Achieve DEI Goals Without DEI Programs | Frank Dobbin, Alexandra Kalev
body
Idea in Brief The Challenge Formal DEI programs and policies are being scaled back or eliminated, and champions of workforce diversity feel their work is being undone. The Solution Many recent management innovations designed to improve performance also boost workforce diversity—for frontline workers and managers alike. And they don’t invite the backlash that formal DEI programs do. The Way Forward If companies start using the high-performance management practices described in this article, their diversity numbers are likely to improve. But that will happen only if these innovations are used to manage all employees. During the past few years, DEI programs have been rolled back in both private and public organizations, and the trend may well continue. For champions of workforce diversity, who feel their work is being undone, this is a difficult time. But there is promising news. A growing body of evidence suggests that many management innovations designed to improve performance actually boost workforce diversity in the bargain—and do so without inviting the backlash that formal DEI programs can incur. That evidence has come together during the past seven years, which have seen unusually low unemployment. In an effort to do more with fewer workers and to keep their workers engaged, smart executives across many industries have begun relying on tools of high-performance management that target different periods of the career cycle: recruitment (bringing in the best people at the start); skills training, mentoring, and work-life support (helping every recruit find the right path and thrive during employment); and retention during hard times (ensuring that if cutbacks are necessary, decisions are made on the basis of performance rather than function or tenure). The animating idea of high-performance management is simple: If you can create a work environment in which all employees are valued, supported, and motivated in ways that allow them to do their best, you’ll get higher engagement from them and better business outcomes. Diversity isn’t the goal—but it is a natural by-product. Our research on workplace diversity—conducted in the United States but relevant globally—confirms that notion. We’ve run statistical analyses of data from some 800 companies in a variety of industries. Many of the techniques that companies use to improve performance have a better record of fostering inclusion than do diversity trainings and grievance processes—popular DEI measures that tend to be counterproductive, as we detailed in HBR almost a decade ago, in “Why Diversity Programs Fail” (July–August 2016). That’s true not just for frontline jobs but for all sorts of positions, including management. This remarkable development should inspire hope. If more companies start using certain high-performance management techniques, we can expect to see their diversity numbers improve—even as formal DEI programs are scaled back or eliminated. The key, however, is that these practices help firms attract, develop, and retain diverse talent only when they are used to manage all employees, not just the tiny percentage tagged as “high potential.” In this article, moving from start to finish through the career cycle, we’ll focus on five companies that built performance-driven HR tools that had the unintended consequence of increasing diversity. Referral Programs at Oracle In 2004 Oracle bought PeopleSoft, an enterprise software company, with a plan to revolutionize employee management. PeopleSoft’s core innovation had been software for HR tracking and accounting. Oracle’s CEO, Larry Ellison, wanted to use the company’s expertise to integrate new performance-enhancing management practices into cloud-based tools—a system called Human Capital Management (HCM). The system was just one component of Oracle’s Fusion management suite, which was designed to improve not only human resources but also resource planning, firm performance, the customer experience, manufacturing, and the supply chain. HCM Fusion would best the competition, Ellison’s team hoped, in part by building on the strength of employee referrals. Academic studies at that time had shown that candidates referred by employees were more likely to be interviewed and offered a job and to accept that offer. Not only that, those candidates performed better on the job, were less likely to leave, and were more likely to be promoted. As Oracle reported in a white paper, the best recruiters for new talent were hiding in plain sight—among a company’s own staff. Oracle first built an employee-referral feature into an HCM platform in early 2012. The feature allowed employers to push out job ads to employees, who could apply for jobs themselves or use a “social sourcing” feature to refer friends. Soon the company integrated a referral feature into the legacy product, Oracle PeopleSoft. Employees who made successful referrals earned bonuses. In 2011 we happened to be interviewing executives at U.S. corporations about recruitment. The topic of referral programs often came up, and in those conversations we were surprised to hear again and again about a benefit that we hadn’t expected: The programs seemed to be boosting diversity. At one Boston hospital, a workforce-development manager told us that in her experience, if you asked workers to refer their friends and family members, one “great hire” would lead to another, and diversity “just kind of happens.” Other HR managers shared similar stories. More recently, when we conducted our 800-firm analysis of workplace diversity, the data confirmed what we had been told in our interviews: Referral programs do indeed boost diversity. They even do so in companies that have long relied on informal referrals. Why? Because people of color in frontline jobs rarely feel empowered to refer family members and friends unless they are actually asked to. A formal referral program provides that “ask.” Even employers with mostly white workforces see growth in their diversity because the few nonwhite workers they do have feel licensed to refer friends. And those referrals amount to two-way vetting—of the candidate as a great worker, and of the organization as a great place to work. The benefits are notable in management jobs as well. If maintained for several years, we have found using advanced statistical analysis, referral programs increase the proportion of Black, Hispanic, and Asian managers by about 5%. Moreover, we found that when companies combine referral programs with employee-resource groups, which can be effective in publicizing referral programs, they see additional increases of at least 7% in the proportion of managers who are Black, Hispanic, and Asian. Today HCM Fusion and Oracle PeopleSoft together control nearly 20% of the human-capital-management market. Others have followed their lead, notably Workday, which in 2014 debuted a recruitment tool with a “refer” button—and now accounts for 26% of the market. Employee-referral systems have the potential to open opportunity across the workforce, and many HCM platforms make them broadly available and easy to use. But to date most companies haven’t realized that potential. Some use referrals for only a handful of hard-to-fill jobs, such as pediatric surgeons and aerospace engineers, and thus don’t realize the benefits of leveraging every employee’s network—the key to improving the quality of all recruits and building a more representative workforce. Even companies that use human-capital-management systems often don’t sign up for the employee-referral options offered. That means they can’t use employees to sell the company to their connections and vet potential recruits. As a result, those firms fail to harness the power of their employees’ networks to recruit more broadly, including among communities that have rarely been tapped and prospects who may not even be looking. Those are huge missed opportunities. Skills Upgrading at Walmart In 2014 Doug McMillon became the new CEO of Walmart. McMillon, who had started at a company warehouse while still in high school, knew the business inside out. At the time, Walmart was losing market share to Costco and Target, both of which were doing a better job of retaining their frontline workers. Costco, for example, had reduced turnover to a manageable 17% a year, whereas Walmart was looking at 44% overall—and as much as 90% in some stores. Better retention naturally translated into a better customer experience. An analysis by Wayne Cascio in 2006 found that by paying higher wages and addressing turnover, Costco had built a knowledgeable and committed staff with double the productivity of Walmart workers. (See “The High Cost of Low Wages,” HBR, December 2006.) Almost any Costco staffer could help shoppers find what they wanted. Not so at high-turnover Walmart. Pierre Buttin With Costco and Target gaining market share, McMillon was willing to raise wages. But how could he improve staff quality in the bargain? Drawing on his decades of experience at the company, he decided that training was the answer. In 2016 he opened the Walmart Academy, which now has more than 200 sites across the country that offer a six-month, part-time training program focused on practical skills. The Academy represented a radical innovation in retail. Associates didn’t just learn checkout procedures. Through hands-on exercises, they learned about the retail business model, teamwork, merchandising, and communication. Essentially, they took Management 101. The most innovative aspect of the program was a module on career paths. In it, each trainee chose a career goal—hourly supervisor, say, or electronics manager—and learned how to get there. The message to employees was clear: Walmart could offer more than a starter job. The academy carried forward Walmart’s long-standing commitment to internal promotions (70% of store managers had started in hourly jobs), and as an additional incentive, it gave graduates a dollar-an-hour raise and invited them to enroll in an assistant-manager course. By 2020 the academy had trained half of the company’s 1.6 million U.S. workers. Walmart tracked 10 metrics to evaluate retention, productivity, and engagement—and confirmed that the academy was delivering results. Those metrics also revealed that the academy’s training boosted diversity among employees, managers, and executives alike. In 2016, for instance, people of color had accounted for 42% of Walmart’s workforce. The company’s training centers had launched by 2017, and by the end of the year promotions of people of color from hourly to management jobs had increased by 5%. Over the next eight years, the percentage of people of color in management positions increased from 31% to 43%. The academy also positively affected women at Walmart. In the 2010s men who could no longer find work in U.S. manufacturing were taking more jobs in retail. As a result, from 2016 to 2024 the total percentage of women in Walmart’s workforce dropped from 56% to 52%. Nonetheless, during the same period female managers held their own, and the number of female executives rose substantially, from 32% to 39%. (See the exhibit “Walmart’s Changing Workforce.”) See more HBR charts in Data & Visuals Our research shows that Walmart’s experience is not unique: Across the 800 firms we studied, skills training boosted management diversity. Offering job rotation stints to prepare employees for different roles, for example, increased the percentage of white women, Black men and women, and Asian men and women in management by 3% to 5%. Walmart has worked to spread the word about the many benefits of skills training through its larger network, including the Walmart Foundation and research partnerships. And in 2022 the company launched One Global Walmart Academy, which is available to all of its 2.3 million associates around the world and has become a model for upgrading the skills of workers. This past November, under pressure from political activists and social-media influencers, Walmart, like many other organizations, made public promises to cut DEI initiatives and to end preferences for women- and minority-owned suppliers. However, the company did not mention its hugely successful training program for frontline workers, more than 50% of whom are now people of color. With this program, Walmart may have figured out how to boost workforce diversity and improve the bottom line in a world in which explicit DEI programming is under attack—an approach that other companies may want to emulate. Mentoring Programs at IBM In the early 2000s IBM was facing a host of challenges, including cutthroat price competition in personal computing. Two years after Sam Palmisano took over as CEO in 2003, he sold the company’s PC business to Lenovo and expanded cloud computing and AI. Palmisano also committed to talent management through knowledge sharing and one-on-one mentoring, which IBM had a storied history of encouraging. His radical innovation in this area was to build a platform with a suite of tools for quickly transferring the knowledge and skills essential for innovations like cloud computing and AI to IBM’s 356,000 workers around the globe. The resources on the platform included guides, success stories, podcasts, virtual speed-mentoring cafés, mentoring cafés for students from historically Black schools, and a real-time “Dear Mentor” chat app. IBM also built international and cross-site mentoring programs to realize the benefits of rotating staff through facilities, an IBM tradition, without incurring disruptions to employees’ personal lives. Palmisano directed managers to recognize mentoring when completing performance evaluations and recruited mentoring champions. Palmisano’s successor, Ginni Rometty, shared his commitment to mentoring and knowledge exchange. When she began to post IBM’s diversity data, in 2018, the firm turned out to be ahead of most in the industry for women in tech, managerial, and executive roles—another example of a diversity boost brought about by a focus on high-performance management. In the five years that followed, IBM continued to see the percentage of women in all three roles grow, even as the industry overall lost women during the pandemic. (See the exhibit “IBM’s Changing Workforce.”) See more HBR charts in Data & Visuals IBM also turned out to be an industry leader in racial and ethnic diversity in 2018 and saw steep increases in it over the next five years, despite industrywide layoffs after 2020 that halted or reversed progress in many firms. Our research reveals that when formal mentoring programs are in place for all staff, they boost the representation of women and people of color in management across industries. In highly skilled industries, such as computing, electronics, and chemicals, mentoring programs increase the share of managers who are white, Black, Hispanic, and Asian women or are Black and Asian men by at least 15% over several years. In fact, in these industries mentoring is the single most effective diversity booster. Schedule Flexibility and Stability at Gap The growth of online shopping in the early 2000s, and its acceleration a decade later with the spread of smartphones, posed a challenge to the popular specialty retailer Gap. As at Walmart, executives believed the retail chain’s survival depended on the customer experience and, ultimately, the quality of its sales force. But the 2008 financial crisis led Gap to cut expenses by hiring fewer full-time associates and allowing managers to adjust associates’ schedules at the last minute. Those moves raised staff anxiety and lowered engagement. In 2014, San Francisco, where Gap is headquartered, passed the Formula Retail Employee Rights Ordinances, which required companies to offer employees some schedule predictability. Seattle, New York City, and the state of Oregon soon passed similar bills. This trend prompted Joan Williams, a professor at the University of California College of the Law, San Francisco, to suggest an experiment to Gap’s executives: Implement some scheduling innovations known to improve employee satisfaction, retention, and performance, and see what they cost. Pierre Buttin The company decided to launch a 35-week trial at 28 stores in Chicago and San Francisco. It involved four changes: First, it introduced a mobile app, called Shift Messenger, that facilitated instant shift swapping without a supervisor’s approval. Second, it gave a “soft guarantee” that core associates would get 20 hours of work every week. Third, it gave those core associates stable week-to-week schedules when possible. Fourth, it established standard start and end times for shifts. Associates were much happier after the changes, and supervisors had no complaints. But the benefits extended beyond just employee satisfaction. Williams and her team found that efficiency had also improved. Store sales increased by 7%, and labor productivity increased by 5%, generating an extra $2.9 million in profits during the trial—100 times what implementing the changes cost Gap. The sales increase could be traced to better customer service: Associates were less stressed and more engaged, the data showed, which meant they were more helpful. That, in turn, led more customers to make purchases—and to spend more when they did. With greater engagement came more associate feedback and suggestions. Workers reported, for example, that items customers wanted were out of stock 40% of the time, so the company built an app to allow for timely replenishment. Since then, Gap has rolled out these scheduling changes at all of its Gap, Banana Republic, Old Navy, Intermix, and Athleta stores. Many other companies have since jumped on the bandwagon, and it is now common knowledge that schedule predictability improves retail performance. But our research reveals a little-known knock-on effect: In scheduling, the combination of flexibility and stability disproportionately helps employers retain and develop people of color and women, which in turn increases their chances of moving into managerial positions. We can see that effect in Gap’s own workforce data, which the retailer began posting in 2011. When we examined the data, we found steady growth in the percentage of people of color working there, with the exception of a brief pandemic reversal. Among store managers, apart from the pandemic blip, that growth accelerated dramatically after 2018, when all the scheduling innovations were implemented. From 2011 to 2017 the number of store managers who were people of color rose from 25% to 29%—and from 2017 to 2023, from 29% to 49%. (See the exhibit “Gap’s Changing Workforce.”) See more HBR charts in Data & Visuals The data on women at Gap tells a similar story. The number of female employees rose gradually from 2011 to 2017, from 73% to 76%, while the percentage of female store managers vacillated between 68% and 69%. But after the scheduling changes were put in place, the percentage of female store managers rose rapidly, to 76%. Does giving workers more-predictable and adjustable schedules always boost diversity? In “The Surprising Benefits of Work/Life Support” (HBR, September–October 2022), we looked at schedule flexibility across industries and reported that it did indeed lead to increases in white, Black, Hispanic, and Asian women in management. That’s impressive, but the real surprise is that Black, Hispanic, and Asian men benefit too. Why? Because these workers are often in dual-income households with children and have to juggle jobs and share childcare duties. White parents in the United States today are more able to pay for help because on average they have more wealth than their Black and Hispanic peers—10 times more, on average. Black and Hispanic men and women who are parents benefit from schedule predictability and flexibility, which allow them to stay on the job, put in the additional effort often needed to prove their talent, and move up rather than transfer to inferior jobs that allow them more flexibility. Nonwhite employees need this kind of support most but are less likely to be in jobs that offer it. And even when they do have those jobs, they’re less likely to feel comfortable asking for support—or to receive it if they do ask. When flexibility and predictability are granted to all employees, the career prospects for Black, Hispanic, and Asian workers improve. Performance-Based Retention at Amazon Jeff Bezos has famously built a performance-centric ethos at Amazon, known to insiders as “purposeful Darwinism.” To make Amazon excel in every sector it entered, whether e-commerce, cloud computing, or personal devices, he rewarded his top performers handsomely—but he also directed supervisors to grade employees on a curve, with at least 5% of them failing. Managers got “unregretted attrition” targets every year, to be achieved through nudging and firings. Amazon grew meteorically from 2016 to 2021, with its headcount rising from 175,000 to 1,120,000. The increase was especially sharp in the final two years of that period, when online sales and streaming exploded because of Covid-19. But as the pandemic subsided, the company—led by Bezos’s handpicked successor, Andy Jassy—announced a major downsizing: Some 27,000 corporate employees were laid off in 2022 and 2023. (To learn more about Jassy’s leadership, see the HBR Interview “Speed Is a Leadership Decision” in this issue.) The layoffs were a test of Amazon’s radical performance-first culture. Experts wondered whether performance-based layoffs might have the same effect as layoffs conducted in the conventional manner, in which newer hires are let go first, and job cuts start in “nonessential” departments such as customer service, legal, and DEI. The effect of conventional layoffs has been well established: Companies that implement them at scale hemorrhage women and people of color. Would the same thing happen with performance-based layoffs? An analysis that one of us (Alexandra) conducted in 2014 suggests that the answer is no: Employers who use performance ratings to determine layoffs maintain overall diversity in management. And in fact, Amazon’s layoffs did not lead to disproportionate losses of women and people of color. Keeping employees with strong performance turned out to be a great strategy for sustaining, and even increasing, workforce diversity in hard times. Part of the reason is that firing the latest hires means losing the most diverse segment of the workforce—recent graduates. Another part is that cutting “nonessential” roles in customer service, legal, and DEI means letting go of a lot of women and people of color, who are often channeled into these areas when hired. Amazon began posting workforce data when its headcount started expanding rapidly, in 2016. As its workforce grew, so too did diversity in its ranks—a natural outcome. Significantly, however, the percentage of managers, professionals, and employees of color continued to rise even during the period of major layoffs, in 2022 and 2023. Women didn’t lose ground, either; the proportion of female professionals and managers continued to grow. See more HBR charts in Data & Visuals These results suggest that layoffs don’t have to have a disproportionate effect on underrepresented groups. We now have concrete evidence to suggest that a focus on performance will ensure that businesses keep their best employees from all groups on staff as they navigate the choppy waters of economic downturns. Laying people off by function or by using “last in, first out” will inevitably cut some excellent performers—and reduce company diversity. And firms can minimize layoffs in the first place by cross-training staff to be ready for new roles. The performance-first approach is gaining admirers. In 2025 Meta, for example, asked managers to cut the lowest-performing 5% of workers—a move straight out of the Amazon playbook. This approach can feel cutthroat to employees, leading some great workers to leave. But if firms need to reduce staff during hard times, they don’t need to sacrifice high performers—or workforce diversity. . . . Oracle, Walmart, IBM, Gap, and Amazon all have distinct cultures and business models. But they share a commitment to excellence, and that has led each to experiment with high-performance management innovations. Their own data, and our analyses of data from 800 firms, confirms that their innovations also boost diversity. When you think about it, it’s not all that surprising. By creating formal referral programs, firms leverage the networks of their entire staff. By creating skills-training and mentoring programs that everyone can sign up for, firms extend the benefits of informal on-the-job training and mentoring to people from all groups. By promising schedule predictability and flexibility, firms help their Black, Hispanic, and Asian employees who lack the means to accommodate rigid, unpredictable schedules when family demands peak. And by basing layoffs on performance, firms retain their best workers rather than shedding the most recent hires and staff in “nonessential” functions, who tend to be women and people of color. For years, diversity and human resources experts have been calling for firms to change their management systems to open opportunity to all. The approaches we’ve identified in this article do just that—even if that was not the intention of their creators. In these tumultuous times for DEI, performance innovations that help employees be their best selves at work may be the best hope for the dream of equality.