If your boss is having his first child, you’d better hope it’s a girl.
That’s the takeaway from a new study published in Administrative Science Quarterly, which examines how employees’ wages change immediately after a male chief executive officer has a child. Economics and business professors from Denmark’s Aalborg University, Columbia Business School, and the University of Maryland’s Smith School of Business studied the salaries of 1.2 million people across 10,600 companies in Denmark and found that when a male CEO has a child, his employees’ wages decrease. It’s not a large decline—just 0.2 percent, adjusted for inflation—but when spread across an entire company, that small amount of money (about $100 per employee) adds up.
At the same time, the CEO pads his own pocket with a wage increase of 4.9 percent. “He has a kid, he thinks immediately, ‘I want more money for my family,’” explains Cristian Dezsö, assistant professor at University of Maryland’s Smith School of Business and co-author of the study. “But it comes at the expense of the employees.”
The study was inspired by a 1991 paper by sociologist Rebecca Warner that linked male politicians’ stances on so-called women’s issues with whether or not they had a daughter. (Not surprisingly, she found that those with daughters were more likely to hold feminist views). “Given my interest in business, when I read that paper I immediately thought, Well, does a daughter influence a CEO’s stance on gender relations in the workplace?” Dezsö says. The answer appears to be yes.
Dezsö and his colleagues focused on Danish companies because of the ease with which they could acquire salary data. “In Denmark, people are issued national ID numbers much like the Social Security numbers in the U.S, and they link that ID to pretty much everything,” he explains. “If I know your number, I can see the company you work for, how much money you make, who is your spouse, your kids, and so on and so forth.”
Dezsö discovered that employees’ wages fall farther if the CEO has a son than if he has a daughter—and they do so at different rates for women than for men. If a chief executive’s firstborn child is a son, female employees’ wages go down by .2 percent and men’s drop by .5 percent. Dezsö is quick to point out that this doesn’t mean women suddenly make more than men—just that the gender pay gap, which exists even in Denmark, narrows slightly. The reason for this might be the executive’s newfound respect for motherhood. “Whenever a CEO has his first kid, he is probably married, and so he sees how hard it is to be a mother … by extension he might sympathize more with women. So the salaries of female employees don’t go down as much as men’s,” he says. When the CEO has a daughter, he’s even more sympathetic—and women’s wages actually go up.
This phenomenon occurs only when executives have their first child. With two or more children, employee’s wages decrease no matter who they are. “Think of it as negative and positive forces working against each other—the CEO wants more money for himself, but he also feels more generous to his employees,” explains Dezsö. “With the first child, positive forces win. With the second or third child, the negative forces win and he just takes more money for himself.”
We were curious to know if these findings hold true for American CEOs as well. (According to Dezsö, the study couldn’t be replicated in the U.S. because privacy laws make it too difficult to procure the relevant information.) While he doesn’t know for sure, he has a hunch. Denmark ranks seventh in the world for gender equality, according to the 2012 Global Gender Gap report, “which means that most people there already hold liberal views of women’s rights,” he says. The U.S., meanwhile, is ranked 22—just above Mozambique. In a more biased society, the economic impact might be even more pronounced.
“If I were to get very cocky, I’d say that what we found in Denmark is a very conservative change in CEOs’ outlooks and maybe we’d find bigger effects in the U.S.” But he doesn’t know for sure. “Until we can collect the right kind of data in the U.S.,” he says, “it cannot be more than speculation.”