Key aspects of corporate sponsorship programs, while designed to advance women’s careers, may end up widening the gender gap rather than narrowing it, according to new experimental research.
“We’re not trying to say that sponsorship programs don’t work or that they’ll always backfire,” says Katie Coffman, an assistant professor at Harvard Business School and co-author of the study. “What we are saying is that certain aspects of sponsorship don’t work exactly how we’d want them to work, at least in the lab.”
The findings are detailed in Laboratory Evidence on the Effects of Sponsorship on the Competitive Preferences of Men and Women, published in the journal Management Science. The study was co-authored by Coffman and Nancy R. Baldiga, an economics and accounting professor at the College of the Holy Cross, who is Coffman’s mother.
It’s well documented that women are underrepresented in the executive suite. For instance, women have comprised nearly 50% of entry-level public accountants for more than two decades, according to the American Institute of Certified Public Accountants. Yet women make up less than 20% of partners in accounting firms.
To address the stubborn gender gap, a growing number of companies have adopted formal sponsorship programs, in which a seasoned manager officially takes on a junior protégé. Sponsorship is essentially mentorship on steroids. While a mentor may provide valuable career advice and an empathetic ear, a sponsor will actively advocate for a protégé’s career—formally and publicly recommending her for high-profile projects or job promotions. If the protégé handles new challenges successfully, it makes both sponsor and protégé look good.
“Sponsorship builds on traditional mentorship by adding a more transactional nature to the relationship,” explains Coffman, who works in the Negotiations, Organizations, and Markets Unit at HBS.
“It’s not just the altruistic view; it’s that if my protégé does well, I do well, too. Sponsorship is a pretty new concept, but it’s becoming more popular, especially in professional service firms.”
Sponsorship programs are meant to boost confidence among protégés, increasing the likelihood that they will compete more effectively against their peers. Research has shown that men tend to favor competitive settings more than women do, which may contribute to the gender gap in the high-stakes senior ranks of a company. Ideally, sponsorships would result in more women entering the competitive pool, which would then result in more women rising to the top.
But the study suggests that key aspects of sponsorship actually benefit men more than women.
Relative interests and a collaborative experiment
Baldiga worked in the professional services industry—as an audit manager at Price Waterhouse—for several years before Coffman was born. (“Gender and work/life issues were close to her heart when she got pregnant in the 1980s,” Coffman says.) Since joining academia in the 1980s, Baldiga has studied issues related to advancement and retention in professional services firms. Coffman, meanwhile, is a behavioral economist who uses experimental methods to study decision making, with a focus on gender.
A few years ago, the two of them decided to team up on a laboratory study on the value of sponsorship. “We took her applied interest—how companies are actually using sponsorship programs to address [the gender gap]—and tried to bring to bear some experimental economics tools in order to understand how well these initiatives might work,” Coffman says. “And it was a great opportunity to work with my mom … It’s always great to work with someone I admire.”
Specifically, Baldiga and Coffman wanted to gauge the efficacy of two isolated ideas related to sponsorship: one, the idea that being personally chosen by a sponsor as a protégé could serve as an important vote of confidence; and two, the idea that sponsors’ compensation being linked to protégé outcomes might serve as further motivation for the protégé.
They set up several rounds of experiments in which 176 men and 178 women—mostly Harvard undergrads—solved basic math problems for cash rewards. “While experimenting directly on executives in the field would yield interesting insights, starting in the laboratory with a student population provides a cost-effective way to glean important information that could inform a more resource-intensive field experiment,” Coffman says. “Plus, students are a nice population for an experiment like this because many of them will go on to have important leadership roles and make important leadership decisions.”
In one round of the experiment, participants could opt for a guaranteed payment of 50 cents per problem solved, or they could compete for higher stakes. Those who chose to compete received a significantly higher payout (ranging from $1 to $3 per problem) if they scored in the top 25% of the problem solvers, but received nothing if they did not. As the researchers predicted, more men than women chose the competition over the guaranteed payment.
The subsequent round introduced the element of sponsorship. In each session, three participants were randomly chosen as “sponsors” who would receive 25 cents each time a “protégé” correctly solved a problem—but only if their protégé scored in the top quartile and decided to choose the high-stakes competition over the lesser guaranteed payment.
Significantly, the protégés were told that they had sponsors before choosing whether to take the guaranteed 50 cents per solved problem or whether to compete for more money. The goal of the researchers: To find out whether participants were more likely to compete if someone were sponsoring them.
The research revealed that sponsorship definitely increased the likelihood that participants would choose a high-stakes competition over a non-competitive guaranteed payment—but this proved true only among male participants. And in terms of the choice to compete, sponsorship had the most dramatic effect among the men who performed the worst on the math problem exercise.
In other words, sponsorship served to boost the confidence of low-ability men.
Among female participants, the introduction of a sponsor had no significant effect on the choice to compete; this was true even when the protégés were told they had been specially chosen by a sponsor.
“So, while you might be hoping sponsorship might close the gender gap by encouraging high-ability women who are reluctant to compete, providing that extra nudge to compete, it in fact has the opposite effect,” Coffman says. “If anything, we see a larger gender gap in willingness to compete [when sponsorship is introduced].”
Designing better sponsorship programs
The researchers emphasize that sponsorship is not necessarily an ineffective way to help women in the workforce, but rather that certain aspects may not work exactly as intended. It behooves organizations to be more thoughtful about how sponsorships are designed.
For instance, a protégé’s success may be less about the mere vote of confidence than about the advocacy skills or the wisdom of the sponsor. Future research may reveal which factors matter most.
“I think it’s great that we’re being more creative about solutions to female advancement, and there might be a lot of good things going on in terms of sponsorship,” Coffman says. “Now we just need to understand better what does work and why.”
This article originally appeared on Harvard Business School’s Working Knowledge.