If DEI Programs Aren’t Effective, What Is?
29:23 Share

Welcome to the HBR IdeaCast from Harvard Business Review. I'm Alison Beard.

In the United States and other parts of the world, there's a big debate happening right now over how public and private organizations should operate. For a long time, many employers were trying to make their workplaces more diverse, equitable, and inclusive, arguing that there was both a moral and a business case for initiatives like anti-bias training or gender quotas.

Recently, though, there's been a rising chorus of people saying that DEI programs and policies run counter to the idea of a meritocracy, where the best get ahead, the rest fall behind, and the operation as a whole succeeds. So who's right?

Our guests today think we all need to look at these issues through a different lens. They want leaders to focus on making work fair for everyone across levels and teams because their research shows that such efforts lead to organizations being both meritocratic and diverse, equitable, and inclusive.

In short, it's about doing everything smarter and better, not as part of any special program, but as an everyday practice. Iris Bonet is a professor and Siri Chalazios is a senior researcher, both at the Harvard Kennedy School. They wrote the book Make Work Fair, Data-Driven Design for Real Results. Thanks both of you for being here. Thank you for having us. Thanks for having us, Alison. It's a pleasure. Thank you.

So this debate over meritocracy versus DEI, it isn't new, but it's a really hot one right now. Why do you think that we're still having it? I actually want to go back to this concept of meritocracy. I had to look it up recently because I realize it's a word that I've been throwing around quite a bit without actually knowing how it's formally defined. So here's what the Merriam-Webster Dictionary has to say. It defines meritocracy as a system in which people advance and are given power according to their skills and talents.

I think it's an incredibly worthy ideal to strive for. But the evidence is overwhelmingly clear that we are not living in a meritocracy today. Women and men get differently recognized and rewarded for equal contributions in group work.

Similar performance evaluations translate to higher rewards and faster advancement for men than they do for women in many workplaces. More than 300 studies show that if two people with the same exact resume have a different name on top that indicates a different gender or race or religion, they don't have the same chance of getting called in for an interview.

So the evidence is clear that what we have today is not a meritocratic workplace. I firmly believe that if we start fixing our structures and the processes whereby we do work and embed fairness into those processes, it will help us move closer to this ideal. And so why is fairness a better frame or lens to use when we're thinking about these issues?

So we introduced fairness as a term for a number of reasons. First of all, we want to make clear to people that fairness is not a program, but it is basically about how we do things. So fairness is only going to work if we build it into everything that we're doing, into our daily works and how we hire, how we promote, how we do performance appraisals, how we run meetings, how we design our products, really everything.

And with fairness, we mean providing true equal opportunities for everyone. And maybe people want to think of a race, right? Where ideally we all start at the same place. Think of the 100-meter dash.

Ideally, we all would have had equal opportunities to train, to practice, equal support to be where we're at now, equal equipment, equal shoes. So that's basically what we're striving for, that we create real equal opportunity for everyone to be their best self. And how do you quantify that?

Whether systems are fair or not, it seems like a slightly slippery thing that might be regarded as subjective. You know, what's fair to me might not be fair to you.

This is where data comes in and the same analytical approach that we take to running our core business, right, is the exact approach that we need to take to measuring fairness. The good news is that so many more things can be measured than we even realize. As we like to say, data tell the tale or data don't lie. Let's look at promotions.

If you examine your internal data and you see that certain groups of employees are spending more time at a given rank before advancing to the next rank, that begs the question, why are we advancing certain groups faster than others?

What is it about the criteria that we use for promotions or the way we make those promotion decisions that's giving a leg up to certain employees over others? Now, this, of course, doesn't mean that every single employee will get promoted at the same rate because promotion

In reality, not everyone has the same aspirations. Not everyone has exactly the same skills and capacities. But if data shows us that there's a lot of systematic differences emerging between groups, a lot of systematic gaps, that would be an indication to dig into the processes that are contributing to those gaps and make sure that there's no unintentional bias baked into them.

For a long time, we thought that DEI programs, which many employers adopted, were going to solve these problems. So for the people who supported diversity, equity, inclusion, they were excited about that. For the people who thought that was actually unfairly disadvantaging majority group inclusion.

members like white men, they didn't like it. What's your view on how effective DEI programs were at leveling the playing field in the way that you're talking about?

So the tool that was actually used most often was the diversity training. Because program is a big word. Most organizations did more than just trainings, but because trainings have been so popular, maybe that's what we want to hone in here for just a moment. So sadly, and maybe to the surprise of many people, trainings actually haven't been shown to have a tremendous impact

They can raise our short-term awareness, but nobody in fact has found very long-term impact on behaviors. And that's really the ultimate goal, right? That we'd like for people to change how they do things. And that's just really hard. And in behavioral science, we've actually found that it is incredibly hard to change mindsets.

Now, Siri and I and some colleagues of ours, including Edward Chang of the Harvard Business School, in fact, tried to work on that. We as researchers, like many others, had kind of given up a little bit.

on diversity training, thinking that we need to have more systemic interventions. But given that they're still so popular, we recently were able to collaborate with a large firm. They're global, active in 100 countries, and they were interested in their hiring managers hiring more inclusively. And that, for them, meant

meant that hiring managers often have a preference for people from their own countries. They wanted to make sure that hiring managers didn't overlook talent just because it came from a different place.

And the second concern they had was that given that most hiring managers, about 80% of hiring managers, in fact, in this firm were men, hiring managers might overlook women. And so the question was, could we help them? And we, in fact, did do a, it's very different from a traditional diversity training program, but a version of a diversity training that was built very much on insights from behavioral science.

It was very focused, it was very targeted on the dependent variable. So we started with the final goal of this company. It was very specific. We want you to hire based on the talent that is available, independent of considerations of gender or, you know, other concerns that we might have.

The second insight was that it had to be timely. So one of the issues that we think happens with traditional diversity training programs, not only are they very generic, but they happen at a random moment in time. So

So they might happen during onboarding or whenever it is appropriate to do these things based on the company's schedule. And so we really wanted these hiring managers just to learn about these insights just before they made these hiring decisions. And then thirdly, often these training programs are delivered online.

by well-intentioned, highly skilled trainers, but not necessarily by an authority figure. So in our case, we also had senior managers of the company deliver the message. And literally, it was a seven-minute video clip

where the senior managers talked about the importance of skills, of complementarities, alluding to the fact that diversity can, in fact, make teams stronger and that this company, in fact, valued that kind of diversity. And we found that it can have an impact. And the impact that was on both women and people from outside of the country and the strongest impact, in fact, was online.

on women from outside of the country who ended up being hired more frequently.

I'd love to pick up on that point about the importance of defining what the goal is of all of our efforts before undertaking them, because we feel like this is one of the places where some of the traditional corporate DEI efforts have gone astray. You know, one example of a ubiquitous practice that really hasn't been evaluated on any kind of large scale is employee resource groups and affinity groups and networks like women's networks.

They have been shown to be able to provide people comfort and connection. They are, in fact, often instituted to mitigate the symptoms of, you know, underlying issues of underrepresentation or biased processes. But there's no evidence to show, for example, that being part of an employee group or going to networking events actually makes you more likely to get promoted.

or changes your performance evaluation. And in fact, that's also often not what companies measure. When they tout metrics related to these ERGs, they'll say, oh, we have seven ERGs and collectively they organized 27 events last year and 3,000 employees attended these events. And those are metrics, yes.

But are they reflective of our ultimate goal? Did people participating in those events or did the existence of those networks actually help to close gaps in hiring, performance evaluations, promotions? Did it make people more likely to get selected for the high potential program that identifies future leadership talent for the company? Did it give people more of a share of voice in meetings?

We would argue that those are some of the actual goals that we're trying to achieve with our efforts. And it's really important that whatever we do is actually aligned to those fundamental goals. So what are some good goals to set when it comes to fairness within our organization? And what's the right way to set them?

Every organization should start with their own data. There is no silver bullet, but also every single organization is starting from a different place, has different opportunities, but also has different challenges. Five or six decades worth of research into goal setting shows that they have to be carefully calibrated. If you set goals for yourself that seem unattainable or unrealistic, it can actually demotivate people because the goal feels so far out of reach.

On the flip side, if your goal is too easy and it doesn't really require you to change anything to attain it, the goal has not yielded any kind of behavior change. So organizations should look at their own data today, see where they have gaps, and then strategically set these stretching yet realistic goals to close those gaps.

One example of a goal that we really like that we started with our collaborator, Oliver Hauser, is this idea of proportionality. In many organizations today, talent exists at the entry levels, whether it's being close to 50-50 women or being close to racially representative of the broader population. But then as we go up in the organizational hierarchy, women's presence, for example, dwindles as you go up the ranks.

So when we set a goal of proportionality, we're essentially saying that we aim to promote people in the same proportion as they are represented at the feeder level. So if 40% of your managers are already women today and the next level director is just 25% women, well, your talent pool is at least 40% women today.

In the goal of proportionality, at least 40% of your promotions would also be women. It really focuses organizations on thinking about how they're growing that talent and make sure that everyone actually has an equal chance of rising up through the ranks. Another goal could also be to try to close the gaps that organizations have identified. So a few years ago, we were approached by a large organization in Australia called

And their goal literally was to close a gap. What they approached us with was that they have people apply to leadership positions and they choose someone. And then they send people back an email saying, thank you for applying. We like your application. There are similar roles like this one available in our organization. Please reapply. And they found that men were about twice as likely to reapply than women. And

And of course, they disliked this gap for lots of reasons, for fairness reasons, and also because they had, in fact, just vetted all these people. And so as an organization, you would like them to be in the pool. In this case, actually, we had a relatively simple idea. We said, well, you know, maybe just add one more sentence to the message that you send back.

And the sentence, in fact, built on something that they had told us before, basically in passing, that they sent these emails to the top 20% of the applicants. And we're like, yes, you should tell your applicants that. So just add one sentence, you know, thank you for applying, please reapply, and your top 20%. And it turns out that one sentence helped us close the gender gap in reapplication rates. ♪

Need to improve your expense management processes? Consider the U.S. Bank Commercial Rewards Card for your business. More than just a card, it's an all-in-one card, expense, and travel management solution that simplifies expense management with a simple, intuitive platform. Get control and real-time insights into spend, set custom policies, and earn rewards, all with no annual fee.

Manage your cash flow better and streamline your operations. Learn more at usbank.com. AI is coming to your industry if it isn't already here. But AI needs lots of speed and computing power. So how do you compete without cost spiraling? Upgrade to Oracle Cloud Infrastructure or OCI.

OCI is the blazing fast and secure platform for your infrastructure, database, application development, and AI workloads. Right now, Oracle is offering to cut your current cloud bill in half if you move to OCI for new U.S. customers with minimum financial commitment. Offer ends March 31st. See if your company qualifies at oracle.com slash ideacastai.

I can see why this would be a focus and perhaps easier for larger organizations. What advice do you give, you know, to small businesses, medium-sized ones, startups even, about how they should be thinking about fairness when they have sort of more pressing issues, I guess, to focus on?

Hmm, it's such a great question. First, I would say it can go both ways. Larger companies often have, you know, more HR staff, maybe a people analytics department, possibly larger budgets at their disposal. They also have

embedded processes that can be more difficult to change because there's more of a status quo established. It's like turning around a small speedboat versus a Titanic or an ocean liner, right? So I actually think some startups and smaller organizations have an advantage there in

It's much easier to build the right processes from scratch and cement them and then scale them as you go to hundreds and thousands of employees as it is to have embedded and kind of established processes that are not optimal and then have to be revised later.

But I also wanted to pick up on what you said, Allison, about more pressing priorities. Because when we talk to senior leaders of organizations, really leaders at all levels, and we ask them, what is the most critical enabler of your success? Usually the first answer is our people.

We need to have the right people in the right roles at the right time doing their best work in order for this organization to succeed. And that is not possible without fairness. If we can't hire the best people, if we don't know how to evaluate them fairly, if we don't give them the opportunities to grow and stretch and learn new skills along the way and then put them into roles where those skills can be leveraged,

We're simply not going to be able to run the most financially successful or otherwise high-performing organization that we could be. So I would argue that embedding fairness is the foundation, the sort of number one priority upon which you build everything else. Can this also be done at a team level? You know, can an individual manager take it upon themselves to make work more fair for their employees?

you know, even if the sort of broader organizational structure and processes aren't necessarily moving in that direction? Well, I think the power of this approach is that really everyone, no matter what your role, there is something that you can measure about your own work. You know, maybe you're the junior intern in the marketing department and part of your role is to manage your company's social media or update the website or write a blog for the intranet.

Well, you might be able to measure who's featured in your pictures and your videos that you're posting. Who are the people in the company that are named and highlighted in the blog articles or the press releases that you're writing? And then look at, is this representative of our employee base? Is this representative of

representative of our customer base or the world more broadly. Or let's say you're a venture capitalist or a loan officer and a core aspect of your job is you're reviewing these applications for funding or for loans. Well, could you track perhaps the gender of the application submitted or the race or even the educational background or the geographic distribution, right? Depending on what's relevant in your job.

Roz Atkins, who's a journalist at the BBC, did this for gender. He started counting who he was featuring as experts and contributors on his nightly primetime news show because he knew that in order for his journalism to be the highest impact and the highest quality, it needed to be representative of the world that it was reflecting on.

And before HR handed him any data, you know, he didn't let that stop him. But he just said to his team, hey, let's take two minutes a day and just count how many men and women appeared on screen during our hour. So all of us can do this and it doesn't have to be complicated or time consuming.

So we hope that people will support this idea of creating fair workplaces. But are there ways to incentivize that kind of change beyond just people feeling it's the right thing to do?

A number of the leaders that we talk to turn to us because they believe it is the right thing to do. So that, I think, is always a starting point. And we shouldn't have to justify equal opportunity, in fact. But then we obviously work with private sector, public sector NGOs, and most of these have other goals to meet.

And so then the question is, what does fairness do for us to actually meet those goals? And Siri just talked about Ross Atkins. It's such a good example because not only were they able to increase the fraction of women on their programming and then actually spread across the whole BBC and eventually was picked up by more than 150 other organizations around the world, but also it increased diversity of the viewership.

So people enjoyed hearing and seeing themselves on the BBC. And that's, of course, an important consideration that most organizations...

care about customers, about clients, and also about benefiting from 100% of the talent pool. And that, I think, is often an argument that resonates with almost everyone, that they just feel that they tend to replicate themselves and don't really benefit from what is available in the world. A study that I really liked estimated that

how much GDP per person in the United States has grown because the talent pool was increased. Basically, what they were asking is, what would GDP per person look like today if the restrictions of the 60s that kept women and people of color from succeeding fully in the workplace, what would happen if those restrictions were still in place?

And they found that 40% of growth per capita is explained by an increased talent pool. What that meant is that people and organizations would meet each other more effectively and that people, the right people, would be in the right jobs at the right time. It's an expanding pie, not one that we all have to divvy up. 100%.

How do you convince leaders across the organization that creating a

more fair work environments is part of their jobs and not just the job of HR or head of DEI when they do have lots of competing priorities. An analogy to communications might be helpful here. A lot of companies have PR departments or corporate communication specialists who handle some of the company's most highest profile communications, like writing the annual report or preparing a speech for the CEO.

But at the same time, up and down throughout the whole organization, people are writing emails on a daily basis, speaking up in meetings, creating slide decks, creating spreadsheets. In other words, everyone has to have a base level of communication skills in order for the organization to function. And Iris and I argue this is exactly how we have to approach fairness as well. We have to think of it as a core competency that's baked into the things that we're already doing.

So our employees, what are they doing? Well, they are writing those emails. They're attending meetings and organizing meetings. They might be interviewing new colleagues either for hiring from the outside or maybe internally for a new project or a new team if people are moving around. From time to time, we're evaluating each other's performance.

When we embed fairness into what we do, we bake it into these types of daily activities so that instead of taking on new things onto our plate, we continue what we're already doing, but a little bit better. And I think this is where that systemic frame can be so important because when we design these systems around hiring and meetings and feedback to be more fair and

and less subjective, we're helping our inherently biased brains make better decisions so that instead of trying to de-bias every single individual in that organization, by de-biasing one system, we get ourselves to behave in less biased ways.

Because there is this rising tide of anti-woke or anti-DEI activism, what advice are you giving to people and organizations that believe in everything you're saying about promoting fairness? How are you telling them to continue their work and sort of push forward to achieve the change they want amidst all this backlash? So first of all,

We are telling everyone that we are not lawyers, that we are behavioral economists and in that sense can't give legal advice, what is safe to do at this point or what is not safe to do. And of course, we work with organizations that are active in lots of different jurisdictions around the world. But then secondly, we still offer them a toolbox, kind of given the goals that they want to pursue, right?

We are here to help them design the workplace that increases the chances that they will actually meet their goals.

And, you know, thirdly, I just want to remind all of us, it's obviously a very complex time, but different organizations have made different choices. So we do have JPMorgan Chase, we do have Apple and other large companies that have gone out and said that they want to continue the work of fairness. So, you know, it's hard at this point to actually know exactly

what the landscape looks like, because we don't have good data on what all these companies have decided to do. But we certainly understand that people feel more uncertain about where to go. I would add at the organizational level, this comes down to managing fairness exactly the same way you manage your core business.

If your goal is to be the market leader in your sector and you're looking to grow and you have a merger on the table or an acquisition on the table, and then it gets blocked by regulators due to antitrust reasons, you don't pack up your bags and go home and say, oh, let's forget it. We might as well stop doing business now.

No, you find another way. You say, okay, we tried A, that was blocked, or maybe it went through, but it didn't work. And so now we try something else. Your goal of achieving market dominance and success remains the same. In our core businesses, we're used to pivoting. We're used to reacting to external market conditions, upturns, downturns, political changes, and continuing to move in the direction that is aligned with our mission.

So at the organizational level, that's what we need to continue to do now. At the individual level, I'd want to suggest channeling our energies to

to where we can make progress and towards the people that are already supportive, rather than focusing too much on the detractors. The human brain sometimes tends toward the negative and says, "Oh, you know, the CEO isn't supportive. I don't know if we can move forward." Well, what if five of the seven of the CEO's direct reports are aligned? And what if they could be your champions and allies in moving the work forward?

I'll go back briefly to Roz Atkins, because as Iris mentioned, what started as just one team's effort, Roz Atkins' team's effort, eventually spread to more than 750 teams at the BBC. And it wasn't because Roz Atkins told other people, forced other teams or incentivized other teams to join. It was because he started spreading the approach on a completely voluntary basis, tapping into journalists' shared problems.

motivations. He said, I'm a journalist just like you, and I know that we both care about the quality of our journalism. My team and I figured out an approach that helped us. Here it is. Start counting. And that's how it started spreading initially. And I think collectively, we could get so much further faster if we focused on tapping into all the support and energy that already exists rather than starting by finding the windmills.

Yeah. And I also think open dialogue about what goals can we agree on? Well, terrific. I've really enjoyed discussing this ongoing debate with you. And thank you so much for your time. It's a real pleasure. Thank you for having us.

That's the Harvard Kennedy School's Iris Bonet and Siri Chalazi. They're co-authors of the book Make Work Fair, Data-Driven Design for Real Results. And we have more than a thousand episodes of IdeaCast, plus more HBR podcasts to help you manage your team, your organization, and your career. Find them at hbr.org slash podcasts, or search HBR on Apple Podcasts, Spotify, or wherever you listen.

Thanks to our team, Senior Producer Mary Du, Associate Producer Hannah Bates, Audio Product Manager Ian Fox, and Senior Production Specialist Rob Eckhart. And thanks to you for listening to the HBR IdeaCast. We'll be back with a new episode on Tuesday. I'm Alison Beard.

AI is coming to your industry if it isn't already here. But AI needs lots of speed and computing power, so how do you compete without cost spiraling? Upgrade to Oracle Cloud Infrastructure or OCI. OCI is the blazing fast and secure platform for your infrastructure, database, application development, and AI workloads.

Right now, Oracle is offering to cut your current cloud bill in half if you move to OCI for new U.S. customers with minimum financial commitment. Offer ends March 31st. See if your company qualifies at oracle.com slash ideacastai.