Long-term changes in domestic structure are a reality with which to cope
Broken homes didn’t cause the shocking increase in shootings and homicides this country has seen since last year. Crime rates changed abruptly, at a time wracked by the murder of George Floyd, the attendant protests and riots, and a deadly pandemic. Family structure has changed gradually for decades.
But those slow-moving changes create a new reality that all our crime-control efforts must struggle against. Children are much less likely to grow up in married two-parent households than they were decades ago. And the decline of stable families is an ever-present risk factor for crime.
The basic statistical trends are striking. There was a large shift away from married parenthood that unfolded over several decades, and then mostly stabilized — with minor improvements in some measures lately.
In the early 1960s, fewer than 10 percent of American kids were born to unmarried mothers. That number rose to about 40 percent in the mid 2000s and remains there in the most recent Centers for Disease Control data. The share of all minors living in non-two-parent families followed a similar trajectory. It started around 12 percent in 1960, rose to nearly a third in the 1990s and 2000s, and has fallen a little since. Notably, in these data from the Census Bureau, many kids classified as “living with both parents” are living in stepfamilies or have parents who are cohabiting instead of married.
Elevated rates of single parenthood are especially pronounced for the demographic groups that struggle the most. Education is one stark dividing line: Per the CDC, only about 10 percent of births to moms with a bachelor’s degree or higher, but more than half of births to moms with less education, are to the unmarried. Race is another: Among African Americans, just 30 percent of births are to married mothers. White minors are about twice as likely as black minors to live with two parents.
And just as it’s undeniable that stable two-parent families have declined, it’s abundantly clear that kids who grow up in unstable homes are more likely to have trouble with the law. For example, one recurring government survey asks prison inmates what family situation they lived in “most of the time” while growing up. About 60 percent say they didn’t spend it with two parents, including stepparents. Recall that, at any given time over the past few decades, fewer than a third of American kids have lived in non-two-parent families.
Of course, it’s notoriously difficult in social science to move from correlation to causation. Family structure correlates with all sorts of other things related to crime, including — as we just discussed — education and race. Is it possible that family instability doesn’t really matter all that much in itself?
Possible, but not too likely. There’s always room for doubt in social science, but there are strong reasons to believe that family instability is bad for kids when it comes to many things, including crime.
First there’s common sense to contend with. Not only do kids experience an immense amount of stress if their parents split up or switch partners, but having a second parent around on a permanent basis means more supervision, more income, and more adults available to serve as role models and help when things go wrong. If a person who’s biologically connected to a child and legally responsible for the kid’s well-being doesn’t make a difference in the child’s life, it’s awfully hard to imagine what could.
Additionally, when researchers statistically control for confounding variables such as race and education, there’s still an incredibly strong link between family structure in childhood and later crime. In a 2004 study, Cynthia C. Harper and Sara S. McLanahan looked at a data set that had tracked a group of American teens starting in 1979. Harper and McLanahan then accounted for variables including the mothers’ education, both parents’ income, the race of the children, and more. They found that while including these variables reduced the correlation between family structure and incarceration, kids from single-mother and stepparent families still had more than double the odds of being incarcerated between ages 15 and 30.
Bradford Wilcox and Wendy Wang of the Institute for Family Studies (where I am a research fellow) did a similar exercise with a newer version of the survey Harper and McLanahan had used — it tracked a group of kids starting in 1997. They controlled for maternal education, the children’s own scores on a cognitive test, and a few other variables, and still found about double the odds of incarceration for kids who grew up in non-intact families.
Some study designs go a bit farther, seeking to find “natural experiments” or other clever ways to tease apart correlation and causation. For instance, siblings are generally similar in their genetics and the environment they grow up in — but in a divorce, the older sibling lived longer in a two-parent family, so researchers can check to see whether older siblings in divorced families tend to have better outcomes. In an important literature review published in 2013, McLanahan, Laura Tach, and Daniel Schneider rounded up studies with stronger designs like this that addressed the relationship between father absence and various outcomes, including delinquency and “externalizing” behaviors such as aggression. Overall, they concluded that “studies using more rigorous designs continue to find negative effects of father absence on offspring well-being” — though weaker effects than less rigorous studies do. This was true of 19 of the 27 studies they reviewed on externalizing behaviors and delinquency.
Now, even these stronger studies are difficult to draw firm conclusions from, and they have their limitations and legitimate criticisms. For example, if a happy marriage turns bitter and acrimonious, perhaps that will have an outsized effect on the youngest child whether the parents divorce or not. (If so, studies with the design mentioned above still measure the effect of a volatile family environment, but they don’t necessarily suggest that divorce in itself is the problem.) Moreover, it’s not the case that every child is best off with both of his or her parents around; some research suggests that child outcomes improve when a parent is incarcerated, for instance, and no one is suggesting that abusive parents are good for kids.
National crime trends also caution against overzealous claims here. As crime rose from the 1960s through the early 1990s, some pointed to the parallel trend of rising single motherhood as the prime driver of the problem. At the end of that period, some worried that an even more troubled generation of “superpredators” was coming of age. But crime dropped in the 1990s, despite no improvement in family structure. Father absence is a risk factor for crime, but it’s not the sole determinant of crime rates nationwide.
Still, why did this happen? How did we end up with so many kids being born out of wedlock and growing up apart from at least one of their biological parents?
A number of different changes all pushed in the same direction. Women advanced in the workplace, making them better able to support children on their own and better able to leave marriages they no longer found worthwhile. Divorce laws liberalized, and America got richer, making it easier for couples to split and then keep two households running simultaneously. The welfare state expanded, as did child-support enforcement. Norms around sex and family shifted.
We’re unlikely to roll much of that back, so to a large extent, the current situation is just a new normal we’ll have to deal with. Yet there are a handful of ways we could try to nudge things back in a healthier direction — bearing in mind that, however slightly, some family metrics have already started to improve recently.
We could eliminate “marriage penalties” in the welfare state, which cause single people on government programs to lose benefits if they wed. We could even give extra benefits to married parents, as Scott Winship of the American Enterprise Institute has proposed, though I doubt the political prospects of that idea. We could try to reduce unwed childbearing by encouraging the use of “long-acting, reversible” contraception, as I suggested in a 2014 piece in this magazine (“On a LARC,” December 31). We could also keep experimenting with programs that encourage marriage, though these generally show weak results. These kinds of measures could at least have some effect on the margin and communicate society’s endorsement of the married, two-parent model.
A half century of family breakdown is not the only reason America struggles to contain crime. But it’s one reason, and we should take it seriously.
______________________
Robert VerBruggen is a fellow at the Manhattan Institute.
This piece originally appeared in National Review