Book Review: The Two Parent Privilege
Why has marriage declined in working class America? What are the consequences? And what should we do about it?
The Two Parent Privilege is a book about two things. It is about the decline of marriage in working class America and it’s about how Melissa Kearney is not socially conservative – she swears.
The first thing is an interesting topic for a book, and Kearney explores the key questions in a data rich and clear way. But she also dedicates loads of pages to how brave she is to be speaking up about this issue and how really she isn’t socially conservative.
The most annoying consequence is that the book has lots of tedious paragraphs like this.
I thought about a conversation I had recently had with a different economist in a different setting who reacted negatively when I mentioned the importance of family structure to children’s outcomes. He bristled, suggesting to me that I sounded “socially conservative”, in a way that implied “not academically serious.” I countered, “You are always talking about the things you are doing for your kids and how much time their activities take up in your life. Why would you be offended by the suggestion that maybe other kids would also benefit from having the involvement of two parents, and in particular a father, in their lives?”
The actual substantial problem is that she does not engage properly with right wing explanations for the data she presents – more and more children are being raised by single mothers and this seems to have a strong impact on their educational outcomes, future earnings, odds of going to prison and odds of becoming single parents themselves. She doesn’t ask obvious questions about whether there’s a genetic explanation for these kids’ bad outcomes.
In fact, she frequently presents data that implies there’s a genetic component without even acknowledging that a person could possibly think this. This is what she has to say when talking about gaps in kindergarten test scores between rich and poor kids.
“Although these achievement gaps cannot be definitively linked to differences in parental investments, the fact that they are observed in kindergarten suggests that they are driven by factors that predate (in the children’s lives) anything that could be attributed to the school environment. That shines a spotlight on home environment, early-childhood experience, and even in-utero experiences.”
(Or maybe, just maybe, there are some important pre-utero factors that can impact a person’s life.)
I think it’s plausible that deadbeat dads have bad genes and pass those genes onto their kids and that is a partial explanation for why those kids have worse outcomes. I’d like to know if that’s true and to what extent it’s true, but I’ll have to read a different book to get those answers.
That being said, the book is data-rich and interesting, and here’s what I learned.
The rise of single-parent households
There has been a substantial decrease in the number of children living with married parents. In 1980, 77% of American kids had married parents, now only 63% do. Most of the decline has happened in families without college-education.1 Educated mothers are getting married almost as much as they used to. (When educated women forgo marriage they also, mostly, forgo children too)
The data suggests that this is not because of happy progressive reasons. These children are not living with parents who are living together in stable unions, but have eschewed marriage because they are too modern for that sort of thing. Nor are they mothered by girl-bosses who picked out a top quality father from a sperm-bank and decided to do it by herself. Most of these children are raised by single mothers with little or no support from the dad and they are born to poor mothers.
Kearney’s graph above is statistically naughty. The proportion of Americans not completing high school is not the same as it was forty years ago. Nor is the proportion of Americans who are going to college. She talks about these groups as though they are a static monolith through this period but that is not the case.
Statistics make it difficult to tell outright lies but they make it easy to wildly exaggerate. She’s presented women with no high school, women with no college, and women with college as though they have been the same group since 1980. Actually the proportions have changed quite dramatically. The 70th percentile woman, by academic attainment, probably didn’t have a college degree and now she does. This means if marriage rates are determined purely by how good at school you are, marriage rates could have been stable overall but declining for each of the education groups that Kearney has created.
This effect overall means that comparing women without high school in 1980 and now is a fool’s errand and it renders a lot of Kearney’s analysis irrelevant.2
While I dislike the way she presents the information, Kearney is right about her core thesis: marriage is declining and that the rise of single parenthood has been concentrated among less educated and poorer mothers.
For unmarried parents, the fathers usually stay around for a few years. At the time these children are born, 74% of the mothers and 90% of the fathers say they expect to get married. But by the time the child is five, only a third of those couples are still together. In another third the parents are separated but the father is somewhat involved. In the last third, the father is completely absent.
Unfortunately, these fathers are not paying for the upkeep of their children. Only about half of single mothers have child-support agreements with the fathers of their children. Of this half, only 44% of them received the full payment owed.
This is a problem. Children raised in single-parent households have much worse outcomes. Children raised by single parents, when controlling for other factors, are less likely to complete college. And, this effect has been getting stronger over time. They also are more likely to have behavioural problems, go to prison, and be single parents themselves when they are older.
To some extent this is a selection effect. If marriage has become the preserve of the middle class, then we should expect the children of married parents to have better, more middle class outcomes regardless – your personality is heritable after all.
But even beyond that, single-parenthood is worse. If you have two parents in a house then there are two adults who can dedicate their incomes and time to their children. That kid then has more resources. Unsurprisingly, the children of married couples spend more time with both their parents. And the children of high-income parents have more money spent on them and are therefore more likely to do sport, athletics, music, dance or art.
Some parents are Bad Actually
Kearney then uses the facts above to promote marriage but I’m dissatisfied with her exploration of the counterfactual. The average dad might have a really positive impact on their kids' lives. But not all dads are the average dad. Some dads are long term out of work and contribute nothing to the family’s finances. They don’t pick up after themselves and add to the household chores rather than help with them. They are addicted to drugs or alcohol, they’re violent, they make for terrible role models, and they burden the psyche of the already-overextended mother with their presence.
And Kearney brings up evidence to support this claim. There is a study from Ohio that measures the difference in outcomes between the children of men who committed the same crimes but receive different sentences. They indicate that if a parent is incarcerated and therefore taken away, then their child is less likely to go on to commit a crime themselves. She also cites the ethnographic work of two sociologists who interview 162 single mothers and write their findings up in a book called “Promises I can Keep: Why Poor Women Put Motherhood before Marriage.” They suggest that many women don’t marry their children’s fathers because they do not see them as a reliable source of economic security and stability.’
Obviously, lots of dads who leave aren’t total criminals. But clearly, some parents are more dedicated to parenting than others. The American Time Use Survey has found that highly educated parents spend less time on leisure activities, sleep and household chores. They spend that extra time with their children even though they also work more hours. I don’t know if this itself is going to make a big difference on a child’s outcomes but I do know that conscientiousness is heritable.
People grow up to be like their parents
Genes and culture are both important and I am going to controversially suggest that people often grow up to be like their parents.
Except for some rare extenuating circumstances, I think fathers who leave their children and skip out on their child support payments are bad people. I think women who have unplanned pregnancies with unreliable men make bad decisions. I think people pass on traits like criminality and impulsivity to their children. Similarly, I think a child born to hyper-educated and conscientious parents inherit some traits that make them more likely to go to university.
If we could wave a magic wand and guarantee that every child grows up with both of their parents, there would still be a disparity in outcomes between the children of the elite and the children of high-school dropouts.
What’s the causal direction?
There is good evidence that suggests that it’s definitely not just bad genes. In the US, different states brought in more liberal divorce laws at different points. Using data from the censuses from 1960 to 1990, before and after the laws were being implemented, you can see that these legal changes increased the number of divorces and subsequently, children had worse levels of education, lower incomes, and more marital churn themselves when compared with similar children in states with stricter divorce laws.
This suggests to me that the act of parents splitting up, not just the traits that make splitting up likely, has a deleterious impact on children.
Girls don’t like boys, they like cars and money
So why is marriage in decline? In the book Kearney puts across one main theory. She says that women don’t really want men around. They like babies and they need sperm for that. They like money, and often men can help with that. But, if men aren’t earning that much more than them, they can make off with their sperm and be happier without them. She hasn’t quite put it like that. (I don’t even think this is what she believes. If I pressed her on it, I’d guess that she would say that she thought cultural factors were more important but she’s an economist, and therefore only argues for the theory that she has data for.)
She says:
Since the 1980s, economic trends have favoured workers with college degrees, making it harder for those without a college degree to obtain a secure footing in the workforce with a steady, well-paying job. Men without college degrees have seen their earnings stagnate and employment rates fall. Women, meanwhile, have experienced increases in average earnings regardless of their education level. This change has stripped many men of their traditional role as breadwinner for the family and, in simple terms, made them less desirable marriage partners.’’
What this means is that between 1990 and 2020, college educated women earned about 70% of what college-educated men earned. Now they earn 76% of what college-educated men earn and this corresponds with a modest drop in marriage rates for college-educated mothers. On the other hand, non-college-educated women went from earning 54% of what non-college-educated men earned to 74%. This has corresponded with a much greater decrease in their marriage rates. Also, in areas where women tend to earn more than men, marriage rates are lower.
She then looks at papers that track marriage rates in the places that have been most heavily hit by declines in manufacturing and sees that yes, a decrease in manufacturing jobs and an increase in the number of robots used – something that we should expect to impact male employment more than female employment – decreases the number of new marriages and increases the number of births to unmarried women.
However the opposite doesn't seem to hold. Areas that benefited from the fracking boom, i.e, places where the employment of non-college-educated men was boosted more than women’s incomes, they found that there was an associated increase in the number of babies born but there was no marked increase in marriages.
This is plausibly because, after marriage rates fall, social norms shift, and people are less likely to want to get married at all. During the coal boom of the 70s and 80s, when single-motherhood was rare, the marriage rate went up as well as the number of children born.
Of course this model can’t explain what’s happening fully – 74% is still less than 76%. If the underlying cause was truly about relative earnings, then we should see a world in which poor women were more likely to be married. In fact, considering the diminishing utility of money, we should expect poor women to be much more motivated to marry, regardless of their relative incomes.
Marriage Penalties
The book completely ignores marriage penalties. If people get poorer by getting married to each other, it makes perfect sense that they wouldn’t then get married. I think it is a massive oversight that she doesn’t discuss this.
If you assume that 20 to 49 year old Americans would marry someone who earns the same as them, the average single American is facing a potential marriage tax of 2.69%. The top-line federal income tax brackets are marriage neutral, but other features of taxes and benefit programs penalise marriage. As a result, the marriage penalty is higher for poorer families who receive more in benefits. So for the bottom quintile, who presumably need the money more, the average marriage tax is 3.71%. For the top quintile it is 1.49%.
They also found that a one percent increase in marriage penalties decreased the odds of getting married later by 3.69% in women with children. (And 1.12% for women without.) For a 25-year-old woman, in the bottom quintile of earnings, with a child, her marriage tax rate reduces her chances of being married at 35 by 7.52%.
There are lots of things we need to consider when designing taxes and benefits. Most policymakers want
The children of poorer families to have the best chance in life – which will often mean removing barriers that stop their parents from marrying and staying together
Women in abusive relationships to feel able to escape their situation without worrying about losing benefits for them and their children
Benefits to be progressive, meaning we give more to people on lower incomes
You could achieve 1 and 2 with more UBI-like systems but those are very expensive. Amber Lapp of the Institute for Family Studies suggests here that they should double the threshold for benefits like Medicaid for married couples. I’m sure this would create some new distortions in the system, but they probably wouldn’t be as bad as the distortions we currently have.3
The end of shotgun marriages
With the rise of abortion came the death of the shotgun marriage. Until about 1970 if a man got a woman pregnant, the expectation from all parties – the couple, her family, his family, the broader community etc – was that he would marry her.
Janet Yellen’s theory is that before abortion and widespread contraception, having a child out of wedlock was so damaging to women that they would only have sex with men if they believed that, in the event they got pregnant, the man would step up and marry her. This meant that women were extracting these promises from men, only having sex with them if they thought they would follow through, and other men were enforcing this norm because they benefited from this norm existing. Close association with a cad and charlatan would probably damage your own standing among the women in your community. If you were married your wife would be angry with you for remaining friends with such a person. If you were single, the taint of this man would rub off on you and you’d struggle to get a girlfriend.
The rise of contraception and abortion meant that women who didn’t mind using contraception, and didn’t have an ethical objection to abortion, no longer extracted the promise of marriage from the men they slept with. This undermined the collective bargaining power of women and sex without commitment became a mainstream relationship norm.
Now, as the choice of whether to have an abortion is primarily the choice of the pregnant woman, a man could reasonably say “I wanted her to have an abortion. I don’t see why I’m on the hook for her choice.” He doesn’t feel the need to step up. And men and women who are happy with contraception don’t want to pressure him to.
And so, outside of isolated communities where abortion is frowned upon, shotgun marriages don’t really happen. Despite the hardship of solo-parenting, lots of women still want children and they are still having these children alone and so begins the “feminization of poverty” that comes from the uptick of single-mother households.
What does Kearney think we should do about it?
This is the weakest part of the book.
In summary, here are the things we should do to address the challenges I have laid out in this book
Work to restore and foster a norm of two-parent homes for children
Work to improve the economic position of men without a college level of education so the are more reliable marriage partners and fathers
Scale up government and community programs that show promise in strengthening families and improving outcomes for parents and children from disadvantaged backgrounds
Have a stronger safety net for families, regardless of family structure
Here are some things I do not think we should do
Accept a new reality where the two-parent family is a thing of the past for less educated, lower-income Americans
Bemoan the economic independence of women
Stigmatize single mothers or encourage unhealthy marriages
Run unsuccessful government marriage programs
Keep government assistance meager under the mistaken assumption that doing so will incentivize more marriages and two-parent families
I don’t think that any of these policy solutions make sense with the evidence she presents throughout the book
Boost men’s wages
“I want to be explicit in stating the following: the growth in women’s earnings and economic opportunities is a positive social trend, unequivocally.”
And then talks about ways she wishes to bolster the earnings of working class men to make them more marriageable. I think it’s worth emphasising a fact from earlier in the book where she acknowledges women still earn less than men. While the gap is shrinking, it still exists. Sure, maybe she wants to widen the gap by increasing men’s wages instead of by depressing womens’ – but she keeps reiterating that women gaining financial independence is a good thing. You can’t have your cake and eat it.
Also, earlier in the book she shows that in areas that have already experienced the collapse of marriage, boosts to men’s earnings, relative to women’s earnings don’t boost marriage rates anyway.
Criminal Justice Reform
She then talks about how we should do criminal justice reform to decrease the number of American dads in prison.
“Criminal justice reform and prisoner reentry efforts will also be needed to reduce the number of American men who spend time incarcerated and to help those who do have criminal backgrounds reintegrate into society and the workforce.”
Earlier in the book she uses fathers with criminal convictions as an example of how some children are not better off with their fathers around. She points to studies which show that when fathers escape prison due to lenient judges their kids are worse off.
Useless Programs
She then writes about a series of Bush-era policies that tried to strengthen marriage.
“An obvious way to help strengthen families would be to run programs that encourage stable marriages. But such efforts have been tried and they generally did not work well. Early in the first decade of the 2000s, the federal government took up the cause of trying to encourage healthy, stable marriages by funding community programs with that goal. The Healthy Marriage Initiative, launched by the Bush administration in 2001, provided federal funding for voluntary programs run by local and state governments and community organizations to promote marriage among low-income couples with children. The results of these studies were not very encouraging. The programs appeared to help people in some ways, but they didn’t meaningfully increase marital stability among participating couples.”
So instead she says that we should have fatherhood promoting programs instead of marriage promoting programs.
“There is not a trend toward programming that promotes strong families (not necessarily marriage) and engaged fatherhood. Fatherhood programs are being designed and implemented around the country, as discussed in chapter 6. Some fatherhood programs work with dads on parenting techniques and conflict-management skills. Others additionally work to address personal barriers in the form of unstable employment or previous incarceration. There is still much to be learned about how to design these programs in such a way that they produce meaningful improvements, but it seems clear that building on and experimenting with programs that engages fathers and that work to improve parental cooperation and active co-parenting between mothers and fathers is more likely to be a successful approach than programs that focus on encouraging or sustaining marriage per se.”
But she’s already shown, in the book, that these programs don’t work either. She mentions one study about one of these programs and it shows that men who participated had better “self reported nurturing behaviour” but didn’t actually spend any more time with their children. Basically, they learned the right, researcher-approved language for talking about their kids.
Welfare
Then, after saying that in order to address inequality we need much more generous welfare payments for all parents – single and married alike. But after that she says,
“Even in Denmark, a bastion of public welfare that includes free college tuition, universal access to high-quality health care, universal high-quality pre-K, and a generous childcare and maternity-leave policy, the influence of family background on many child outcomes is about as strong as it is in the US.”
Basically, everything Kearney proposes to counteract the decline of the marriage contradicts other things she says. On her own terms, she has nothing to offer.
That shouldn’t surprise us. Basically everything I’ve listed above – more welfare, less incarceration, raising wages for people without college, spending money on random ineffectual feel-good programs – is just part of your average Democrat platform. She’s written a book about something the Democrat establishment doesn’t really care about and then concludes that all their solutions were right anyway.
Based on what Kearney has said, if I were serious about countering the decline in marriage in the working class US, I would not conclude that more of the same was good enough. I would probably pay people to get married – either through tax breaks or some kind of cash payment, and then I would make it difficult to get divorced.
But I also don’t think she’s convinced me that the main problem even is the decline of marriage. I’ve been convinced that on the margin, more people should get married and once married, we should stop them from getting divorced.
But I don’t think the main issue here is that people aren’t marrying enough. I think it’s because there aren’t enough good men in the US. 9% of all American men are incarcerated at some point in their lives, 19% still live with their parents and 15% don’t work. Who knows how many are addicts, abusive, or criminals who haven’t yet been caught? What right-thinking woman is going to take on that burden?
Now, I don’t expect that men were better in the past. They were probably just as violent and useless – if not more so. But it used to be much worse to be an unmarried woman. Women were paid less, single mothers were much more maligned, and it was easier to fall into life threatening poverty.
The fact that it’s possible, through a combination of welfare and job opportunities, to live an adequate existence without male support means that lots of women and children are much better off. The consequences aren’t “unequivocally good”, but I don’t think they’re so dire that I’d wish to reverse this trend.
I think that the decline in marriage is, in large part, the result of women looking at what possible men are on offer and saying “no thanks.” It’s difficult to know in what proportion of these cases, the women are making choices that are good for them and their children, and in what proportion of these cases they are missing out on resources, love, and support that would make everyone better off. But I think that a lot of these women are canny. At the point at which these children are born, the women are less likely to think that they will eventually get married than the men do. They don’t seem to be choosing motherhood while completely ignorant of what their future is going to look like.
Kearney, towards the end of her book, has a story about an Uber driver she talked to. He said that he has a daughter, but she’s not with his kids mother. Kearney asks why and he gives a half-hearted answer and says that maybe they’d get married and live together if he saved up enough money. She concludes that surely, if he loves his daughter and gets along with his ex, they could make it work.
I also have an Uber driver story — who the fuck doesn’t. When I was in New Mexico, I had a half hour journey with a driver, a 29 year old woman who I’ll call Jane. Jane was a single mum with a seven year old son. She had been training to be a nurse but during the pandemic her son’s nursery closed down so she switched to doing UberEats deliveries so she could take care of him while working. He’d hang out with her in the back of the car.
Jane and I talked about boys. A man she had a situationship with tried to call her and she rejected the call to keep talking to me. The problem with him, she said, was that he just wasn’t responsible enough. He had a factory job, but he had only just started it and had a history of not sticking with anything. He’d been arrested a few weeks ago for a DUI. “He’s just not in the position to be a good step-father.”
I then asked her about her son’s father. She said that she was twenty one when they were dating and he had been much older. She found out, after she was pregnant with her son, that he had been lying to her about his age and mysterious “other things.” She broke up with him and moved in with her mum.
I think Jane has done the right thing for her son by breaking up with his father. Her ominous tone implied to me that he had done something quite nasty. If I had to guess, I’d say he was sleeping with Jane when he already had a wife and family at home. I think it’s likely that Kearney’s Uber driver, even if he seemed like a nice man for the twenty minutes they were together, was, like the men in Jane’s life, too unreliable to be a father.
In the UK, half of all children are now born out of wedlock. Because the decline of marriage has been recent, 65% of families are still headed by married couples.
This effect means that it’s difficult to make comparisons across time by education levels. Here’s an interesting X dot com thread from an economist who was looking at mortality and education levels, and he explains how he tried to correct for this effect.
Universal credit in the UK doesn’t have quite the same problem. Couples are assessed for their benefits if they live together, regardless of whether they are married. This probably creates a disincentive from cohabiting with a partner if you’re a single mother on universal credit who is thinking about moving in with a man who earns a salary that he may be unwilling to share with you. But once you’re living together there’s no further block on marriage.
This is a great review and discussion.
The question that occurs to me now, which I almost can't believe I'm asking: Is single parenthood bad for children?
Could it be that we always had a distribution of circumstances and outcomes for kids, some better than others, and that the real change is that today children on the bottom end of the distribution are more likely to have unmarried parents, but that it's not the unmarried part that's doing the legwork?
I'm not 100% sure how you'd get at this question, but widening gaps in outcomes over time as single parenthood became more prominent might be one signal.
If I remember right (which maybe I don't), Charles Murrary once looked at children who had a single parent due to the death of a parent vs the more conventional absent second parent, finding that children of a deceased parent had better outcomes. That would support the idea that it's factors correlated with non-marriage, not the non-marriage itself, that is causal.
Beautifully written.
I agree that a significant portion of American men are not marriage material. However, women's expectations of their partners are quite high. I would argue too high. They are much more likely to leave a less-than-ideal partnership, albeit not abusive, regardless of its negative impact on their children.
Interestingly, gay male marriages are the most stable while lesbian divorces far surpass straight divorces.
Men need to step up. And women need to lower their standards, just a bit.