When Olivia "Mandy" O'Neill entered the PhD Program at Stanford in 2000, the question she most wanted to answer was one that had been puzzling researchers and employers alike: Why were promising professional women failing to realize their potential in the workforce? Previous research into this question, which tried to disentangle people's biological sex from their gender identity, left everyone scratching their heads. For example, some studies looking at who succeeds had found that stereotypically masculine people of both sexes earn more than feminine people, while other research showed almost the opposite — that masculine women suffered a backlash for violating gender norms.
Even today, when her own investigations have given her much clearer answers, O'Neill sounds a note of frustration as she describes the confusion in the scholarly literature at the time. "The research wasn't coherent and really didn't explain the problem, which is: Why were women opting out, particularly the ones who looked like they should have the highest potential?"
O'Neill's dissertation advisor, Stanford's Charles O'Reilly, not only shared her interest; he also had a rich data set that held important clues. Back in 1987, when he was a professor at the University of California, Berkeley's Haas School of Business, he had surveyed MBA students there about their career plans and personalities. He then did follow-up surveys at four-year intervals to see, among other things, how much the women and men were earning. The Haas School provided an ideal sample, O'Neill explains, because its graduates in those days were embarking on a wide range of career paths, so their experiences could say something true about people's career paths in general.
In starting their work together, O'Neill and O'Reilly thought — just as past researchers had — that gender, apart from biological sex, would bear on people's income. But instead of looking at participants' answers to questions about their gender asked in stereotypical ways ("Do you like playing with children?"), the two tried to get at something they thought might be a better sign of people's career ambition: Would they rather work in so-called feminine organizations, meaning ones that are supportive and nurturing, or in masculine ones, which are competitive and aggressive?
In 1991, just four years after graduation, women were earning at least as much as men — in fact, women who had expressed an interest in working in a masculine organization were out-earning men. And members of both sexes earned more if they had expressed a preference for a masculine organization. Put another way, gender mattered more than sex.
But eight years after graduation, in the 1995 survey — by which time the MBAs' income ranged from $15,000 to $2 million per year — women had fallen behind men. To see why, O'Neill and O'Reilly dug deeper into the data, which revealed an interesting answer: The number of hours put in at their jobs made all the difference. Quite simply, eight years into their careers, women were working fewer hours than men on average, even when compared with men who had the same number of children.
That may sound like meritocracy at work: Rather than suggesting that qualified women were being passed up for promotions and raises, the relationship between hours worked and income earned seems both straightforward and equitable. But the situation isn't as simple as that. Many organizations set up workers' rise up the career ladder as a "tournament," the researchers note, borrowing an idea first described by Stanford GSB economist Edward Lazear. Rather than seeing their rewards accrue in direct proportion to their productivity, employees in a tournament vie for top rank at each stage, from department head to chief executive, with the winners at each tier earning the lion's share of raises and promotions to the next level.
A tournament system is sensible from the point of view of employers, who can't always easily measure professionals' output. But since workers' success depends on outperforming others in the organization, tournaments foster intense competition. More surprisingly, they also create changing incentives over time. Workers' raw potential matters a lot in the beginning, when they are getting job offers straight out of school, but as more and more people get winnowed out of the running over the years, the remaining contenders are increasingly similar in their ability.
That leaves only one way to get ahead of the pack. "As the tournament unfolds, what starts to matter more is effort, or hours put in," explains O'Neill, who received her PhD from Stanford GSB in 2005 and is now an assistant professor of management at the George Mason University School of Management in Fairfax, Va. "Working hard and being willing to make sacrifices make a bigger difference at later stages of the tournament."
It's at these later stages that women start to lag men, who eight years after graduation were not only working longer hours but also expressing a greater willingness to relocate for work.
Why do women work fewer hours than men do? Demands back home may play a role. O'Neill's study shows that women were doing twice as much housework and child care as men. And, as she puts it, "Time is a zero-sum game, so hours they were spending doing that is hours they weren't spending gunning for the next promotion."
Biology surely plays some role as well, since women's fertile years — unlike men's — are limited and, in many fields, coincide with peak promotion years.
Another intriguing possibility is that some women use family responsibilities as a convenient way to avoid facing possible career failure. Telling yourself you had no choice but to cut back when you had your baby, after all, is easier on the ego than going all in and losing the tournament anyway. The tendency to self-handicap, as psychologists call this form of self-deception, certainly isn't limited to women, and men might want an excuse to opt out, too — but they lack an alternative path as well-trod and socially acceptable as taking care of the home front. Still another possibility is that women sacrifice time at work for time at home without much thought, O'Neill says, perhaps in response to subtle messages from parents, spouses, and bosses.
In short, it's complicated: Sex, social expectations, and individual psychology intermingle in ways that can be hard to foresee at the start of one's career. At least part of the answer to the riddle of why highly capable women don't always realize their potential in the workforce, O'Neill concludes, is that people are more than their potential. What they value affects what they pursue, and values can change over time.
Executives who are serious about retaining talented women, the research suggests, need to understand the ways that mainstream systems for promotion — and organizations in general — aren't as gender-neutral as they might appear. Employees, for their part, would do well to think of their careers as tournaments, and taking the long view from the very start. Planning and patience can pay off in the form of work-life balance: A woman who becomes a boss and earns a high salary, O'Neill points out, is in a better position to set her own schedule, find good child care, and keep the career she wanted.
Members of both sexes can benefit from looking at their life as a whole. Though this particular study focused on income, O'Neill says we need to expand our definition of success to include overall happiness and good health. "If you're dead," she says, "you can't be the CEO."
Mandy O'Neill is an assistant professor of management at the George Mason University School of Management. Charles O'Reilly is the Frank E. Buck Professor of Management at Stanford GSB. Edward Lazear is the Jack Steele Parker Professor of Human Resources Management and Economics at Stanford GSB.