Why Can’t Your Company Just Fix the Gender Wage Gap?

Beyond the stalling, waffling, denying, punishing, and ignoring, some answers.

Natasha Lamb of Arjuna Capital.

Photographer: Laurel Golio for Bloomberg Businessweek

At Citigroup Inc.’s annual meeting in New York City in April, after Chief Executive Officer Michael Corbat had summarized the company’s activities over the past year—1 million cardholders added! $23.6 billion in climate-change-reducing investments!—Citi Chairman Michael O’Neill gestured to two microphones placed in the audience and invited shareholders to speak. This was Natasha Lamb’s cue. She stepped to the mic.

Lamb is a 34-year-old managing partner at the Boston investing company Arjuna Capital. Her investing philosophy at Arjuna, which calls itself an “enlightened” firm with a $200 million portfolio, is that environmental sustainability and social equality lead to long-term profitability. For the past year she’s been systematically submitting shareholder proposals on behalf of Arjuna’s clients, first to nine tech companies and then to six banks, asking them to analyze and make public any differences in the salaries of men and women they employ. The week after the Citigroup meeting, Lamb planned to make a similar presentation at American Express Co. The annual meetings of Bank of America, Wells Fargo, Mastercard, and JPMorgan Chase would soon follow.

Shareholder proposals to Citigroup are limited to 500 words, and Lamb had only a few minutes to state her case. She started by reciting statistics: A study by the job-search website Glassdoor showed that even when accounting for education and experience, women in finance make about 6.4 percent less than men with the same jobs. At Citi, specifically, the crowdsourced compensation data company PayScale Inc. put that number at 6.1 percent. Would Citigroup prepare a public report on any pay gap the company may or may not have?

Lamb had barely finished speaking when another shareholder, a retired budget officer from Ridgewood, N.J., named Russell Forenza, rose to announce his formal opposition to her idea. Forenza started with an anecdote about his daughter, who holds a degree in chemistry but teaches high school for a living and makes less money than she might otherwise. “Sometimes it’s personal decisions,” he said. “Or they leave and have children. That greatly affects their earning power. Men don’t have to do that, at least not yet.”

Lamb responded to Forenza. “There’s a gender pay gap in our country that arises perhaps from ‘personal decisions,’ or having children and leaving the workforce. That’s one thing,” she said. “But the disclosure we’re looking for is like positions, like seniority.”

Featured in Bloomberg Businessweek, June 26, 2017. Subscribe now.
Photographers (clockwise from top left): Damien Maloney, Laurel Golio, Lucas Foglia, and Carlos Saavedra, all for Bloomberg Businessweek

She turned to O’Neill and continued. “Your U.K. peers are putting up the whole gap, saying, ‘We pay this much to men in our whole company, we pay this much to women, and there is a 30 percent gap.’ ” Lamb was referring to the British banks Virgin Money Holdings UK Plc and Schroders Plc, which had recently disclosed that men make 36 percent and 31 percent more money than women at their companies, respectively. But those figures represented the difference between median pay for all men and all women—and men, as is the case at most financial companies, held a disproportionate number of the highest-paying positions, while women tended to be grouped in lower positions. This is also true for Citi, where almost three-fourths of its administrative and clerical workers are women, a figure that’s remained unchanged for almost 20 years. “We’re not even asking you to go that far,” Lamb said. “We’re asking for equal pay for equal work. What is the gap?”

O’Neill said he didn’t see a point in conducting such a study. “The issue is, you want a number,” he said. “It is a simplistic way to address this problem. There are so many factors that make that one number not terribly relevant without a lot of analysis behind it.”

“We’d be very pleased with multiple numbers and detailed analysis,” Lamb replied. But Citi didn’t want to offer that up either. When Lamb took her seat, the company asked stockholders to vote down her proposal.

Over the past 50 years, women have achieved an astounding level of equality in the U.S. They have become astronauts and U.S. Supreme Court justices and have come so close to winning the presidency it’s easy to forget that until 1974 they couldn’t get a credit card unless a man co-signed the application. Women are the primary breadwinners in half of all U.S. families. They’re more likely to hold a bachelor’s or master’s degree than men, and should they choose to marry, about a third of them will keep their last name. But one of the most intractable and measurable differences between women and men in the U.S. is the kind of jobs they hold and how much they’re paid. The most frequently cited figure when discussing this gender wage gap is that full-time working women in America earn roughly 80 percent of what men do. But that number, though true, doesn’t offer much insight into the underlying social and economic forces generating the inequality, nor why it’s remained relatively unchanged for the past 20 years. It also doesn’t explain how incredibly difficult it is to change the status quo.

The current pay gap is a smaller, much narrower version of the one women have endured for generations. In the past, beliefs that women should be at home, or that they should be shielded from certain work, or doubts that they could perform jobs as well as men, led to discrimination in pay. In 1869 the U.S. Treasury employed 500 female clerks and paid them half the rate of men. Sometimes a world war would break out, and women would be allowed to hold jobs previously reserved for men; in 1918 the U.S. Employment Service published a list of what was considered acceptable “women’s work” (electric machine operator, automobile worker) during wartime. Labor unions often lobbied, sometimes successfully, for women to be offered men’s salaries, so when returning soldiers took their jobs back—which they did—employers wouldn’t balk at suddenly having to offer them their old pay.

Cindy Robbins, who, with Leyla Seka, led Salesforce’s pay-gap analysis.
Photographer: Laurel Golio for Bloomberg Businessweek

Until now, the main attempts at narrowing the gap have come from state and federal governments. Last year the Obama administration mandated that companies with more than 100 employees report salary data broken down by race, ethnicity, and gender. A 2016 California law requires companies to justify why employees who do “substantially similar” work are paid differently. In Massachusetts companies can’t ask prospective employees about their salary history; the idea is that unjustifiably lower pay will carry over from job to job if new salaries are based on old ones. A similar law in New York City takes effect on Oct. 31.

But legislation and executive action—not to mention enforcement—are subject to the whims of the controlling political party and can only do so much. California’s law can’t change this: According to a 2017 University of Southern California survey, only 4.2 percent of film directors are women. And despite President Obama’s support for equal pay, an American Enterprise Institute analysis found that the average pay for women in his administration was about 11 percent less than men’s. (The institute hasn’t run an analysis of the Trump administration.) If laws were a sufficient fix, none of this would be an issue, because gender discrimination at work has been illegal under the Equal Pay Act since 1963.

There are two ways of looking at the wage gap. The first is to take a bird’s-eye view. The most significant reason women still make 80 percent of what men do is that they’re clustered into lower-paying fields, as seen at Citigroup. According to the Institute for Women’s Policy Research, of the 20 most common careers for men and women, only four overlap. In other words, men still work in primarily male-dominated fields and women in female-dominated ones. Roughly 79 percent of elementary and middle school teachers are women; 80 percent of software developers are men. Industries such as law and medicine are more mixed, but most secretaries and administrative assistants are still women (94 percent), and most electricians are men (97 percent). The traditionally male-dominated fields pay better—so much better, in fact, that a man with just a high school diploma is likely to make more money than a woman with some college education or even an associate’s degree.

Salaries, of course, aren’t based on a set of immutable scientific facts. Are female-dominated occupations worth less to society? Are they easier to do? It’s true that electricians, who earn on average about $57,000 per year, according to the Bureau of Labor Statistics, need vocational training to ensure they can properly wire buildings. But anyone who’s struggled to find a good teacher to educate their child ($55,000, despite requiring a bachelor’s degree) or round-the-clock care for an elderly parent ($21,900) knows the value in that work and the skill required to do it well.

Women aren’t earning less money solely because they pursue or are pushed into low-paying work. According to a 2009 study published in the academic journal Social Forces that looked at 50 years’ worth of U.S. Census Bureau income data, when a substantial number of women move into a field, as has happened in biology and design, the average wage drops for everyone in that field, men included. The opposite can be true, too. In the early days of computer science, when programming was considered a monotonous, low-level task akin to typing or filing, there were more women in the field, and the pay was much lower. Now that we think about it as a way to design the technological future, more men are entering computer science, and the pay has risen. The share of women pursuing degrees in the field has dropped from a high of 37 percent in 1984 to 18 percent today.

Career clustering leads to an even wider pay discrepancy when the gap is broken down by race. For example, black and Hispanic women are more than twice as likely as white women to work in the service sector. As a result, they make about 63 percent and 57 percent of what white men do, respectively. (White men are still the largest subgroup in the labor force, so they’re often used as the default against which other race and gender combinations are measured.) According to employment data that Citigroup has provided to the U.S. government, about 60 percent of black women who work there are in administrative or clerical roles, a figure that has remained relatively constant since at least 1999. It’s often debated whether this segregation stems more from a lack of opportunity or outright discrimination—though considering the rising frequency with which blacks and Latinos are obtaining college degrees, it’s unlikely to be attributable to aptitude or preference.

The second way to analyze the wage gap is to zoom in on job-to-job differences in pay—what Lamb is pushing companies to disclose. Do men and women, working in the same job, with the same level of education and experience, make the same amount of money? You can slice women’s careers any number of ways—by industry, by specialty, by seniority level, sometimes even by company—but all too often the answer is no.

A study published in July in JAMA Internal Medicine analyzed salaries of 10,000 physicians employed by medical schools and found that on average, women made $51,000 a year less than men. The difference wasn’t only because fields such as gynecology and pediatrics, which are largely female, pay less (though they do). The gap also occurred within specific fields. Female orthopedic surgeons made $41,000 less than male orthopedic surgeons. The salaries of female oncologists were $38,000 less than those of male oncologists. And so on. This pattern isn’t unique to medicine. Last year, Glassdoor reported that even after accounting for age, experience level, and education, the average salary for a computer programmer drops 28 percent if that programmer is a woman.

Most compensation gaps in comparable jobs aren’t that large; both Glassdoor and the nonprofit research company Catalyst Inc. estimate that across the U.S. economy, it’s only about 5 percent. Still, it’s a pretty consistent finding. And year after year, that lost 5 percent begins to add up.

And that’s assuming the disparity doesn’t grow. According to the Census Bureau’s 2015 American Community Survey, women make almost the same as their male colleagues just out of college and when they are 22 to 27 (97 percent). Low-wage workers also make similar amounts because, as Randy Albelda, an economics professor at the University of Massachusetts at Boston, puts it, “those jobs are lousy for both men and women.” But the gap increases as women age and also as they advance into higher-paying careers. “When women start to have kids, it gets really obvious,” she says. A recent paper by the National Bureau of Economic Research that looked only at married men and women found that on average, college-educated men made about $355 per week more than college-educated women.

Leah Farmer, a senior director in software development.
Photographer: Laurel Golio for Bloomberg Businessweek

Forenza, the shareholder who countered Lamb’s request for more salary transparency at Citi, might ascribe this white-collar discrepancy to “personal decisions,” but for many women, it’s often not a choice that’s theirs to make. Child-care costs sometimes cause women to drop to part-time work or stagger their hours so they leave the office early and work from home at night when their children are in bed. A schedule shift theoretically shouldn’t affect one’s pay, but work done at home isn’t as visible, and therefore as appreciated, as that done in the office. It appears that women work fewer hours than men; according to the U.S. Department of Labor’s 2015 American Time Use Survey, for full-time workers, that difference amounts to 24 minutes per workday. Of course, most full-time workers aren’t paid hourly.

Women aren’t working 24 fewer minutes per day because they’re less dedicated to their jobs. According to the survey, the extra time is going toward housework or child care, the majority of which, even in dual-income households, is still provided by the woman. In fact, as of last year, women were doing almost double the amount of cleaning, cooking, and child care as men.

It’s not easy to change something as complex and ingrained as pay inequity. Doing so requires an overhaul of so many cultural norms—the factors that lead to career clustering, assumptions about how women should look and act, working parenthood as it relates to office hours and promotions—that sometimes it can feel out of any one person’s, or company’s, control. Still, some companies are trying—not out of some altruistic sense of doing the right thing (though their public-relations departments will often tell you that), but because in a country with a labor force that’s almost half women, it’s getting harder to ignore. And, companies are increasingly discovering, it makes financial sense. Fairness can be fruitful.

Women make a disproportionate share of consumer decisions in the U.S. Boston Consulting Group estimates that they control about 65 percent, or $12 trillion, of the consumer-spending market. That figure includes, along with single women and ones in same-sex relationships, the female halves of heterosexual partnerships, who are still more likely to do the daily household shopping. Women also make the majority of auto-purchase decisions. The country has already benefited from their increasing economic power: McKinsey & Co. estimates that about a third of the U.S. gross domestic product can be attributed to women’s increased participation in the labor force since the 1970s.

Of course, that’s the macroeconomic view. On a more granular level, the experiences of Ida Tin, the Danish entrepreneur behind the period-tracking app Clue, highlight how financial opportunities can be missed in the absence of a diverse workforce. Tin started her first round of fundraising in 2012, pitching Clue to investors, who were usually men. “A lot of investors said, ‘I like investing in things that I can use myself.’ ” They had no need for a period-tracking app. In her first round, Tin raised only $56,000. Five years and 5 million users later, her last round brought in $20 million.

Buffer, a 75-person social media management service, is aware that its homogeneous workforce still puts it at risk. “We had basically the typical tech company problem: the vast majority of men in engineering jobs and the vast majority of women in marketing and customer service jobs,” says Courtney Seiter, who holds the decidedly nonengineering title of director of people at Buffer. In 2013, Buffer decided to publicize its employees’ salaries, under the assumption that if everyone at the company knew what everyone else made, nobody would be treated unfairly. So the company was taken aback last year when an internal audit revealed that women made on average $9,500 less per year than men. “We just assumed that when we ran the numbers we’d see parity. We were very surprised when we didn’t,” Seiter says.

A lot of Buffer’s gap stemmed from both the gendered clustering problem Seiter mentioned and the fact that men held most of the highest-paying positions at the company. At the time, Buffer hadn’t done much active recruiting, relying instead on employee referrals. But those were often—surprise—demographically similar to Buffer’s existing employees. To mitigate that, the company contacted organizations such as Women of Color in Tech and 20/20 Shift, a consulting and training startup that aims to help tech companies find and hire more women and minorities.

Buffer found a job-to-job gap, too. Salaries at the company are determined by a number of factors, including the somewhat arbitrary idea of experience level. “There was a lot of guessing on that one,” Seiter says. On the whole, managers tended to give less weight to women’s experience than men’s. She says Buffer couldn’t come up with a reason it was ranking women this way. “I can only say this anecdotally, but from my experience, men would be more likely to push back a little on how they were rated, and women said, ‘Oh, wherever you put me is fine. I’ll just work hard, and you’ll notice me that way,’ ” she says. So Buffer scrapped its old evaluation system and created a method based on an employee’s clear, technical understanding of the company’s products.

Buffer admits it still has work to do. Despite participating in one of 20/20 Shift’s classes—students designed social media campaigns for Buffer products—it’s yet to hire someone through the group. Buffer is still 77 percent white and 69 percent male (90 percent male in its tech department). It’s hired or promoted a few female executives and managers; one is even director of engineering. Now, at the managerial level or higher, women actually make more than their male counterparts. But because of the persistent clustering problem, the average man’s salary is $95,221—still $2,404 more than a woman’s.

Leyla Seka.
Photograph: Laurel Golio for Bloomberg Businessweek

On a larger scale, the San Francisco-based cloud-computing giant Salesforce.com Inc. has also struggled with a persistent wage gap. In 2015, on the advice of two female executives, Leyla Seka and Cindy Robbins, an unusually detailed analysis of its pay data found that about 6 percent of its employees, then about 1,020 people, were being inexplicably underpaid. The gap couldn’t be accounted for by the number of hours they worked, their education, seniority, or how good they were at their job. So Salesforce spent $3 million adjusting salaries to bring everyone up to par. This year it added race to its analysis and spent another $3 million adjusting the salaries of 11 percent of its employees. Salesforce says it needed to do this second rejiggering because last year it acquired 14 companies and their 7,000 employees, thus inheriting their preexisting pay gaps. When Salesforce delved into the reasons for these gaps, it found they often began during salary negotiations. “If your starting compensation is negotiated from your pay level at your previous company, how do you negotiate effectively based on that?” says Robbins, Salesforce’s executive vice president for global employee success. Salesforce has stopped the practice of pegging new hires’ salaries to their old ones.

Buffer and Salesforce are relatively young companies that have been tackling the wage gap for only a few years. Corning Inc., the industrial glass and ceramics manufacturer in upstate New York, provides a rare opportunity to look at challenges over the long term. In 1985 an internal survey revealed that women were quitting the company at twice the rate of men. Corning was wasting $3.5 million a year on additional training and recruiting to replace them.

In exit interviews, women said they were leaving Corning because they didn’t have opportunities to advance; at the time, almost all managers and executives were men. So the company put in place simple metrics to make sure women were promoted at the same rate as men. Some women also said they struggled to balance child care with a full-time job. In 1988, Corning started subsidizing day care, a practice it continues today. (The company underwrites the operating expenses for certain day care centers and allows employees to pay a sliding scale based on their income level.) Eventually, female attrition slowed. For years now, it’s been the same as men’s.

In 2004, Corning made a push to hire more minorities and women by using Census Bureau employment data to determine what the gender and racial makeup of its applicant pool should look like for any open position. It then asked recruiting companies to send it an appropriately diverse set of applicants. It took a decade, but with this technique, the company has tripled its number of minorities.

Corning, though, is part of a heavily male-dominated industry. Despite all its efforts, only a third of its North American workers are women. Still, those hiring and promotion metrics have ensured that 30 percent of its corporate officers and 37 percent of its managers are women (or minorities), too. Four years ago, Corning teamed with a consulting company that specializes in pay-gap analysis; it found a slight gap at 99.2 percent that seemed to coincide with women’s tendency to fall lower on the predetermined pay ranges for any given job. “We worked on that for about three years and can now say we’re at 100 percent parity,” says Christy Pambianchi, vice president for human resources, who’s been with Corning since 2000. Her team runs its pay-gap analysis four times a year to avoid any large, retroactive adjustments.

Achieving pay parity while also addressing clustering has required decades of persistence. Pambianchi likens it to an attempt to lose weight. “You can read magazine articles about how to lose 10 pounds; it’s pretty simple,” she says. “But doing it is a real challenge. You’re correcting long-held habits. Everything about this is hard.”

Companies that are motivated to make these kinds of changes, or at least talk about the problem, are the exception. For many women, even detecting the wage gap in the first place can seem like an act of subversion.

“I only found out because someone slipped me an anonymous note in my mailbox with the names of three guys I worked with and their salaries,” says Lilly Ledbetter, who worked as a production supervisor at a Goodyear Tire & Rubber Co. plant in Alabama for almost 20 years before finding out she was earning 25 percent to 40 percent less than her male co-workers. “It wasn’t just the salary. The contributions to my 401(k) were tied to how much I made. My Social Security check is, too.” Ledbetter sued for back pay in 1999, but she ultimately lost because she hadn’t filed a complaint within 180 days of receiving her first unequal paycheck. Her case prompted Congress to pass the Lilly Ledbetter Fair Pay Act in 2009, which changed the 180-day rule so that it reset with every paycheck, extending workers’ ability to take action when they discover they’ve been underpaid. But they still have to uncover that discrimination on their own. “If I hadn’t received that anonymous note,” Ledbetter says, “I’d never have known.”

“I was involved in hiring, so I knew what people were making,” says Leah Farmer, who’s worked at Expedia Inc. and Amazon.com Inc. among other places and who now leads a software team at a large nonprofit hospital system. “Earlier in my career I thought, Maybe I’m just not ‘tech’ enough, or I’m young. I assumed it would eventually just get fixed.” Farmer says she received promotions and raises as her career progressed. But as she got older, the gap between what she made and what her male colleagues made widened. Farmer finally realized that someone working under her and with many fewer years of experience was making only $5,000 less a year than her. “I don’t think it was malicious, like, ‘Ha-ha, we can pay her less,’ ” she says, “but I think there was this idea that for whatever reason, this is what I was worth.” (Both Amazon and Expedia have publicly said that men and women in equivalent positions make the same amount of money.)

A former vice president at the Disney Channel, who asked that her name be withheld because she signed a nondisclosure agreement and still works in Hollywood, says her first sense that something wasn’t right came when a man transferred to her department from another area of the company. “I received a call from HR, and in normal course of business, because he was now reporting to me, they said, ‘This is his salary,’ and I was like, ‘What?’ It was more than I was making,” she says. Her suspicions about the pay gap within the company were reinforced when she was promoted and offered a salary much lower than she knew was typical for the role. She grew more concerned when she promoted her assistant, a young Latina in her 20s, to be a coordinator for a television show. The woman happened to have a close male friend, also in his 20s, who received the same assistant-to-coordinator promotion on the same day. When they compared salaries, the female coordinator found hers was inexplicably lower than her friend’s. She brought it to Disney’s attention but says the company refused to change it. The female coordinator, who also asked that her name be withheld, had to find another job offer—which took about six months—before Disney bumped her salary up to her friend’s level. Disney declined to comment.

Last year, when Arjuna Capital submitted shareholder proposals at nine tech companies, seven of them issued press releases saying they had virtually no difference in pay between women and men in the same roles. (Six came forward voluntarily, and a 51 percent shareholder vote forced EBay Inc.’s hand.) Microsoft Corp. and Apple Inc. in particular made headlines for reporting that they’d achieved equal pay, though neither company acknowledged that fewer than a third of its employees were women. (At the executive level, the percentages drop into the teens.) Facebook Inc. wouldn’t provide Arjuna with a pay-gap number; on Equal Pay Day in April the company announced—on Facebook, naturally—that it had completed “thorough statistical analyses” and found that men and women earn “the same.” Lori Goler, vice president of people at Facebook, says the company has been running pay-gap analyses focusing on job-to-job comparisons since at least 2009 and double-checks them with a third party.

Kellie McElhaney of the Center for Gender Equity & Leadership at Berkeley.
Photographer: Laurel Golio for Bloomberg Businessweek

Google Inc. remains silent. Last year it refused to supply the Department of Labor with the salary and employment data legally required of all federal contractors (Google, like many large companies, does business with the government), prompting the department to open an investigation that’s led to allegations of widespread gender discrimination in pay. “We found systemic compensation disparities against women pretty much across the entire workforce,” Janette Wipper, a Department of Labor regional director, testified in San Francisco court in April, the same month that Google put out a public statement claiming it had “closed the gender pay gap globally.” Google, which denied these accusations in court, declined to comment.

“I’m highly suspicious of companies that do their own data analysis internally, because you can slice numbers any way you like,” says Kellie McElhaney, an associate adjunct professor and the founding director of the Center for Gender Equity & Leadership at the University of California at Berkeley’s Haas School of Business. She’s seen all manner of corporate foot-dragging: CEOs who hire her, then refuse to meet with her; companies who brag about achieving pay equality but don’t address clustering issues; male executives who are asked to oversee gender-related programs simply because they have working wives or daughters.

“Don’t even get me started on the diversity speaker trail,” says Ariel Lopez, founder and CEO of 20/20 Shift. “I no longer want to be on panels, speaking about diversity. It’s the same thing over and over again. Everyone knows it’s a problem, so OK, what are your metrics? How many people do you need to hire to change this?”

At Citigroup’s annual meeting in April, O’Neill repeatedly emphasized the bank’s commitment to paying women and men the same. “We completely support the agenda,” he told Lamb. “But I don’t think there’s much point in beating a dead horse.” Lamb pushed him to explain why a company that claims to be committed to equality would be so reticent about checking to see if that equality really existed. “This has been wrapped up in a black box for so long, making it hard for women to see,” she said. Without specific numbers, everything women go through is just conjecture. A hunch. They needed the equivalent of Ledbetter’s anonymous note. “We just want to know what we’re dealing with,” Lamb said.

O’Neill assured her that Citigroup ran its own internal pay analyses; it just didn’t want to make the results public. And the company won’t have to. When shareholder votes were tallied, 86 percent of them sided with Citigroup. At other banks, Arjuna’s equal pay proposals met the same fate.

Lamb isn’t deterred. She says she’ll file the same proposals again next year, and she’s already in talks with several retail companies to see if they’ll release their numbers without the need for a formal shareholder proposal. In San Francisco, McElhaney has started borrowing tech companies’ preferred phrases to get her point across. A few months ago she gave a presentation to the cloud-based communications service Twilio Inc., which has grown from 90 to more than 800 people in only a few years but suffers from the same lack-of-women problem that the rest of Silicon Valley does. “One of their stated company values is something like, ‘We tackle hard problems,’ ” McElhaney says. “I told them, ‘OK, you pride yourself on solving unsolvable problems, disrupting things that need to be disrupted. Well, here’s a problem. Solve this. Disrupt this.’ ”