Whether sorting packages in the mailroom, coding in Python, or tending to patients, everyone starts their career somewhere. Yet in most industries, a first job is merely an entry point. What happens over the course of a career is crucial to building an individual’s human capital.
Human capital is formally defined as the knowledge, skills, competencies, and attributes that individuals possess.1 Its accumulation begins in childhood and continues throughout educational stages and working life. The value of human capital is realized when people put it to work—that is, by gaining work experience—and pay is an important signal of that value.2
Work experience, pay, and human capital itself are linked in complicated ways, and the threads are hard to unravel. One worker may enjoy pay hikes as she moves from one role to another and acquires additional skills, a pattern of work experience that enhances both her human capital and the way it is valued through her pay. But another worker may see her human capital eroding over time as her skills go unused in a lower-paying role that doesn’t require them. In this case, both human capital and its value diminish. Meanwhile, two workers who started out possessing similar skills may go on to earn differing levels of pay when they switch into roles with varying organizational and industry characteristics, indicating that the same human capital is valued differently in the two jobs.
Overall, however, work experience is vital to both individuals and economies. For individuals, work experience underpins nearly half of lifetime earnings. For economies, work experience reflects how effectively human capital is matched with employers’ needs to raise productivity. In this context, comparing the work experience trajectories of men and women assumes its importance.
The gender pay gap can be measured in myriad ways, varying by the metrics used or the groups compared. The most commonly cited statistic for the gender pay gap in the United States indicates that women earn 84 cents for every dollar men earn.1 However, this figure considers only full-time workers and measures only annual median wages. When including part-time and seasonal workers and measuring either annual median wages or hourly wages, larger gaps, ranging from 17 to 22 percent, are generally observed.2 The gender pay gap can vary by age cohort, and some studies find that older women tend to experience larger pay gaps than younger women.3
In our analysis of the gender pay gap, we take a more comprehensive approach than most studies by including workers of all ages, both part time and full time, as well as those in seasonal and nonseasonal jobs across the United States. Instead of focusing on hourly wages or median annual wages, we examine mean annual wages. We estimate the pay gap to be 27 percent, observed in year ten of an average career in our data set (see sidebar “Our data, scope, and methodology” for details of our data, which skew toward higher-skill professionals). Interestingly, mean earnings data for the entire US workforce reveal a similar pay gap.4 Our chosen approach allows us to isolate the effects of various drivers over time to determine how they contribute to the pay gap, which we explore in this article. It is important to note that the 27 percent gender pay gap applies only to people in the workforce, and women currently constitute 47 percent of that total.5 If we accounted for women who were not working, proportionate to their share of the working-age population, the income gap observed would be even larger.
Women have narrowed and even reversed the gender gap in education in the United States.3 Yet only 58.7 percent of women participate in the labor force, compared with 70.2 percent of men.4 And what women and men do at work diverges significantly over time. Our research finds that that divergence, over a decade or more of work, drives almost 80 percent of the gender pay gap of 27 cents on the dollar—what we call the “work-experience pay gap.” (See sidebar “Definitions of the gender pay gap vary.”)
To arrive at this conclusion, we analyzed how men and women go about accumulating work experience—switching jobs, returning after breaks, climbing the corporate ladder, making lateral moves, downshifting, and more—and how they realize the value of human capital differently (in terms of pay).5 This privileged, close-up view is possible because we use a data set of some 86,000 de-identified online career histories of real people in the American workforce. Our data set is overweighted with white-collar, higher-paying jobs because people with public-facing, online work histories are more likely than the general population to hold them—and they are of particular interest to talent-scouting employers.
This research focuses on the extent, nature, and impact of divergence in work experience patterns and its effect on the pay gap between men and women. While gender pay gaps have been well studied by other researchers, we add to the discourse by dissecting the dynamics of work experience gained over time (Exhibit 1).
Data sources. Our research uses licensed, de-identified data from US-based online public professional profiles to trace self-reported employment statuses and job changes. Our data set contains information on gender, education, job title, employer, and date, but it does not contain information on individuals’ names, locations within the United States, race, ethnicity, or other personal characteristics.
We selected a randomized sample of one million men and one million women that was gender identified based on a machine learning model that predicts gender using names, birth years, and name origin. From this data set, we selected a subset of individuals who declared their education, changed roles at least once, and had at least ten years of work experience following their highest educational attainment. This winnowed the sample down to 50,529 men and 35,235 women for a total of 85,764 individuals in the United States.1 It excludes those who may have exited the workforce altogether. This selection procedure in and of itself could result in a skewed sample, which we explore in more detail in the next section. These approximately 86,000 profiles include about 36,000 unique job titles, which can be mapped to 705 occupations classified by the US Bureau of Labor Statistics. Because sectors are determined at the company level, note that individuals can work in different sectors while performing the same occupation.
We integrate these individual-level career profile data with several occupation-level data sets for the overall US workforce, including wages and wage distribution by occupation from the US Bureau of Labor Statistics, gender wage gap within occupations from IPUMS (formerly Integrated Public Use Microdata Series), working hours from the US Census Bureau’s American Community Survey, and occupational characteristics (such as competitiveness, flexibility, physical demands, and more) from the Occupational Information Network (O*NET). We also validate our findings on the distribution of the workforce by seniority levels (entry level, manager, top manager, C-suite position) against McKinsey’s Women in the Workplace survey.2
Scope of data. Our sample reflects online work histories and skews toward men (59 percent in the sample, compared with 53 percent in the US workforce) and higher-educated workers in higher-paying occupations such as managers, STEM professionals, and business and legal professionals (exhibit). The sample of men is somewhat differently skewed compared with the sample of women, perhaps reflecting women’s lower representation in higher-paying occupations and, consequently, in online profiles.
We did not reweight our sample to mirror the US occupational mix, because predicting each stage of an individual’s career path across various occupations would be prohibitively difficult. Although this approach means our findings are not representative of the entire economy, it allows us to use reported data more directly and minimize assumptions. Even though our analysis covers the first ten years of a career, our findings for the next ten years have been directionally similar.
Methodology. In earlier research by the McKinsey Global Institute (MGI), we measured the value of human capital in terms of lifetime earnings, while in this research, we focus on annual wages. For our analysis, we built a data set with details on each worker for every year of work experience, along with their corresponding role, occupation, and wage. This data set allows us to distill insights on changing roles, occupational trajectories, career breaks, pay gaps, and more.
About two-thirds of workers in our sample started their careers after 2000, with about half entering after 2005. We use real annual wages in 2022 as the base year for average wages in any occupation in the tenth calendar year of a career, and then apply the 2022 gender wage gap by occupation.3 We use this approach because career pathway patterns are similar for women irrespective of their entry year into the workforce, and the same holds true for men.
To understand differences in career trajectory patterns between men and women, we categorize all occupations into quintiles based on their average wages in 2022. We then track the movements of individuals between these quintiles from the occupation they began their careers in to their occupation at year ten, keeping the quintile classification of the occupation constant throughout the entire period.
To analyze the drivers of the gender pay gap, we hold all variables except one constant to isolate its specific effect. For example, we hold wages constant while analyzing the wage distribution that women would have if they followed the same career pathways as men.
We separately control for average, occupation-specific working hours of men and women, and we model differential pay levels within each occupation based on role titles and tenures. We account for the impact of education level and type of degree on the pay gap by analyzing cohorts of individuals who began their careers in the same occupation and tracing their career arcs.
We also recognize that an individual’s age and years of experience can influence the pay gap. To control for this, we focused on individuals with at least ten years of work experience, examining their career trajectories from their entry into the workforce to year ten. As discussed earlier, we also validated our findings for a separate subset of approximately 27,000 individuals who had at least 20 years of recorded experience, distilling the drivers of the pay gap at year 20, and found that the results remained consistent, with continuing differences in career trajectories between men and women in the second ten-year period.
In this report, we typically present average trends for men and women, while recognizing that there are variations in these averages. We also checked for statistical significance across all metrics represented in this article and included only those that are statistically significant at a 5 percent level.4
Limitations. While we draw on a detailed and rich data set, our analysis reflects the challenges of working with de-identified, self-reported data. First, our data set does not capture all the nuances of how companies structure compensation or implement wage increases. Second, the accuracy of reported career breaks and their reasons can vary. There may be instances of unreported breaks within the same role that could have influenced tenure and career trajectories but are not reflected in our data. Third, we project future pathways based on historical trends without accounting for an externally determined skills supply, wage premiums, or other exogenous labor signals. Finally, our data include only binary gender classifications, implying that individuals identifying as nonbinary are likely categorized as either male or female.5
This methodology does not lead us to predictions; rather, it empirically demonstrates some of the differences in men’s and women’s work experience journeys. Labor markets are highly adaptable and may go on to behave in ways our methodology does not anticipate.
See the technical appendix for more details on the data and analysis undertaken in this research.
We emphasize that we do not directly investigate the reasons that they diverge. Women and men may intentionally choose to pursue different paths for a variety of reasons relating to opportunity and personal agency, with complex underlying factors that are difficult to untangle. For instance, personal preferences might lead women and men to opt for different kinds of work, or they may assign meaning to their work in different ways. At the same time, not all doors may remain open to all workers at every stage of life. As other research has explored, women may bear more responsibility for caregiving and household chores, while men may shoulder greater breadwinning responsibilities, which can restrict career choices for both.6 Whether or not those traditional or stereotypical responsibilities hold sway lies outside the scope of this article. (For details, see sidebar “Our data, scope, and methodology.”)
The gender pay gap highlights differences in how men and women realize value from their human capital. Over a 30-year career, we estimate, women earn about $500,000 less than men, on average.7 This loss of pay—and productivity, by implication—takes on particular importance in the context of tight labor markets and future demographic headwinds, with fewer workers potentially needing to support more retirees and fuel the nation’s economic engine. As automation and AI transform the nature of work and the skills required in the economy, optimal talent utilization is becoming a critical issue.8
Women actively change roles and traverse similar skill distances when compared with men, but they spend less time in paid work, on aggregate, and they navigate their careers in different directions (Exhibit 2).
The arc of work experience can be relatively smooth. But frequently, it involves stops and starts as people go about their personal lives alongside their careers, taking breaks to juggle work and nonwork priorities. We added up all the self-reported career breaks in our data set, looking at the directional patterns.1
We found that both men and women in our sample reported taking career breaks, defined here as a gap of one month or more between role moves. Perhaps unsurprisingly, women took breaks more often than men did, with as much as an eight-percentage-point difference in how often breaks occurred relative to all role moves at an estimated age of 37 years. With every break, women also spent more days out of the workforce, averaging an extra four months off per break.
Women’s career breaks are most apparent during their child-rearing years, likely linked to maternity leaves as well as caring for young and school-age children.2 The women in our sample with lower educational levels tended to take longer breaks from paid work compared with those with advanced degrees.3 Looking at other research, we infer that this may be due to the cost of care, which can be prohibitively expensive for lower wage earners in the United States, or perhaps related to the higher incidence of nonmarital childbearing among people without college degrees.4 Notably, we also see women taking breaks beyond their child-rearing years, at rates consistently higher than for men.
Household responsibilities are beyond the scope of our current study, but other research finds that male and female household members tend to allocate time differently on a daily and weekly basis.1 For example, one long-term study of fathers and mothers concluded that the proportion of time dedicated to paid work by fathers in the US workforce decreased from 42 hours per week in 1965 to 37 hours in 2011.2 Mothers increased their time spent in paid work from eight to 21 hours per week over the same period. Although fathers allocated some of this time away from work toward caregiving and household responsibilities, averaging about 17 hours per week in 2011, they still fell significantly short of the approximately 31 hours that mothers dedicated to these activities, according to the study.
Overall, in 2022, working women clocked 7 percent less time on paid work compared with men.3 This divergence is a combination of two factors: women hold a larger share of part-time jobs and work slightly shorter workweeks in their full-time jobs. (Of course, it’s important to remember that these are aggregate numbers; we also see women leaning in with more-than-full-time jobs and men taking on part-time roles.)4 Along with their career breaks—women take 42 percent more days on breaks than men over the course of a decade—we estimate that women accrued just 8.6 years of work experience compared with ten years for men, a 14-percentage-point gap in accumulated work experience.
The loss of work experience arising from both part-time work and career breaks could affect women’s promotions and pay in ways that compound over time. Based on career progression patterns in our database, we estimate that the differences in skills acquired by women compared with men through work experience account for $400,000 of their $500,000 earnings difference over a career.5
Despite accumulating less work experience in terms of time, women in our sample changed jobs at a pace similar to that of men after their first roles. Both men and women took on new roles, either within the same organization or at a different one, at a similar rate—an average of about 2.6 times over the ten-year period examined. For women and men alike, about 80 percent of role changes involved moving into new occupations and new organizations.
Based on our review of job postings, role moves typically come with new skill requirements.1 The share of new skills acquired and deployed that go beyond those exhibited in a prior role is what we refer to as “skill distance.” To be clear, traversing skill distance can happen when moving to a higher-paying job, a lateral position, or even a lower-paying job. If a computer programmer becomes a taxi driver, new skills are required.
In our analysis, women and men traversed similar degrees of skill distance in their career trajectories—another interesting similarity.2 More specifically, for both women and men, an average of 40 percent of the skills needed for a new job were different from those required in previous roles. Not only did women expand their skill sets with each new role as much as men in our sample, they also moved into new posts as early in their careers as men did. At least as reflected in these data, women are no more reluctant than men to enter unfamiliar career territory.
Even when workers traverse similar skill distances, not all job changes translate into better wages. For example, someone who started her career as an accountant might acquire new skills taking on an auditor role without a bigger paycheck—or become a financial manager and climb into a whole new income bracket.
Despite evidence of similar levels of willingness to try new roles, when men change occupations, we found that they are more likely than women to move to new occupations where the average pay is higher. When women change occupations, they are more likely than men to go into lower-paying fields. Of course, this trend reflects averages: There are men who fall behind and women who accelerate their earnings. In our sample, 20 percent of men moved into lower-paying occupations over ten years, while 33 percent of women advanced into higher-paying occupations during the same period. Outside our sample, which we have noted is overweighted with white-collar jobs and follows only the career pathways of people who remain in the workforce, other researchers have noted that some men without college degrees are dropping out of the workforce altogether, citing health reasons.1 In fact, while the median weekly earnings of all Americans increased over the past 40 years, the wages of men who did not go to college declined.2
Based on the income-level averages of the occupations of individuals in our sample, we group them into quintiles at the start of their careers and at year ten. That helps us observe a general pattern that holds across all five quintiles: men often climbed up through the quintiles into higher-paying occupations, while women were more likely than men to descend to lower-paying occupations by the tenth year.3 This might be because women are potentially overrepresented in the lower salary ranges within each quintile, given that they generally earn less than men on average within each occupation (a topic we explore in more detail in the next section).
Although half of the women in lower quintiles in our sample moved into higher-earning salary segments, they didn’t reach the top quintile as often as men. Considering men in our sample who started in an occupation in any income quintile excluding the top one, roughly 45 percent of them moved into top-quintile occupations over ten years, compared with only about 30 percent of women. That means men were about 1.5 times more likely than women to reach the top quintile, no matter what rung of the income ladder they started on.
Take our example of accountants, a second-quintile occupation: 46 percent of men who began as accountants had advanced to top-quintile occupations like financial managers by year ten, while 35 percent of women did so. Conversely, 5 percent of men transitioned to lower-quintile occupations like auditing clerks, compared with 10 percent of women.
Even when men and women move into the same occupation, slightly more men tend to hold higher-paying senior roles than women. In our sample, 3 percent of all women occupied the C-suite, compared with 6 percent of men, and 44 percent of women were managers, compared with 48 percent of men.
Pay isn’t everything, but it reflects the value realized from human capital. Women’s careers, even highly credentialed ones, tend to plateau when compared with men’s—and this translates into a pay gap over time. We investigate the link in the following section.
The many choices that women and men make over time lead to career arcs of decidedly different shapes. There are almost as many career trajectories as there are people in our sample. But looking at aggregate views, a few patterns stand out. While women earn more undergraduate and advanced degrees and start out strong, their career trajectories generally are flatter than men’s over time. Meanwhile, regardless of where men start out, they are more likely on average to climb into jobs that are more highly valued in the labor market.
Focusing on the overall US workforce in 2022 rather than our sample, higher-paying occupations tend to have more male representation than lower-paying ones. There are exceptions, from female majorities in high-paying radiology jobs to male majorities in low-paying bricklaying work. Yet grouped by occupational pay brackets, women in the US workforce hold almost 60 percent of the jobs in the very lowest bracket, which includes retail salespeople and assistants.
When considering total wages within occupations, women on aggregate tend to earn less than men within almost all job categories tracked, from hourly workers like retail salespeople to top-paying occupations like financial managers. Of course, many factors, including role-level seniority and tenure, can influence pay variations within occupations, as we discuss in the next section. That said, the occupational overview reveals a gendered pattern that indicates a 27 percent pay gap.
Our sample points to a 27 percent gender pay gap on annual mean wages observed in year ten of an average career. We track cohorts of men and women who start their careers in the same occupation and follow their career trajectories over time to isolate and analyze the defining traits of the paths they pursue.
The “work-experience pay gap”—attributable to both accumulating less experience (hours worked) and diverging career pathways—accounts for nearly 80 percent of the overall gender pay gap.
The difference in pay between men and women is apparent at the outset of their careers, as women often enter the workforce in lower-paying occupations. Fifty-one percent of men in our sample entered the workforce in a top-quintile occupation, compared with 35 percent of women. But first jobs don’t determine the future. Even starting in a low-paying job such as customer service representative, one worker might advance to become a top-quintile financial manager, while another might progress only to a third-quintile sales floor supervisor. We find that the effect of this initial occupational difference accounts for only a fraction of the gender pay gap—three percentage points—by year ten of work experience.
Pathways over time are crucial. If women had followed the same occupational trajectories as men starting from the same occupation, they could have narrowed the pay gap by eight percentage points. For example, among those who began their careers as sales managers, more men stayed and advanced in the field, while more women transitioned into support roles like training and development or administrative services managers.
Seniority levels vary even within the same occupation. Women in corporate roles often hold a smaller share of senior job titles than men—and less than their share at entry level. In our sample, among corporate workers, women made up 44 percent of entry-level positions, 39 percent of managerial roles, 36 percent of “top managers” such as vice presidents, and 24 percent of C-suite positions.1
Differences in seniority tend to be more relevant in large, multitiered organizations, where disparities in advancement between men and women can accumulate over time. This gap may be less pronounced when considering noncorporate roles. For example, among professional workers such as doctors and electricians, tenure matters more, and women tend to fall short on tenure, signaling lower in-role work experience. Put together, lower seniority and tenure account for at least three percentage points of the gender pay gap.2
Nearly a quarter of women hold part-time jobs, compared with 15 percent of men. And among full-time workers, women clock 1.5 fewer hours per week than men. Assuming a linear relationship between hours worked and pay, we find that labor input accounts for about a quarter of the pay gap. This is a conservative estimate because the linear relationship doesn’t usually hold for non-hourly-wage jobs, due to overtime and other factors.
Explanations for the last fifth of the gap are hard to tease out. One contributing factor could be differences observed in industries. Men and women in the same occupation may be employed in companies in different industries.1 For example, more IT managers in higher-paying tech sectors could be men, while those in lower-paying manufacturing sectors could be women. However, 85 percent of the workers in our sample were employed in occupations with no significant variation in industry representation between men and women.
Another factor is a potential wage penalty on an hourly basis for both hourly and part-time workers compared with non-hourly workers, even within the same occupations.2 This is likely because non-hourly workers often assume supervisory responsibilities. Additionally, workers who took long career breaks and returned to the workforce might be paid less than those who did not take breaks in the same job.3
We were unable to quantify some other factors that could explain the pay gap that other studies have explored. These include firm-level pay differences, such as more men being employed by higher-paying firms and more women by lower-paying ones.4 Unionized jobs, which typically don’t have a pay gap, might be more prevalent in sectors that employ fewer women.5 Additionally, traits like risk preferences and consequent gender-based differences in job searches could play a role.6 The field of study chosen before entering the workforce could also serve as an indicator of how individuals are differently positioned to succeed, even when starting from the same jobs.7 Finally, unconscious and conscious bias might lead employers to prefer men for better-paying jobs over equally qualified women or to pay women less for the same role.8
While individual stories vary widely, the big picture indicates that diverging occupational paths and shortfalls in accumulated work experience drove most of the pay gap. To explore why women tend to end up with lower-income jobs than men, we examined occupational characteristics.
Not all starting points led to pay gaps for women in our sample, although the vast majority did. Some starting occupations—compliance officers or travel agents, for example—led to ten-year career paths where women ended up outearning men who started in the same occupation. Some 10 percent of women were on these “pay premium” trajectories. However, more than half the women in our sample experienced large ten-year pay gaps due to their occupational trajectories. Examples include those starting out as medical assistants or nurses.1
Among measurable traits, two workplace characteristics defined by O*NET stood out as significant factors differentiating men’s and women’s trajectories: workplace competition and flexibility.2 The workplace competition score quantifies the extent to which jobs require workers to compete or be aware of competitive pressures, such as commission-based work for real estate agents. The flexibility score measures the “full-time nature” of an occupation, represented by the proportion of all workers in an occupation who are employed full time.3
We identified three starting occupation cohorts: those where women’s career paths lead to significant pay gaps compared with men, those resulting in moderate pay gaps, and those leading to pay premiums. We found that, on average, workers starting in all three cohorts ended up in occupations ten years later in which the full-time and competition scores were lower for women than for men, on average.
Women generally moved to occupations with lower workplace competition and more flexible work arrangements, and this was pronounced in trajectories leading to larger pay gaps. For example, consider the starting occupation of a brokerage clerk, in which men ultimately significantly outearn women over time. Women starting as brokers often seek roles like credit counselors, which have lower competition and full-time scores. In contrast, men starting as brokers tend to become investment fund managers more often, entering highly competitive environments.
We illustrate diverging career trajectories with examples drawn from the men and women who began their careers as tech professionals and went on to follow different paths. Here we group tech professionals into three subgroups: tech managers, tech engineers, and tech support. Overall, women make up about one-quarter of the workforce employed as tech professionals and experience a lower-than-average 23 percent pay gap. Yet they were more likely than men to leave this field, and when they left, more of them typically moved into lower-paying occupations.
Among the women who started out as tech managers, for example, only about one-third remained in the tech profession, compared with half of all men. Specifically, just 6 percent of women continued as tech managers, while 16 percent of men did. Among the women who left, about half transitioned to similarly high-paying roles, for example, marketing managers, while the rest moved to lower-paying positions, for example, community service managers. In contrast, of the men who exited tech, three-fifths (29 percent out of the 50 percent) shifted to other high-paying occupations.
Similarly, only about 40 percent of women who began their careers as tech support staff stayed in the tech field, compared with 60 percent of men. Among those who stayed in tech, a slightly higher proportion of women than men advanced to better-paying tech positions like engineers and managers. However, 40 percent of women who left tech (or 24 percent of all women) left for lower-paying occupations like customer service representatives, while only one-third of the men (or 12 percent of all men) did so.
Managers. In this male-dominated, high-paying occupational group, women were less likely to remain in management and more often shifted from frontline functions to lower-paying support roles. When women left management, they tended to secure higher- or similar-paying jobs like real estate broker or organizational consultant (Exhibit A).
Nurses. This women-dominated occupation represents an interesting difference. Women more often remained in these roles, with nearly half staying on as nurses, compared with just over a quarter of men. More male nurses advanced to even higher-paying occupations in the health field, such as medical and health services managers (Exhibit B).
Office support workers. Even in low-paying occupation groups likely to shrink due to automation and AI—such as office support—the trend is similar. Women who left were more likely to find work in other low-paying occupations, like counter attendants and customer service representatives (Exhibit C).
Production workers. A similar trend is on display in production work. When men left, they tended to transition to higher-paying occupations like procurement managers or industrial engineers. When women left, they were more likely to take on lower-paying work as administrative assistants or stock clerks, which are also being automated (Exhibit D).
Other examples also illustrate this trend. In the sidebar, we present two occupations currently dominated by women—nurses and office support—and another two similar to tech with male majorities—managers and production workers. Two of these are growing, as tech is—nurses and managers—and two are shrinking—office support and production workers. (See sidebar “Diverging career trajectories among managers, nurses, office support, and production workers.”)
The mix of occupations needed by America’s economy is evolving. Demand for workers in some fields—notably, healthcare, technology, and management—is expected to grow through 2030 as adoption of automation and AI technologies accelerates, while some roles in office support and production work will disappear in aggregate by then, according to previous MGI research.9 Millions of US workers will likely need to transition out of “shrinking” fields of work into “growing” ones.10
Despite their outsize presence in growing healthcare occupations, women are underrepresented in growing occupations overall. Only 59 percent of women are in these occupations, compared with 67 percent of men. Conversely, women are overrepresented in shrinking occupations.
Looking back, we also see that women moved into and remained in growing occupations less often than men in our sample. Women accounted for just 42 percent of the workers who transitioned from shrinking occupations to growing ones or stayed back in growing occupations over time.
For example, historically, more women than men have moved into healthcare, a fast-growing occupational group, and into creative and education industries, which are expected to grow more modestly. The question of why more men don’t enter these high-growth “pink-collar” occupations as often as women is also something to consider.1 At the same time, women are underrepresented by a wide margin among workers moving into fast-growing STEM jobs and management roles.
Employers can be expected to adjust wages and work arrangements to meet their future demands, providing an incentive for qualified individuals to move into growing occupations, as they have in the past. For example, in 2022, 18 percent of women in the workforce were in management fields, up from 15 percent in 2012. Similarly, the share of men in these fields increased, albeit at a slower pace, from 17 percent to 19 percent. Conversely, in shrinking fields like office and administrative support, the concentration of women decreased from 19 percent to 16 percent, while the share of men declined from 6 percent to 5 percent.
Overall, historical patterns suggest that women’s current underrepresentation in growing occupations could become even more pronounced over time.2 That is, while the proportion of women working in growing segments could rise to nearly two-thirds by 2030, the proportion of men could increase more rapidly so that more than three-fourths of them are working in growing segments by then. The disproportionate presence of men in growing and higher-paying occupations, coupled with the overrepresentation of women in shrinking occupations, could cause the pay gap to stay the same or increase over time.
Some companies are at the forefront of building human capital for all workers, both men and women, while at the same time delivering top-tier financial performance. We labeled those employers People + Performance Winners (P+P Winners) in previous research. In that research, we found that employees who pass through P+P companies go on to have higher lifetime earnings on average than those who did not.1 (See sidebar “What are ‘People + Performance Winners’ and what might they show us?”)
Now, we build a gender-disaggregated view for 12,000 career histories spanning 1,100 US-based employers. We find that, adjusting for industry mix, women who passed through the halls of P+P Winners at any time in the first ten years of their careers ended up with higher average salaries and lower gender pay gaps than those who had passed through People-Focused Companies, Performance-Driven Companies, or Typical Performers.2 Although these are correlations and we cannot directly ascertain the causes, we found that, on average, both men and women who worked for P+P Winners moved more often into growing occupations, experienced more internal role moves, and had longer tenures at the company. In our analysis, we did not find higher women’s representation in P+P Winners per se; instead, we identified a strong emphasis on both expecting and facilitating skills development for all employees, including women.
Importantly, P+P Winners provide workers with clear organizational vision, translated for the specific contexts of individual teams and roles. They offer a performance culture with transparent expectations and incentives that challenges employees while empowering them, as well as a focus on innovation that encourages risk-taking with appropriate coaching. By contrast, People-Focused Companies excel on human capital development measures like internal rotation and training but lack P+P Winners’ performance-oriented culture, while Performance-Driven Companies focus on defined performance goals, efficiency, and a strong external orientation rather than empowering employees. The distinctive organizational culture of P+P Winners offers lessons to employers intensifying efforts to attract talent and reskill workers given trends expected for the future of work.
In MGI’s 2023 report “Performance through people: Transforming human capital into competitive advantage,” we introduced a framework that classified 1,800 global companies according to two factors relative to their sector peers: (1) how much they focused on developing human capital; and (2) whether they financially outperformed.
We classified the majority of the companies we analyzed as Typical Performers, which do not stand out in either dimension; a small subset as People-Focused Companies, for the significant resources they devoted to developing employees, even if performance was unremarkable); and another subset as Performance-Driven Companies, for their top-tier financial results, even if human capital development was average or typical. But it is the fourth set, People + Performance Winners (P+P Winners)—about 10 percent of the companies we analyzed—that we want to draw attention to here.
P+P Winners were the select subset of firms that we saw excel at creating opportunities for their employees to build skills, which we measured through internal mobility opportunities, training hours, and organizational health scores. At the same time, this subset consistently cleared the highest bar for financial performance. We found that P+P Winners spanned all sectors in our sample and showed greater resilience and more consistent earnings relative to their peers.
Drawing on our previous research, in this work we updated data for the 1,000 US-based global businesses analyzed previously, added about 100 US-based companies, and categorized them into the four subsets described above. Of the approximately 86,000 individuals in our present study sample, we found that 7,409 men and 5,067 women (12,476 unique workers in total) had worked at one of the 1,100 companies analyzed at some point in their careers. We note that more than 50 percent of the individuals employed by these businesses in our sample were in management or STEM professional positions; that is, they were not representative of the workforce as a whole.
For those 12,000 individuals in our sample, we analyzed the occupational trajectories of both men and women by the type of company they passed through. We studied which occupations they moved to (whether growing or shrinking), the number of role moves they made within the company, the time they spent in the company, and the associated gender pay gaps.
We did the same for workers classified based on which type of company they began their careers at, where they had worked early in their careers, and which one they had worked at the longest. Our findings about the relative benefits of P+P Winners remained consistent across all four approaches.
Drawing from the P+P Winner advantages observed in our study samples, some broad ideas stand out that could be useful to other employers. P+P Winners cultivate internal mobility opportunities that can build skills and retain talent. They also target training and apprenticeship programs, especially for midcareer talent. Other evidence documents the promise in these approaches. For example, skill-enhancing opportunities within organizations have increased employee satisfaction and retention.11 According to LinkedIn’s Global Talent Trends 2020 research, employees tend to remain at a company 41 percent longer if the company regularly promotes from within.12 And a recent Gallup survey suggests that midlevel and mid-tenure employees, who constitute the majority in most organizations, would benefit from targeted training and development programs.13 As labor markets remain tight amid demographic shifts, employers should keep in mind that workers will have more options in choosing their work and choosing their employer. Both men and women can, and will, make strategic career moves that help them grow. Each individual’s choice is their own, but employers can do more to position themselves for this future.
This research was led by Anu Madgavkar, a McKinsey Global Institute partner in New Jersey; Kweilin Ellingrud, a McKinsey senior partner and director of MGI in Minneapolis; Sven Smit, a McKinsey senior partner and chair of MGI in Amsterdam; Chris Bradley, a McKinsey senior partner and director of MGI in Sydney; Olivia White, a McKinsey senior partner and director of MGI in San Francisco; and Kanmani Chockalingam, an MGI Fellow in San Francisco.
Kanmani Chockalingam led the working team, which included Ananya Sivaraman, Dapo Folami, Mackenzie Manofsky, Paige Hasebe, Rebecca Solcia, and Sirui Wang.
This project benefited immensely from the perspectives of McKinsey colleagues. Our thanks go to Dana Maor, Davis Carlin, Eric Kutcher, Jason Bloch, Mekala Krishnan, Sarah Gitlin, and Tracy Nowski.
We are grateful to the academic advisers who challenged our thinking and sharpened our insights: Sir Christopher Pissarides, Nobel Prize winner and Regius Professor of Economics at the London School of Economics; Matthew Slaughter, Paul Danos Dean of the Tuck School of Business and the Earl C. Daum 1924 Professor of International Business, Dartmouth College; and Jesús Crespo Cuaresma, professor of macroeconomics at the WU Vienna University of Economics and Business.
The article was edited and produced by MGI senior editors Cintra Scott and Stephanie Strom, together with senior data visualization editor Chuck Burke. We also thank David Batcheck, Rachel Robinson, Rebeca Robboy, and Rishabh Chaturvedi for their support.
We thank McKinsey Global Publishing for designing this visual narrative, including team members Charmaine Rice, Dana Sand, Diane Rice, Katrina Parker, Mary Gayen, and Nathan R. Wilson.
As with all MGI research, this work is independent and has not been commissioned or sponsored in any way by any business, government, or other institution. While we gathered a variety of perspectives, our views have been independently formed and articulated in this report. Any errors are our own.
A person’s first job is just the first step in a long journey. Work experience matters and defines the quality of workers’ career arcs as well as their pay. Gender pay gaps that grow over time reflect different choices made by men and women in utilizing and building their human capital. Companies can influence those choices by fostering organizational cultures that emphasize role mobility and skill building for all workers, ensuring that both women and men realize more value from their human capital over their careers at the same time that they prepare for the future of work.
Tough trade-offs: How time and career choices shape the gender pay gap – McKinsey & Company
