The gender pay gap is a labyrinth of unconscious prejudice and blatant discrimination, a byproduct of unbalanced professional opportunities, educational prospects and deeply-entrenched societal norms. With so many forces at work, getting to the root of glass ceiling issues within a single firm or industry is difficult, if not impossible.
Janice Fanning Madden, a Wharton real estate professor and a professor of regional science and sociology at the University of Pennsylvania, had the rare opportunity to examine the mechanics of the pay gap among male and female stockbrokers while acting as an expert witness in class action lawsuits filed against two large brokerage firms in the early 2000s. Women from the firms claimed that the pay structure at the companies unfairly benefitted men. Leadership at the brokerage houses contended that both men and women were paid using the same commission-based system and that women were paid less on average because they were worse sales people than men.
Madden found that women were assigned inferior accounts, which led to them earning lower returns and smaller commissions. As Madden notes in "Performance-Support Bias and the Gender Pay Gap among Stockbrokers," this perpetuated a vicious cycle because the firms doled out amenities that often aid in better returns -- such as bigger or more attractive offices, support staff and mentors -- based on employees' sales records. Thus, women not being given the opportunity to handle a lucrative account today also hurt their chances of being given advantages that could have a significant impact on their future efforts.
"I wasn't able to see what kind of offices [the women] had, what kind of secretaries or what kind of titles," Madden says. "But the decisions of assigning those things to women were made by the same people who were stiffing them on the accounts. They were probably dealing with less support in these other areas, even when they had the same type of account [as men] to work with."
Using a "natural experiment" based on sales generated by accounts that were transferred by management from one broker to another, Madden was able to analyze performance when the playing field for accounts was equal -- when female employees were working with clients who had same potential to produce high commissions as those handled by male stockbrokers. Madden shows that the women produced sales that were at least equivalent to those produced by men. "I was quite surprised that women were such strong performers. I had sort of thought that the brokerages were right and that women were less likely to churn their accounts than men. That was not beyond belief to me," she says, pointing to evidence showing that women overall tend to invest their own money more conservatively than men. "But they're given the same incentives as men to sell on their accounts, so they behave in the same way."
A Gut Reaction
Over the past 15 years, women have accounted for about one-third of all full-time stockbrokers. During that time, Madden notes, their earnings have grown from 54% to about two-thirds of the salaries earned by men -- but the job still has the largest gender pay gap among sales positions, although it is also the highest-paid of those positions overall. The two organizations Madden studied were large, national full-service brokerage houses that sold financial products primarily to individual investors. Women made up a relatively small percentage of the stockbroker ranks at both firms -- 11.2% at one and 13.8% at the other.
"But the gender gap in wages at these two big firms is much smaller than the gender gap for stockbrokers, [according to] the Census," Madden says. "It's not because these two firms discriminated less but because women are more likely to work in small firms that are less successful. These firms also had fewer women proportionally than the occupation as a whole so some of the difference comes from the difficulty of getting a foot in the door of a major stock brokerage firm."
At both businesses, stockbrokers' salaries were entirely determined by commission by the end of their second year after passing the Series 7 exam required to become a registered broker. At the time the lawsuits were filed, there were "significant" gaps in compensation among male and female stockbrokers, Madden writes. "There are really only two reasons for a gender gap in compensation derived from commissions on sales.... Women are less effective at sales, on average, than men; and/or performance support bias -- the sales opportunities, including account assignments and the various supports used to produce sales assigned to women, are inferior to those assigned to men."
Madden examined both possible factors using data on more than a billion individual transactions on customer accounts within the firms between 1994 and 1996. She notes that there are several possible reasons why women may achieve worse sales figures than men: It could be that they simply have less innate ability, but it could also be due to societal factors like clients being less willing to work with, or make purchases when dealing with, women stockbrokers.
The "natural experiment" was created by observing sales yields from stockbrokers of both genders on accounts with similar histories that were assigned to incumbent stockbrokers after the original overseer of the account left the firm. At one firm, women were significantly less likely than men to receive these transferred assets. But when they did get these accounts, women had stronger sales achievements than men. Men and women were equally likely to be given transferred accounts at the other firm, and in that case their performance on accounts with similar histories and yields was equally strong.
These results provide evidence that the performance differences do not stem from innate ability, Madden says. She also did not find strong support for the hypothesis that customer bias played a significant role in the gender pay gap. Instead, Madden discovered that women at both firms were less likely to receive the types of accounts that produced higher commissions. This in turn hurt their salaries and their ability to compete for benefits that were awarded based on performance.
"Both brokerage firms alleged that women receive inferior account transfers because they generated lower commissions in the prior year, and the prior year's commissions were used to allocate transferred accounts," she writes. But employees weren't operating on a level playing field in the way clients were parceled out. And the consequences of being relegated to "inferior" accounts only balloon over time, Madden adds: "The effects of small annual differences in the distribution of accounts or of other forms of performance support accumulate over a career as early career differences allow brokers to qualify for additional benefits, such as titles [and] office space, based on account size or production."
The lawsuits were settled before going to trial and, as part of those agreements, the two firms agreed to change their policies so accounts were distributed to brokers using a standardized system. Madden heard from lawyers working with the plaintiffs that people from one of the firms commented on her findings "and said [the company] had clearly been making mistakes in assigning accounts.... They had not realized how well women were doing."
According to Madden, the study is the first to show how gender bias affects employees working under a seemingly objective performance-based pay system like commissions, rather than one where salaries are determined by subjective performance evaluations. She notes that the research highlights how important it is for firms to examine the ways that subjectivity can creep into, and damage the effectiveness of, systems designed to encourage equal treatment in the workplace.
"In lots of occupations, women or minorities can get the short end of the stick because people are operating without thinking," Madden says. "They would never purposely ... discriminate against somebody, but they're reacting with their gut and unfortunately, when we react with our gut, we sometimes bring in prejudices we're not aware that we have."