AMERICAN MOTHERS typically earn less than childless women, while fathers earn just as much as childless men. The “motherhood penalty” has been documented for years, and it’s the primary reason for the overall gender gap in earnings between men and women. But a just-published study throws open new questions about why it exists — and why it’s been so hard to eradicate.

The earnings discrepancy creates a problem for moms and their families. Children put extra financial pressure on a household, and the motherhood penalty makes it harder for women to help their families meet those needs. Moreover, the hit to earnings leaves women vulnerable should their partner’s income disappear.

Previous research has suggested that mothers might earn less because they switch to smaller or lower-paying firms that offer more flexible schedules. Other scholars have suggested that households are making a rational decision to prioritize the higher-earning partner’s career — and that in opposite-sex partnerships, the higher-earning parent tends to be the father.

But the new study, by economists Douglas Almond and Yi Cheng of Columbia University and Cecilia Machado of Brazil’s FGV EPGE School of Economics and Finance, suggests that’s not what’s happening.

Drawn from two decades of data from 1990 through 2010 for more than 800,000 parents, the study first confirms what so much other research has shown: The birth of a first child doesn’t meaningfully affect men’s earnings but leads to a substantial drop in women’s earnings. On average, mothers lose about $2,000 a quarter — $8,000 a year. That amounts to a relative drop of about 51% compared with their pre-child earnings.

That drop doesn’t meaningfully change at big versus small companies; in other words, women aren’t switching to smaller or more flexible firms. Nor are couples making a calculated choice to maximize the household’s earning potential because, the researchers conclude, the motherhood penalty persists in households where the woman earns more. In fact, in households with breadwinning moms, the penalty is even larger. These women experience a drop of 60% in their pre-child earnings relative to their lower-earning male partner.

Part of the gap is because some women end up leaving the workforce, either temporarily or permanently. According to census data, in households with children younger than 6, 35% of mothers are out of the labor force — compared with just 4% of fathers.

Yet mothers who keep working also find their earnings penalized. “This indicates that the effect is not only due to women leaving the workforce, but also due to lower earnings among the employed,” Almond says.

For all mothers, the penalty is durable — it didn’t meaningfully shrink in the six years after the birth of a first child and in fact even grew a little bit over that time. The penalty is also depressingly hard to escape. It persists among the college-educated regardless of whether the boss is a woman and regardless of whether the company has a predominantly female employee base.

The die is cast when the baby is born. “The benefits of childbearing accrue to both parents,” the economists write, but “the burden of childbearing comes at the expense of women’s labor market outcomes.”

Of course, not all women who leave the paid workforce to raise children are miserable about it. But they are taking a financial gamble. Even in affluent households, relying on a single income is precarious: A woman is putting her financial future at risk if her spouse ends up out of work, seeks a divorce, suffers a serious illness or dies early. And having your own income can be an important way to maintain a degree of influence and independence.

The motherhood penalty isn’t as pervasive in Scandinavian countries. Previous research on women in Sweden, for example, finds that in couples where women earn the same or more than their male partners, these breadwinning mamas face no motherhood penalty.

What is Sweden getting right that the US could copy? My first thought was “paid parental leave.” The US is notoriously the world’s only wealthy country not to guarantee paid maternity leave, and Sweden has generous parental leave policies for both mothers and fathers. But the researchers say that paid leave is unlikely to be enough. “Germany and Austria had the largest motherhood penalties but also more generous leave policy than the US,” notes Cheng, citing previous research led by Princeton’s Henrik Kleven.

What about making day care more affordable? While that could help some women — particularly lower-earning women — the researchers don’t see it as a silver bullet. After all, in the families where women bring home the bacon, even expensive day care is a sensible investment. But their data suggest that those families aren’t making that kind of calculation.

Instead, solutions may lie in cultural change — getting both mothers and fathers to see parenting as a joint venture. “Ultimately, the traditional thinking is that child care by default is the responsibility of the mother,” Almond explained. “When you visit a playground in Sweden, there are as many fathers there as mothers.” This isn’t what he sees strolling around New York City. “The difference is striking.”

Ending the assumption that mom is in charge by default — and that dad, when he’s parenting, is “babysitting” or “helping” — would also combat one of the most pernicious reasons for the motherhood penalty: bias against moms in the workplace. Study after study has shown that employers see mothers, unjustly, as less committed and less competent. That translates into fewer opportunities and lower pay. “Without a doubt, mothers who stay in the labor force face biases in their perceived commitment,” Cheng says.

And that’s a way of thinking that compounds women’s financial vulnerabilities. Women of nearly all races and ages have higher poverty rates than men; make up two-thirds of minimum-wage workers; hold two-thirds of student-loan debt; and have lower retirement savings than men. All of this, of course, is bad for children, who typically follow their mothers — but not their fathers — into penury.

I agree that a cultural change is needed. But I also believe in the power of policy to make room for those changes. The uptake rates on unpaid leave — the only type American companies with more than 50 employees are required to offer — aren’t inspiring. But companies that have lengthened their paid maternity leaves say they see more women return. And offering an equal amount of paid leave to new dads would put men and women on an equal footing at work — and open a door for fathers to be more involved during a baby’s early months, laying the groundwork for more equal parenting later. Yes, a cultural change would be needed to get those dads to actually use their paternity leave — but they can’t use a benefit they don’t have.

Similarly, expanding day care access would send a cultural signal that mothers were a valuable part of the workforce and children were an important part of society. That could help change underlying assumptions about which employees are the most “committed.”

A woman who gives birth puts her body at risk to create new life. With smarter policies and some meaningful cultural change, it doesn’t have to be a financial risk, too.

BLOOMBERG OPINION