I think traditionally, sadly, the rigor that we might have applied in our engineering departments or in our finance department, or even our marketing department have not always applied to our HR departments. So, I think partly we just have not been focusing on our employees' data with the same kind of scrutiny that we might have focused on our customers' data.

So, I think that is one of the reasons. I think a second one is also one that you have just mentioned, and that is that some of the procedures were very interest-bearing, very informal and very objective. And so, we don't have all the information on what's happened.

And then, thirdly, coming back to technology, and that's probably not going to surprise you, but it surprised me initially, but many of the big companies are still, how shall I say this, digging their way out of having a thousand different systems designed in different places, different parts of the world, not really communicating, speaking with each other. So that's also just a big challenge.

MS. ROSENWORCEL: I know, digitization in our largest organizations hasn't fully occurred. All of our systems tend not to talk to one another. Sitting here in the government, I can tell you there's some truth in that.

MS. BOHNET: Yes, right. I can imagine.

MS. ROSENWORCEL: All right. So, I like to close things out by asking everyone a few questions at the end. The first is what's the first thing that you did on the Internet?

MS. BOHNET: That is such a good question. I don't know what the first thing was that I did on the Internet. I have to say Google probably was early on in my life, kind of Google searches, looking for things. I don't know whether that was my very first thing.

But I do remember, probably not, so I don't know what my first thing was. But I have this vivid memory of thinking what amazing search tool, search tools we now have available and just look for things, plucking at words. So, I do think the search engines, yes, whether it's Google, et cetera, and they have been quite important early on for me.

MS. ROSENWORCEL: Absolutely. It's revolutionary what you can access right now. So, what's the last thing you just did on the Internet? Hopefully, memory serves better with that?

MS. BOHNET: So, I'm serving as the academic dean now --

MS. ROSENWORCEL: Congratulations.

MS. BOHNET: -- so my job here at Harvard is faculty hiring, faculty demotion, et cetera. And so, I now find myself a lot, including this morning, looking at people because there's so much information now available on the Internet.

You know, plugging in a name of somebody who is at some other university that some colleague or some research paper had alerted me to, and finding out more about that person -- like wow, maybe we should try and attract this person to Harvard. So, yes, that's what I've been doing.

MS. ROSENWORCEL: So, look at that, that's your academic research and your managerial responsibilities coming together.

MS. BOHNET: Yes.

MS. ROSENWORCEL: All right. So now a little more thinking about the future. What do you think and hope the future of the Internet and digital life looks like?

MS. BOHNET: So, I mean there's lots of hope to have, but I'm going to go back to a bit of our early discussion. I hope we will get a better grasp of and grip on the kind of bias on the Internet, harassment in the digital world. I'm quite involved in research and research communities and algorithm bias and kind of --

So, I am hoping that going back to the search engines that we just talked about, and maybe not just search engines but everywhere that we -- the Internet becomes kind of more inclusive in how it presents itself, for example, when talking a search term.

I'm also hoping, and there's one more other hope, that it becomes more generally available to many more people in this country and around the world because as I said before it is an amazing tool and obviously we have to learn how to use it. Our children have to learn how to use it. But as we're improving the tool and making it more inclusive and less biased, I do think it's amazing what we're trying to do for the world.

MS. ROSENWORCEL: Oh, I couldn't agree more. What a terrific note to end things on. So, thank you so much for joining us.

And that wraps up another episode of Broadband Conversations. Thank you for being here. Thanks to everyone for listening. Take care.