Getting to Diversity

In the 1950s the civil rights movement brought hope that the opportunities that the American economy had afforded to white men would be extended to others. A flurry of federal laws and court decisions opened the way for change. In 1961 John F. Kennedy ordered companies with federal contracts—mostly big firms—to end racial discrimination. Three hundred of America’s leading employers—American Airlines, AT&T, Colgate-Palmolive, and Lockheed among them—began meeting to share ideas for integrating the workforce, calling their initiative Plans for Progress.

In 1963, at the March on Washington for Jobs and Freedom, Martin Luther King, Jr. told the nation that he had a dream of a future free from racial discrimination and animus. In 1964 Congress passed the Civil Rights Act, making it illegal for all employers to discriminate on the basis of race or sex. In part as a result of these events, corporate America as a whole made slow but steady progress on opening opportunity in the 1960s and 1970s.

That progress was based on corporate commitment to change and fueled by political unrest through the 1960s that kept the cause of racial justice front and center. Firms in Plans for Progress, and beyond, experimented with all sorts of programs to promote equity. Some worked, others failed, and in the mash-up of different programs, it was often difficult to know which innovations were leading to change.

It is still difficult to know which workplace diversity and inclusion programs are working and which are not. In the heady days of the early 1960s, when revolution seemed to be around the corner, firms began to open all kinds of jobs to women and people of color for the first time. Through the 1970s, firms made progress on diversifying the ranks, including the ranks of management. But by the mid-1980s, progress had stalled. Firms were hiring more women, Black, and Latinx workers for entry-level jobs than ever before. But between 1985 and 2018 the odds of a Black or Latinx man in corporate America being a manager held steady at 6 percent and 7 percent respectively. The odds for Black women rose from 4 to 5 percent, and the odds for Latinx women rose from 4 to 6 percent. White women’s chances of making it into management also dimmed after 1985 or so. Only Asian American men and women continued to see meaningful growth, but that growth was not in keeping with their growing educational attainment.

Firms made progress before the mid-1980s in part because the political pressure for change was kept up. Both Democratic and Republican presidents had backed affirmative action. And in the early 1960s, leading firms had implemented changes to career systems to open opportunity to women and Black, Hispanic, and Asian American workers. Plans for Progress signatories initiated largely symbolic measures—known as window-dressing then, and “virtue signaling” now—such as race relations workshops and diversity mission statements. But they also initiated systemic changes in how they recruited and mentored people, and trained them for better jobs. During the stagflation of the 1970s, firms let many of those programs languish because they were no longer needed to expand their workforces. By the early 1980s, they were left with the symbolic measures and had mostly taken down the systemic changes designed to bring in, and train up, people from different backgrounds. Then in 1981 Ronald Reagan came to power on a pledge to cut red tape for business, including affirmative action and equal opportunity regulations. Thus, as the economy recovered, companies had only symbolic measures in place and the pressure from Washington had abated. 

In our recent book, Getting to Diversity: What Works and What Doesn’t (Harvard University Press, 2022), we try to understand why progress on opening opportunity to America’s best jobs—management jobs—slowed during the 1980s and has not returned to the rapid pace of the 1960s and ’70s. We crunched data from hundreds of firms with millions of workers between 1971 and 2015 to examine the effects of different equal-opportunity and diversity programs on the workforce. Our analyses suggest that the symbolic, or “paper,” policies that many firms rely on today are ineffective or can backfire. They suggest that changes to corporate career systems, which were popular in the 1960s but then fell out of favor, are much more effective. Other systemic changes, popular in a smaller number of firms today, including diversity taskforces and work-life supports, have also proven quite effective.

Our research offers a way forward for employers, including law firms, that want to go beyond mere symbolic compliance with antidiscrimination laws to make real progress on diversifying their workforces in a sustainable way. Our test of policy efficacy is straightforward —we look at whether innovations change the racial, ethnic, and gender diversity of corporate managers during the years those innovations are in place.

We focus here on what works, not on what doesn’t work, because a growing number of legal and HR experts now recognize what doesn’t work, including common forms of diversity and harassment training, HR rules to prevent managers from discriminating, and grievance systems for pursuing complaints. In most workplaces, these symbolic, or “paper,” interventions typically impede progress, as we detail in Getting to Diversity. But what should employers replace these failed initiatives with? Our analyses point the way. 

The science of workplace inclusion

For years, scientists studied inequality by examining which individual characteristics helped people get ahead: education, technical skills, experience, connections, grit. But as women and people of color gained ground on all these fronts, it became clear that the individual characteristics that guarantee career success for white men don’t necessarily guarantee it for people from other groups: employees with disabilities, LGBTQ+ workers, nonwhites, women, and so on.

Social scientists like us started asking new questions: What goes on within companies to prevent women and people of color from flourishing in the way that white men with the very same qualifications flourish? Most of these studies focused on the barriers firms erect. Some focused on how firms can tear those barriers down and work toward true inclusion. ((An excellent survey of the research from behavioral economists on de-biasing the workplace can be found in Iris Bohnet’s What Works: Gender Inequality by Design (Cambridge, MA: Belknap Press of Harvard University Press, 2016).))

In our research, we discovered all sorts of hidden barriers in company career systems. Most companies pride themselves on being meritocratic. Many employers understand that individuals may be biased and offer diversity training and grievance procedures to fight bias, or at least to signal that they get it. But most view their career structures and formal policies as recognizing and rewarding talent. In practice, under the banner of meritocracy, many have maintained career systems that all but exclude women, people of color, and other groups from key career resources. By excluding many talented people from recruitment drives, mentoring, training for better jobs, and work-life support, firms narrow the pipeline of talent at each career stage. We review here what we demonstrate in Getting to Diversity about how firms can turn career systems that exclude underrepresented groups into systems that include them—how they can make access to career opportunities truly democratic.

Democratizing recruitment

Many formal recruitment systems that look open and fair on paper don’t reach communities of color. Many neglect white women. In a company with mostly white men in skilled and management jobs, recruiting through the channels they have used before often nets mostly white men. The solution is as simple as casting a wider net. In addition to recruiting at majority-white high schools, companies can recruit at high schools in Black and Latinx neighborhoods. In addition to recruiting at the (mostly white) University of Michigan, they can visit Howard University for Black recruits and the University of New Mexico for Latinx recruits. In addition to recruiting at an Ivy League computer science department, whose graduates are mostly men, they can recruit at a Women in Technology forum. For midcareer prospects, companies can go to organizations such as the Society of Hispanic Professional Engineers or the GLMA (Health Professionals Advancing LGBTQ Equality). For top executives, they can use search firms that specialize in presenting diverse slates of candidates, such as the Savannah Group. A diversity manager at a commercial bank told us that their bank’s multichannel management recruitment program includes all of the above. Smaller companies turn to professional associations and NGOs for access to diverse candidates.

What makes the effect of targeted recruitment programs long-lasting is that they change not only who comes in the door but also the managers involved in recruitment. White men managers who join recruitment trips tell us that the experience of visiting Howard University or the University of New Mexico opens their eyes to talent pools they hadn’t thought about. They then draw on this pool in future searches. They also tell us that when they recruit someone on one of these visits, they look out for them just as they would for a friend they brought on. Our data shows that targeted recruitment programs significantly increase management diversity for years to come, but only about one in five companies has one.

Referrals can also be used to diversify recruits. At many companies, informal referrals are the number one source of new employees. For good reason: Recruits who come through friends and family are typically less likely to quit, and more likely to succeed, than those who come through job sites. A referral program ensures that recruits arrive with a buddy on the inside, which is a strong tonic against quitting. But most informal referrals come from managers, who tend to be white and whose networks tend to be white. People of color are often reluctant to refer friends and family, believing they stand little chance of being hired. A formal referral incentive program can increase referrals from employees of color. Because people of color are concentrated in frontline jobs in many firms, referral programs have to encourage all employees to make referrals.  

Tryouts are another way to diversify the pipeline. Companies that offer paid internships and short-term contracts end up hiring more women and people of color for permanent jobs. ((Adina Sterling and Jennifer Merluzzi, “Tryouts as Alternative Hiring Arrangements in Organizations,” Research in Organizational Behavior 39 (2019): 100–22.)) Here’s an important side note: Experience as a temp worker has been shown to hurt white men looking for a job, but to help Black men because temp agencies do due diligence on vetting workers, which helps overcome hiring-manager bias. ((David S. Pedulla, Making the Cut: Hiring Decisions, Bias, and the Consequences of Nonstandard, Mismatched, and Precarious Employment (Princeton, NJ: Princeton University Press, 2020).))

When executives tell us their targeted recruitment and referral programs don’t advance diversity, we often find that they need to be tweaked. The numbers would change if they sent their own managers to recruit at a historically Black college, for example, rather than letting an outside recruiter do it and if they encouraged everyone to make referrals so that referrals are diverse.

In the table we show the effects of targeted recruitment and referral incentive programs on the share of managers from each group. Numbers indicate the average percent change in the years following adoption of the program among firms that adopt the program. Across these programs, we observe the program in place for seven years in the typical firm. In 2015, during the period we observe—1971–2015—57 percent of managers were white men, 29 percent were white women, 2.6 percent were Black men, 2.2 percent were Black women, 3.3 percent were Latinx men, 1.6 percent were Latinx women, 2.5 percent were Asian American men, and 1.4 percent were Asian American women. Thus, for instance, targeted recruitment increases Black men in management by about 11 percent, which would raise their share from 2.6 percent to 2.9 percent in the typical firm. Targeted recruitment and referral incentive programs show a clear pattern of positive effects here. It is worth noting that referral incentive programs are not DEI programs; they are all-around recruitment programs. 

Once new hires are in, other career resources need to be democratized to give everyone a chance to move up. 

Democratizing networks

New recruits need mentors who can help them find their way and spotlight their talents for others to see. Unfortunately, too many new recruits are left out in the cold. One problem is that many employers leave mentoring to happen “naturally.” That is a recipe for excluding most workers from underrepresented groups, because most managers are white men who tend to pick other white men to mentor. But some companies have worked hard to democratize mentoring, opening it to everyone.

The key to building a wide-open formal mentoring program is to make sure that everyone is offered a mentor and that matches are made on shared career interests, not demographics. Thus employers can’t rely on mentoring programs for “high potential” workers, which don’t reach qualified women and people of color who are flying under the radar. They need to offer mentors to everyone. And employers can’t match mentors based on demographics. For one thing, there are rarely enough women and people of color in senior ranks to go around. For another, white men managers are more likely to be at the top of the firm, and women and people of color need access to them as mentors. Matching based on interests also guarantees that mentor and protégé have something in common. It is as simple as asking protégés what skills they’d like to develop and asking mentors what skills they could impart—whether coding, finance, or teamwork.

Mentors in formal programs do the same things that mentors who match informally do. They counsel protégés, of course, but they also sponsor them for training programs, nominate them for awards, and recommend them for teams developing new software or ad campaigns. These programs pay for themselves by boosting retention and helping executives find neglected gems in the workforce.

Mentors, too, benefit from these programs. Mentors in formal programs have lower quit rates and higher promotion rates than peers not in these programs, but the biggest benefit may come from contact with their protégés. Mentoring programs can clue white men mentors into the lived experience of women and nonwhites at work. They may learn for the first time about the obstacles—distrust, expectations of failure—that their Black, Latinx, and Asian American protégés face.

In the table above, we see that mentoring programs have quite strong positive effects on Black women, Latinx men and women, and Asian American men and women. They raise Black women by about 14 percent, from 2.2 percent to about 2.5 percent. For mentoring programs, effects vary significantly among industries in which most nonmanagers do not have college degrees and industries in which most do. In the latter—in electronics and chemicals, for instance—we also see strong positive effects for women and people of color. In most industries, managers must hold college degrees, so if nonmanagers do not hold college degrees, it is more difficult for mentoring programs to help them move into management.

Democratizing skill and management training

Much like mentoring, company training often excludes women and people of color when it is informal. Skilled workers and executives, who are mainly white, may hand-pick trainees who look like them for informal shadowing and on-the-job instruction. Even in formal training programs, access may be limited because of narrow eligibility criteria or because employees can join only through a nomination by a manager.

The benefits of a formal company training program for diversity are clear. They help people move from dead-end jobs into high-wage jobs with promotion prospects. They allow management trainees to learn important skills in making presentations, budgeting, and managing others—skills that the usual suspects identified as “management material” may have acquired through informal training or college courses. A key is to make the system open to all: first by making sure eligibility criteria, such as holding a full-time job, and training technicalities, such as training hours, do not needlessly exclude certain groups; second, by allowing self-nominations to prevent supervisors from excluding anyone; and third, by encouraging supervisors to nominate  women and people of color. Employers should ask themselves whether other groups are getting full access to training, including workers with disabilities and LGBTQ+ employees.

The most effective training programs, when it comes to boosting diversity, are cross-training programs that break down the walls between informally segregated work groups and expose workers to different departments, employees, and managers. Trainees circulate through half a dozen jobs for a month or so each to learn firsthand the skills used throughout the firm. Local managers and coworkers show them the ropes, which builds everyone’s networks, including those of the trainers. Managers get a chance to observe the skills of people from groups they have never hired from before—women who can code or Latinx men who are great team leaders. Soon the trainer is asking around to find a post for a trainee that will use the math skills they gained working in payroll or the contract-writing skills they honed in an earlier internship. Cross-training, then, builds both skills and networks and, our data confirms, managerial diversity.

In the table above, cross-training shows positive effects for white women, Black men and women, and Asian American men and women. It raises white women from 29 percent to about 30.4 percent. Cross-training shows a negative effect for Latinx men, and no effect for Latinx women. This may be because, to go back to the issue of college degrees, Latinx men are least likely among these groups to hold college degrees, and Latinx women are second least likely. Training alone cannot help members of those groups without college degrees move into management in firms that require managers to have completed college. But Latinx Americans are quickly closing the college completion gap, so cross-training may soon show effects for them. 

Democratizing work-life balance

Work-life supports are often informal as well, as when managers give their favored workers whatever help they need to keep them on the job. If that means letting them leave early two afternoons a week for the daycare pickup, so be it. Formal programs can give all employees a license to devise arrangements that work for their managers and themselves. This can democratize access to schedule flexibility, for instance.

Most people need tools for managing work and life. But this has especially been true for women, who are often responsible for more of the care work at home. Support is also especially vital for Black, Latinx, and Asian American women and men, who are more likely to be single parents or in dual-career families with incomes that don’t permit either parent to work part-time for a spell. Some are caring for aging parents to boot. Managers are also more likely to suspect that women and people are not as committed, or talented, as white men and thus must work harder to prove themselves. This makes work-life management more challenging. Without work-life support, these workers may have to quit to find a job that provides more flexibility for tag-team parenting or elder care, setting the seniority clock back to zero in the process and delaying the chance for a promotion to the first rung of the management ladder.

Workers most often ask for flexible scheduling, family leave, and childcare support. Formalizing these programs has been shown to reduce stress and improve performance for all workers. ((Erin L. Kelly and Phyllis Moen, Overload: How Good Jobs Went Bad and What We Can Do About It (Princeton, NJ: Princeton University Press, 2020).)) We found that these programs have the added advantage of increasing the numbers of all four groups of women in management, as well as Black, Latinx, and Asian American men. Even the cheap versions of these initiatives, such as flexible starting and ending times, unpaid leave, and childcare referral services, can help keep employees from these groups on the job for long enough to move up.

Companies can boost the effects of these programs by making them accessible to all and getting managers on board so that they help team members access work-life programs. This, and providing supervisors with tools to manage flexibility and leaves, can transform middle managers from stragglers in the work-life parade to leaders. Companies have to make it clear that no one who requests support will be stigmatized or face career penalties. Some CEOs have taken the lead by talking up work-life balance, but nothing beats CEOs who themselves use family leave, flextime, and company childcare.

Until they democratize work-life programs and start changing management traditions that are stacked against workers with family demands, employers will continue to lose valuable workers, absorb turnover costs, and see slow progress on management diversity.

But as we see in table 1, work-life supports can boost diversity in the management ranks. Flexible scheduling programs, which specify how employees can ask for flexibility but leave it up to their supervisors to decide whether flexibility is practicable, show positive effects on all seven historically disadvantaged groups. Parental-leave programs show significant positive effects on white and Black women, and Latinx and Asian American men. Childcare referral services show significant positive effects on all seven historically disadvantaged groups, while on-site childcare shows positive effects on four, and vouchers to subsidize childcare (on-site or off-site) show positive effects on three. Based on the 2015 prevalence of these groups in management, flexible scheduling increases Latinx men from 3.3 percent to 3.7 percent; parental leave raises white women from 29 percent to 31.5 percent; childcare vouchers increase Black men from 2.6 percent to 3.0 percent of managers. Work-life programs are not always part of the DEI toolkit, but they clearly show impressive positive effects. 

Bake reform into management systems

Opening opportunities requires minor modifications to career systems that are already in place, such as casting a wider recruitment net and formalizing mentoring, training, and work-life programs in ways that make them available to all. But how can employers make sure that old traditions will not creep back in? How can they adjust newly democratized employment systems to changing circumstances to sustain inclusion? How can they expand inclusion in other systems, such as pay, promotion, and layoff systems? How can they ensure all disadvantaged groups are addressed? The best bet is to build structures that ensure regular assessment and updating. Firms can’t outsource that task to consultants or leave all the work to diversity and inclusion units. All managers have to be involved.

If a company wants a permanent structure that gets managers involved, there is nothing like a diversity task force. Task force members come from different departments, groups, and ranks and report to a senior executive. IBM’s Mark Gerstner created task forces organized to focus on eight dimensions of diversity as early as 1995. FedEx’s task force includes managers from all functions, from HR to engineering. In a Chicago electronics firm we interviewed, they included insiders who had climbed up the ranks and knew where the problems were. Task forces can help in several ways. Members spot problems by examining the company’s HR data. They may discover that people from one group get stuck on the second rung of the ladder and that people from another leave after their second child is born. The task force then brainstorms for solutions. Task force members may well become champions in their own corners of the company for the solutions the task force develops and hold managers in their units accountable for change. They can nudge supervisors to let people use flextime, for instance, or encourage fellow managers to join the recruitment visit to a Historically Black College or University. These task forces need to be permanent because there is no end to the challenges. Just when you’ve solved the problem of recruiting Latinx engineers, you may discover you can’t keep them. Our statistical analyses show that even after accounting for the effects of the programs those task forces put into place, creating a task force boosts management diversity on its own.

Appointing a permanent diversity manager is another way to guarantee that scrutiny of the system, and reform, are built into the organizational culture and routines. Like a task force, a diversity manager can diagnose problems, identify solutions, and put them into place. The mere presence of a diversity manager can also alert line managers that they may be asked about personnel decisions. Asking after the fact may not get Marketing to reverse the decision to hire Chad rather than Carla, but the next time around, Marketing may take more care to give everyone a fair shot. Managers change their behavior when they know that someone is keeping track of their decisions. As with task forces, our statistical analyses show that hiring a diversity manager boosts management diversity above and beyond the effects of the particular programs they put into place.

When diversity task forces and managers examine the workforce data from HR, their aim may be to spot and solve problems quietly. But lately more firms have been sharing their basic workforce data with the world, posting the numbers online and even setting public goals and dates for achieving them. Public scrutiny is not a solution in and of itself, but it can be a key to committing managers to reform. To be part of the solution, every manager has to know where the company stands and where it needs to be. A little hand-wringing about your company’s slow progress may go a long way toward motiving executives to figure out solutions. When companies have specific goals for change, active diversity managers, and well-designed diversity task forces, our research shows, diversity among managers rises.

The upshot

In this article, we’ve focused on the programs that actually boost diversity. The upshot of our analysis is that some programs that are not widely used these days can be strikingly effective at promoting diversity if implemented correctly. Most of the programs we have discussed are not diversity programs. Mentoring and skill training are standard talent management practices, but when implemented in a way that allows access to everyone, they also increase diversity. Work-life programs are not typically seen as DEI programs that help support Black, Latinx, and Asian American men, but they do just that. 

When it comes to explicit diversity programs, such as targeted recruitment, employers sometimes fear backlash. Sadly, many executives have resisted creating recruitment programs that target diverse candidates for fear of alienating white men in their workforces. Surveys have consistently shown that white men do not favor special programs to open opportunity to women and people of color. But at the same time, they show that white men who work for firms with such programs actually support them. The bigger lesson here is that firms should not worry about backlash against diversity programs that get managers involved in solving the problem.

We find that rewiring employee networks through mentoring programs significantly boosts manager diversity. But many companies still lack mentoring programs that are open to all. The first question we ask managers who tell us that their program doesn’t advance diversity is, “Do you have a system for matching mentors and protégés?” If the answer is no, because “mentors and protégés find one another and we support them,” then they don’t really have a mentoring program, which means that plenty of women and people of color are left out. To democratize mentoring, you can’t rely on informal connections—you have to offer to make matches. 

Democratizing career resources continues with opening skill and management training to all. In firms with only informal training, only white men may get additional training. Without skill or management training, women and people of color, often placed in dead-end jobs, have no chance to move into supervisor roles.

We found that cross-training programs, which rotate workers through departments and jobs, can be especially helpful for diversifying management. Employers typically implement cross-training for other purposes—to reduce turnover or improve worker aptitudes and flexibility. But cross-training is a secret weapon for promoting diversity through training. It gives women and people of color a chance to try out jobs that they had never thought of and that managers had never thought of placing them in.

The United States, unlike its European counterparts, leaves most work-life supports up to the employer. Flextime, paid leaves, and childcare are rarely provided, or even supported, by government policies. But it has become increasingly difficult for today’s workers to juggle work and family demands. Advances in technology, growing competition, globalization, and the decline of unions have expanded work demands. And the changing nature of the workforce, with the rise of dual-career couples and single parents, has expanded demands on the home front. Formal work-life programs, from flexible hours to family leave to childcare benefits, can make it easier for workers to balance work and family needs, and can signal that workers will not be penalized for doing so. These programs help white women, and Black, Latinx, and Asian American women and men, move into management.

Firms can institutionalize change by hiring a diversity manager and creating a diversity task force, the goal being to analyze career patterns and introduce changes needed to democratize elements of the career system. Diversity managers and task forces can also make middle managers feel accountable for myriad decisions about the careers of their underlings. Diversity managers often tell us they can’t tell managers whom to hire or whom to keep on during downsizings. But they can ask managers why they hired or laid off this one, and not that one, after the fact. Task force members can play a similar role, making colleagues in their home departments feel accountable. Diversity task forces have the added benefit of engaging managers from across the firm in making diversity their own problem to solve.

Diversity comes in many forms, and so do its obstacles and solutions. To make real change in an organization, the trick is to inspect existing career systems to detect obstacles they create for groups; tear down those obstacles to democratize access to positions of leadership; and build systems to sustain change. 

---

Frank Dobbin is Henry Ford II Professor of the Social Sciences and sociology department chair at Harvard University. Dobbin studies organizations, inequality, economic behavior, and public policy.

Alexandra Kalev is associate professor and chair of the department of sociology and anthropology at Tel-Aviv University. She studies organizations, work, law, inequality and diversity across national, organizational and economic contexts.

1. We’d like to be able to look at other groups—LGBTQ+ workers, differently abled workers, older workers, indigenous American workers—but data on those groups is not collected by the government, or the groups are too small to show effects in statistical models. [↩]