This cohabitation-causes-bad-parenting thing

W. Bradford Wilcox is getting bolder in his public campaign against cohabitation, and it’s moving him further from the evidence.

Today’s NY Times Room for Debate follows the last week’s appearance on KPCC radio, and his claim of a causal effect of cohabiting on parents’ parenting is growing more expansive.

Last week on the radio he was slightly cautious:

Part of that is what we call selection in the sciences, where certain types of people are more likely to select into a cohabiting relationship, and these are folks who generally have less economic resources and less commitment to each other. But there also seems to be kind of a causal effect of cohabitation, where the very fact of being in a relationship without that same degree of commitment and that same level of trust, and fidelity for instance, that in turn makes them more likely to be poorer functioning parents.

This week the cautious tone is gone:

But is cohabitation really the problem, or some deeper factor — like poverty or relationship troubles that predated the cohabitation? The truth is that these other factors account for some of cohabitation’s negative impact but the best studies suggest that cohabitation also has an independent negative effect, precisely because it does not institutionalize commitment in a way that is easily understood and honored by romantic partners and their friends and family.

Now the causal effect is found in “the best studies,” which somehow also are able to distinguish the reason for this causal effect, “because it does not institutionalize commitment…”

This is a dishonest representation of the scholarly record.

I can’t find any study that accounts for selection into cohabitation while finding a negative effect on parenting or children’s outcomes — and the claim that any negative effect is “because it does not institutionalize commitment” is essentially impossible to substantiate.

After Wilcox mentions “the best studies,” he quotes one that does not make a claim for an independent effect. He writes:

Children in cohabiting families are about twice as likely to drop out of high school, use drugs, or end up depressed, compared with children in intact, married families. They are also at least three times more likely to be physically, sexually or emotionally abused, according to a recent federal report.

But that study includes this:

Factors such as parents’ labor force participation, household socioeconomic status, family size, and family structure and living arrangement are not only associated with the incidence of maltreatment but are also correlated with each other. Further analyses could determine their independent relationships to maltreatment, such as whether households with more children have higher incidence rates even when household socioeconomic status is taken into account.

In a comprehensive 2008 review in Annual Review of Sociology titled “Family Structure and the Reproduction of Inequalities,” Sara McLanahan and Christine Percheski wrote:

Do children from families with two stably cohabiting, unmarried parents fare as well as children from married parent families? The evidence so far suggests they do not (Brown 2004, Artis 2007), although the research on this topic is limited.

If you follow those two references, you find that both scholars responsibly point out that they do not account for selection into cohabitation:

  • Artis (2007): “ these data cannot address how selection into marriage may influence these patterns.”
  • Brown (2004): “Selection likely plays a role in the family structure and child outcomes relationship but could not be addressed because of data limitations.”

On that radio show, Stephanie Coontz did a great job of explaining the risks of “causal generations,” and in today’s piece she bluntly titles her contribution, “Cohabitation Doesn’t Cause Bad Parenting.”

Sharon Sassler summarizes the actual evidence correctly, as I see it:

But a growing body of evidence suggests that any advantages from marriage are more a result of selection than causation. In other words, the most educated and economically established adults are the most likely to wed, and overwhelmingly defer childbearing until after marriage. The benefits redound to their children.

Unfortunately, the one-side-other-side format of these debates just provides fuel for Wilcox’s campaign, reducing the “debate” to a mind-numbing he-said-she-said in which the public is encouraged to choose the evidence that fits their preconceptions. In fact, Wilcox makes a direct appeal to the Times‘ readers’ preconception:

Anyone who disagrees should answer this question: When was the last time you saw a cohabiting couple enter their relationship by vowing, in front of their closest friends and family, to love and cherish one another, for better for worse, for richer for poorer, in sickness and in health, till death do they part?

That is just ridiculous, a public relinquishment of his credential as a social scientist. Teaching people not to reason like this is one of the great contributions of sociology.

Today’s cohabitation-causes-bad-parenting is the 2010s version of the single-mothers-and-welfare-cause-crime hype from the 1980s. Both are dressed in a stated concern for the well-being of poor people, but both speak to the non-poor, creating a stigmatizing social distance from the poor, undermining real efforts to improve their conditions and reinforcing the “common sense” belief that systemic poverty results from poor individuals’ intimate decisions.

17 thoughts on “This cohabitation-causes-bad-parenting thing

  1. Point 1: Just because there is no causation does not mean that the correlation is not a strong marker of offspring drug abuse, school dropout rates, etc, etc, etc.

    If in fact they are strongly correlated. Just be honest and say, “these are strongly correlated”.

    Point 2: I wonder what the actual ratio of “stably cohabiting, unmarried parents” (SCUP) to “married parent families” and ratio of SCUP to “unstably cohabiting, unmarried parents” . If these ratios are small, then that also should tell us something.

    Point 3: It wasn’t too long ago that poor people actually *did* get and stay married for long terms. What happened?

    Point 4: The Moynihan Report warned 45 years ago of the ill effects of non-nuclear families.


  2. Could you explain why Wilcox’s statement is ridiculous?

    I’m talking about this statement: “Anyone who disagrees should answer this question: When was the last time you saw a cohabiting couple enter their relationship by vowing, in front of their closest friends and family, to love and cherish one another, for better for worse, for richer for poorer, in sickness and in health, till death do they part?”


    1. It’s ridiculous for two reasons. First, it’s a marriage “vow,” so asking whether cohabiting couples take it is just meaningless (and, as many people have pointed out, married people are much more likely to have *broken* this vow…)

      But second, and more to my point about social science, is that good social science shows us that our personal impressions and anecdotes are highly susceptible to selection and bias of many kinds. Social science aims to get beyond simple personal impressions and answer questions socially – not individually.

      This is not just misleading but also pernicious in Wilcox’s case in the NYT, because he is speaking to a well-educated (on average) readership which is less likely to have direct experience with the poverty, insecurity, uncertainty and stress associated — and with family decision-making under these conditions. So it is an “othering” strategy, asking us to reinforce our perceptions that people who make these decisions are not like us. I believe this is harmful.


      1. What if the people who make those decisions actually are *not* like “us”?

        The problem is in accurately defining “us”.


      2. Can we point to any qualitative data that shows poor and working class young people perceive cohabitation as an arrangement that means to love and cherish the other for better or worse, for richer or poorer, in sickness and in health, until death do us part?


  3. Yes, precisely. And, once the Wilcox and the rest of the Christian Right defines “us” as rich, white, heterosexual, conservative Christians, they can get on with the business of eliminating all social support for anyone else. You want your social security? You’ll have to show your Christian Card. Parent’s aren’t married? Sorry, no Christian School Voucher for you!


    1. Except that there are quite a large number in the Christian Right who are raising their grandchildren and even great-grandchildren because their children and grandchildren are meth heads.


  4. In response to this post, W. Bradford Wilcox sent me an email asking me to post this reply. It is unedited below. I will have a new post with some responses in the morning.

    * * *

    Given the frequency and intensity with which Philip Cohen addresses my work and public comments (including yesterday’s post), I would like to point out some common ground before turning to areas of obvious disagreement.

    First, Cohen and I are both concerned with growing social and family inequality, as well as the welfare of children. And we both agree that shifts in the contemporary economy and polity have made it more difficult for adults to establish stable, high-quality relationships that redound to their own benefit and to the benefit of their children. This is a line of argument I made in When Marriage Disappears (2010), and I assume Cohen would agree with this strand of the argument about contemporary families.

    But we disagree about whether family structure per se has an impact on the welfare of adults and children beyond the factors that help to determine who “selects” into marriage, single parenthood, cohabitation, etc. In fact, at the end of his post, Cohen makes a dismissive reference to the “single-mothers-and-welfare-cause-crime hype from the 1980s.” Evidently, Cohen believes that single parenthood per se has nothing to do with crime.

    Say what? What about Harper and McLanahan (2004), LaFree et al. (2010), and Sampson et al. (2005), among others? What does Cohen make of Sampson’s (1995) observation that “family structure is one of the strongest, if not the strongest, predictors of… urban violence across cities in the United States?”

    More generally, what does he think about this conclusion offered by Haskins, McLanahan, and Donahue (2005) about contemporary family trends and children? “Although it was once possible to believe that… high rates of… cohabitation and nonmarital childbearing represented little more than lifestyle alternatives brought about by the freedom to pursue individual fulfillment, many analysts now believe that these individual choices can be damaging to the children who have no say in them and to the society that enables them.”

    Are Sampson and McLanahan succumbing to 1980s-style right-wing hype? Of course not.

    Rather, their research and their conclusions about that research suggest that family structure per se has played a role in fueling the very social inequality that Cohen is rightly concerned about and has also had an independent effect on the welfare of children and adults.

    This brings me to cohabitation. What has struck me in reading the work of Artis (2007), Brown (2004), Cavanagh (2008), and others is that the association between cohabitation and negative social and psychological outcomes like externalizing, depression, and drug use is comparable to the association between single parenthood and these outcomes. And that the association persists even after controlling from obvious “selection” factors like parental education, race/ethnicity, and income. There has also been good work (e.g., Schnitzer and Ewigman [2005]) indicating that the association between step-cohabitation and child abuse persists even after controlling for potential selection factors like race, age, family size, and access to health care. Indeed, the link between some outcomes and parental cohabitation (Cavanagh 2008) persists even after controlling for a host of factors–including family instability and parenting behaviors. From my perspective, then, we have growing evidence that cohabitation–especially step-cohabitation where an unrelated adult male enters the household–is causally implicated in some negative outcomes for children.

    We will have to wait for more longitudinal studies, twin studies, etc. to nail down the extent to which the cohabitation-negative child outcomes story is about selection versus causation. But I’m confident that it’s both, even though Cohen seems to be confident it’s only a selection story.

    Finally, Cohen is scandalized that I ended my New York Times piece by making the point that cohabiting couples do not typically enter their relationships by making commitment vows to one another in front of friends and family. The reason I did this was to underline the fact that cohabitation is insufficiently institutionalized. There are very few social rituals, norms, or commonly-understood practices to help guide and stabilize cohabiting relationships. Indeed, Stanley et al.’s (2011) research finds that most cohabiting couples report that they either “slid into” their relationship, or it “just sort of happened,” and that the provisional character of their relationship helps explain the instability and lower quality of cohabiting relationships, relative to marriage.

    Cohen seems to think children and adults can get by well enough without the durable domestic institution of marriage to lend meaning, direction, and order to family life–especially the bearing and rearing of children. By contrast, I am one of those “analysts” who believes, in the words of Ron Haskins, Sara McLanahan, and Elizabeth Donahue, that “high rates of… cohabitation and nonmarital childbearing… can be damaging to the children who have no say in them and to the society that enables them.”


Comments welcome (may be moderated)

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s