Moms Matter: Evidence on the Gender Wage Gap, Parental Influence, and Welfare Reform
Editor’s note: Every Monday, the National Bureau of Economic Research (NBER) releases new “working papers” from leading teams of economic researchers from around the world. The NBER will publish hundreds of these working papers every year. These items usually haven’t been published elsewhere, have not yet been fully peer-reviewed, and reflect brand new research. As such, they represent the most cutting-edge findings in the economic literature. Applied economist Lyman Stone will be writing a regular column on E21 summarizing the most interesting findings from these works. He also posts reviews of these papers on Twitter every Monday under the hashtag #NBERday.
Do moms really matter? The question seems almost too obvious to ask: of course moms matter! But the extent to which parents or parenting matter for child outcomes is hotly debated. Some researchers have suggested that parents really don’t have much influence on how children develop. But new research supports the idea that parenting counts.
Using data on tens of thousands of Israeli students in the 1980s and 1990s, a team of researchers has provided compelling new proof that parents influence kids. Their theory is pretty simple: how much time kids spend with a given parent varies in systematic ways. If a parent dies, time spent with that parent going forward will fall to zero, and time spent with the other parent will rise. If parents get divorced, whoever has custody gets more time with the kids. If one parent works more outside the home, that parent will be around his or her child less. Using this natural variation in how much time a student spends with a given parent, the authors show that spending more time with their more educated parent predicts better exam performance.
Imagine a student has two parents, a father with a PhD and a mother with just a high school degree, and both parents work. Now imagine they get divorced, and the mom receives custody. The student’s eventual test scores are likely to be worse than if the father received custody. Crucially, this is not about genetics. It turns out that the extent to which a parent impacts a student’s test scores depends on how long before the test the critical event, such as a divorce or parent death, occurs. That is, if the student takes the test at age 17, the death of a well-educated parent at age 14 has a smaller effect than if that death had occurred at age 6. But the year a parent dies is essentially random with respect to a child’s genetic makeup. The authors also show that deaths or divorces just after the test have no association with test score differences. In other words, this is truly about parenting.
The authors also find that moms matter more, perhaps because moms interact more with their children. In other words, a mom with a master’s degree has a bigger effect on child performance than a dad with one. Thus, women who earn advanced degrees but who then work as stay-at-home-moms may still be seeing very real benefits from their degrees: their children do better academically.
Given that maternal education may have a big influence on child performance, it’s worthwhile to turn to a new study looking at how having a baby impacts a mom’s earnings. Last year, a landmark study from Denmark showed that, even in a Nordic paradise, there is a huge “baby penalty” to women’s earnings: having a baby reduces their incomes. Since then, similar trends have been reported in the US, the UK, Germany, Austria, and Sweden. This new paper, from the same authors as the Denmark study, combines all these results and shows how the “baby penalty” varies across countries. It turns out that the Nordic countries actually do have the smallest baby penalties: women who have children ultimately make 21% less in Denmark and 27% less in Sweden than women who don’t have kids. The penalty in the US is 31% and in the UK it’s 44%. But in Austria, it’s 51%, and in Germany, it’s 61%!
The authors argue that these outcomes are not due to policy differences: is Germany’s social policy really different enough from Denmark’s to explain that huge gap in the baby penalty? And how do you explain the US having a relatively smaller penalty when its policies are so different from those of the others?
Rather, the driver here seems to be cultural values. The researchers find that the share of people in each country who believe women with young children should stay home is a pretty good predictor of the size of the baby penalty. This is no surprise, since the Danish case at least showed that a huge share of the effect is driven by the large number of women who choose to be stay-at-home-moms after they have children rather than by wage differences between mothers and women who don’t have kids.
Last year’s Denmark paper also shows that the baby penalty explains a large share of the gap in earnings between men and women, too. And most of the baby penalty can be explained by the cultural values of families, and what they want for their kids: do the parents want mom to stay home, or not? It should be no surprise that in countries like Austria, where 44% of women aged 15-50 say moms with young kids should not work, more moms stay home, and thus the baby penalty is bigger. Meanwhile, in Denmark, where under 5 percent of women aged 15-50 say moms of young kids should stay home, more moms return to work, and thus the baby penalty is just 21% of earnings. Much of the baby penalty is basically about personal and family preferences.
But this also complicates policymaking. If the major driver of the baby penalty, and thus much of the gender earnings gap, is about women’s own preferences to stay home with a child, it is not clear to what extent policy can or should try to close that gap. Who is a legislator to tell a mom she’s wrong for wanting to stay home? If that’s what she wants to do, and if it makes sense for the family’s finances, why push her to do otherwise?
The risks of pushing against families’ preferences are made clear in a third paper out last week. Welfare reform in the 1990s did a lot to reduce welfare caseloads and spending, and to boost female labor force participation. But what effect did getting all those women, and moms, into the workforce have on their kids? Well, it turns out that the rollout of welfare reform increased truancy, substance abuse, and criminality among kids in impacted households! That is, when welfare access was restricted and incentives for moms to work were increased, kids fared worse. But one cautionary note should be included: the channel for this effect is unclear; it does not appear to be directly driven by families where the mom entered the workforce. The fact that work incentives alter child outcomes but not through work itself is peculiar enough to merit taking this study with a grain of salt.
But speaking of cases where specific parenting norms impact children, another new study looked at how traditional gender norms impact young girls’ math performance. The researchers’ method was pretty clever: they used formal estimates of “son preference” in Florida families to deduce which families might have more “traditional” gender values. “Son preference” is a way of estimating the odds that having a son versus a daughter alters the odds of having additional children. On the whole, the US does not have much net preference for sons: some groups appear to even have daughter-preferences! But some American families do have measurable amounts of son-preference. By using that measured son preference as a way to get at which families have more traditional gender norms, the paper is able to analyze a much bigger sample size (tens of thousands of students in Florida) than they could using survey-based metrics. Their finding is that a girl who grows up in a more son-biased family has about 3% lower math scores.
The authors make a big deal of this finding: traditional gender norms reduce girls’ math scores! But some perspective is needed. This difference is equivalent to about one-fourth of the difference between growing up in a household with a mom with no high school degree versus a mom with just a high school degree. In other words, large differences in family cultural attitudes have much smaller effects on a child’s school performance than fairly small differences in social class.
Overall, these new papers paint a coherent picture: parents, and especially the parent who stays home with the kids, have a big influence on their children. Having good parents matters for how kids turn out. The sacrifices many parents make, and especially moms if we’re being honest, in terms of workplace success, pay off for their kids. It’s a bit early for Mother’s Day, but it’s never a bad time to thank your mom.
Lyman Stone is an economist who specializes in regional development and demographics. He is also an Advisor at Demographic Intelligence, a Senior Contributor at The Federalist, and a Research Fellow at the Institute for Family Studies. He and his wife Ruth currently live in Hong Kong, where they serve as missionaries in the Lutheran Church.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, e21 delivers a short email that includes e21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the e21 Morning eBrief.
Photo by iStock