Women managers are significantly more likely to be promoted to CEO at companies organized into separate business units—each with its own revenues, costs and bottom line—than at companies that operate as a single, stand-alone entity, a new study led by the University of Massachusetts Amherst finds.
Women still hold fewer than 10% of CEO positions in S&P 500 companies, despite gains over the past two decades.
"As recently as 2015, there were more men named John running major U.S. companies than there were women," says Ulya Tsolmon , associate professor of strategy in the Isenberg School of Management at UMass Amherst and lead author of the study published in the Strategic Management Journal . The gender gap is "not just any disparity. It's a huge disparity."
Tsolmon and co-author Tingyu Du, of California Polytechnic State University, reveal the explanation may have less to do with company culture and more to do with accounting.
When a manager runs a division with its own profit-and-loss statement (P&L), her results are measurable, comparable to those of the executives running other divisions—and, importantly, difficult to credit to someone else. This narrows the space for subjective judgment, where research has repeatedly found bias creeping in.
"It's clear who has led the unit," Tsolmon explains. A division with its own P&L makes it "much easier to say, 'Yes, this person ran this business, and here are the results.'"
The study draws on 25 years of personnel records from the LexisNexis Directory of Corporate Affiliations, covering 616,754 managers at 49,135 U.S. companies from 1993 to 2017. Promotion to CEO was rare in any given year—occurring in about 0.56% of roughly 2.4 million manager-year observations.
Overall, the study estimates that women in multi-unit firms had a 37% higher probability of internal promotion to CEO than women in stand-alone firms. Men, by contrast, fared slightly worse at divisionalized firms.
Tsolmon and Du are careful about what the findings do not show, noting the results describe associations rather than proof that restructuring a company will produce women CEOs. They also caution that divisionalization doesn't necessarily address inequality elsewhere, especially when it comes to who gets handed a division to run in the first place.
"The prescription is definitely not that everybody should just go out there and become divisionalized," Tsolmon advises.
Companies do not choose their structures at random, and each carries trade-offs. Multi-unit firms delegate more, which means mentoring managers and linking incentives to results. More centralized firms rely on internal control through hierarchy, relationships and impression management, which could be more difficult for women if they have fewer chances to break into the networks that allocate resources and promotions, Tsolmon suggests.
"We tend to think about organizational structure as a question of efficiency, coordination and control," she says. "But structure also determines what information organizations produce about people—and that can affect who gets ahead."
She would like to see future research examine whether the gender gap emerges earlier on the career ladder, and whether the visibility of performance affects men and women differently as they climb.
"Until we really understand the mechanisms behind the disparity, it's hard to prescribe solutions," Tsolmon adds.
