Friday, September 19, 2014

Understanding Why Marriage is Associated with Positive Child Outcomes: Commitment, Signaling, and Sequence

by Scott Stanley

Why is marriage associated with positive child outcomes when compared to non-marriage? In The Marriage Effect: Money or Parenting?, Kimberly Howard and Richard Reeves argued that the primary mechanisms through which marriage benefits children are based on income and parenting quality. Emily Badger wrote a piece on their work with a title that captured what the take-away was for many: Children with married parents are better off— but marriage isn’t the reason why. While I agree with a number of their points, I believe something substantial is missing in Howard and Reeves’ overall argument. Brad Wilcox outlined one set of concerns. My focus here is in a different direction—on the development and timing of commitment.

Howard and Reeves unflinchingly declare that there are substantial advantages for children raised by married parents compared to others. While I think the notion of “settled science” is conceptually dangerous, this fact seems broadly recognized.[i] Howard and Reeves are particularly interested in two questions: (1) “is it the marriage itself that matters? and (2) if not, what do we do to help? Their findings show that the association between marriage and positive child outcomes may be substantially accounted for by greater income and more engaged parenting among marrieds. Based on this, they argue that intervention efforts should focus on parenting and not on marriage, per se.

I respect this logic. My colleagues and I have long argued that relationship oriented interventions should focus on variables that are dynamic (i.e., putatively changeable) rather than those that are relatively static.[ii] That may seem obvious, but people sometimes misunderstand why prevention-focused experts will put more emphasis on variables of lessor predictive power that are arguably dynamic (e.g., the number of sexual partners) than on variables that account for more variance but are immutable (e.g., race). Knowledge of static risk factors is nonetheless also important because it points to where to concentrate efforts to help people. For example, while there has been a lot of press on the meager outcomes from recent federal studies on relationship education provided to couples at lower incomes, there is some good news about who may benefit the most related to relatively static risk factors.[iii]

Marriage: a Mere Commitment Device?

Emily Badger quotes Reeves on the income and parenting engagement points: 

"Those two factors taken together explain most of the better outcomes for the children of married couples," Reeves says. "Not all. But most. And I think the takeaway here is not to mistake a commitment device – which marriage is – for an explanatory device."

The argument is further clarified in this quote from Howard and Reeves in their original piece: “Is it marriage itself that matters, or is marriage the visible expression of other factors, that are the true cause of different outcomes? And if so, which ones?”

I think this argument shows a serious under-appreciation for the importance of “visible expressions” of commitment. Signals of commitment are important across a wide swath of societal life because people will often make better decisions with clearer information about the level of motivation in others,[iv] and signals about commitment are, arguably, of great importance in the development and maintenance of romantic and family relationships.[v] Reeves seems to be arguing that the signal value of marriage is not as consequential as behaviors such as parenting, but what that view fails to account for is how marriage has most typically been a potent signal of commitment with a distinct placement regarding the sequence and timing of childbearing. At the root of it, what is signaled by marriage is a commitment comprised of “us with a future.”[vi] Sure, reality has very often been messier than the tidy ordering of love, marriage, and a baby carriage; and many marriages do not go the distance. But marriage is likely, in some large respect, explanatory regarding child outcomes because marriage most often is a strong and credible signal of commitment prior to childbirth.

Put another way, Howard and Reeves seem to focus on childrearing (parenting) with little emphasis for the role marriage often plays in sequencing of commitment and childbearing. I believe that the quality of the parenting a child will receive is situated in the context of the level of commitment his or her two parents have to parenting together. Danielle Kurtzleben at Vox.com highlighted one key part of this puzzle related to that idea of “together”:

There is a common-sense reason to why this bump is so great. A pair of mediocre married parents will have way more time to spend with their kids than even an exceptionally devoted single dad . . .

Even here, there may be an under-appreciation for how (and if) the partnership to parent formed in the first place. The fact is that marriage is associated with a far greater likelihood that a child’s parents will continue to parent together than other contexts.[vii] At one end of a spectrum are parents who married before the child arrived, where those children have a relatively high likelihood of receiving extensive co-parenting. At the other end of this spectrum would be children born to parents who had not developed much, if any, commitment to each other beforehand, much less a commitment to parent a child together before having one. Those children, on average, have more of an uphill climb in life, and, as Howard and Reeves note, economic and social mobility are impacted. Such children are not disadvantaged because they don’t have a parent who cares, but because they are going to land, on average, the furthest from having the economic and social capital of two people pulling together to start them off in life. One can argue that the benefits of having two committed parents can exist apart from marriage. I agree. So why would I argue that marriage has special explanatory value regarding child outcomes?

Back to signals and sequence.

While not always, and perhaps less so now than before, marriage serves as a strong signal that two people are tacitly committed to raising a family together. Further, and for more complex reasons than I want to develop here, signals are the most informative when they are fully under the control of those sending them—by which I mean, when the behavior has fewer prior constraints so that it reflects something true about the individual. That means that signals about commitment are more informative before a child arrives than after because having a child increases life constraints. When marriage precedes two people having a child, the question of intention about a shared long-term time horizon was settled before things got messy with baby drool and poop. For couples with this foundation already in place, even unplanned and mistimed children are still landing in a relatively rich context regarding bi-parental commitment. One can (and should) believe that various socio-economic disadvantages govern a lot in this big lottery of life, but we should not lose sight of how sequence plays a consequential and causal role in child outcomes.

I am far from alone believing this. I think the greatest change in families impacting children in this era is that so many are born into low commitment contexts. This seems to be exactly the point that Isabel Sawhill argues in her forthcoming book, Generation Unbound: Drifting Into Sex and Parenthood Without Marriage. In her New York Times piece on 9-14-2014, she wrote:

We’ve been worrying about these trends for years, and wondering: Can marriage be restored as the standard way to raise children? As much as we might welcome a revival, I doubt that it will happen. The genie is out of the bottle.

I would love for Sawhill to be wrong about marriage, but I share her pessimism. Further, by arguing for what is needed, Sawhill draws attention to what is increasingly missing.

What we need instead is a new ethic of responsible parenthood. If we combine an updated social norm with greater reliance on the most effective forms of birth control, we can transform drifters into planners and improve children’s life prospects.

In her book and article, Sawhill focuses a lot of attention on complex issues related to birth control. I will sidestep that issue for now to focus on drifting versus planning. This is familiar territory for me and my colleagues. Whether you think about drifting versus planning or sliding versus deciding,[viii] the underlying point is that it matters how and when (and if) intention forms when it comes to the consequences of life altering transitions such as having a child. Commitments are decisions, and decisions support the strongest follow-through.

What about Howard and Reeves’ finding about engaged parenting? They note that “It is plausible that parents who commit to each other through marriage may also have a stronger joint commitment to raising their children.” That’s exactly what I believe is being given short shrift in the current discussion. In fact, I suspect that their parenting variable is partly a proxy for the mutual commitment to parent that is implicit in marriage.

While I can see plenty of value in efforts to provide more education about parenting to both couples and single parents, I also believe we need to work to increase the odds that children are born into high commitment contexts. Such efforts might include helping people better understand how sliding into having a child together, in a relationship with an unclear future, leads to worse outcomes for adults and children. Emphasizing this reality may be unpalatable to some who worry that such messages can be retroactively stigmatizing for those who are already downstream from consequential drifting. If so, the importance of emphasizing this may be as controversial to some as Isabel Sawhill’s suggestions about birth control are to others. Either way, it does not reflect how life really works to ignore sequence as we all grapple with solutions.

Marriage is, indeed, fading in front of our eyes, and with it goes a lot of signal clarity about commitment in the context of sequence. Maybe those elements can be constructed behaviorally on a broad scale, but we should recognize the difficulty we face in trying to make up for the loss of something with real explanatory power.




[i] Sawhill, I. V. (2014). Generation unbound: Drifting into sex and parenthood without marriage. Washington D.C.: Brookings Institution Press. 
[ii] e.g., Stanley, S. M. (2001). Making a case for premarital education. Family Relations, 50(3), 272-280.;  Markman, H. J., Stanley, S. M., & Blumberg, S. L. (2010). Fighting for your marriage. San Francisco: Jossey-Bass. NOTE: It is a separate and challenging matter in social interventions to demonstrate that the variables targeted are the mechanisms of change. 
[iii] Who Benefits Most from Family-Strengthening Efforts?
[iv] The seminal paper by Nobel Prize winner Michael Spence is: Spence, A. Michael. (1973). Job Market Signaling. Quarterly Journal of Economics, 87(3), 355-374. In his Nobel award speech, Spence noted that that “the information carried by the signal can be productive itself. This will occur if there is a decision that is made better or with greater efficiency, with better information.” I believe this is relevant to the points I make here (though I make no claim to understanding all of the nuances of Spence’s work).
[v] Rowthorn, R. (2002).  Marriage as a signal.  In A. W. Dnes and R. Rowthorn (Eds.), The Law and Economics of Marriage and Divorce (pp. 132 - 156).  New York: Cambridge University Press.; Nock, S.L.  (2009). The Growing Importance of Marriage in America.  In H. E. Peters and C. M. Kamp Dush (Eds.), Marriage and Family: Perspectives and Complexities (pp. 302-324). New York: Columbia University Press.; Stanley, S. M., Rhoades, G. K., & Whitton, S. W. (2010). Commitment: Functions,formation, and the securing of romantic attachment. Journal of Family Theory and Review, 2, 243-257.
[vi] Jones, W. H., & Adams, J. M. (1999).  Handbook of interpersonal commitment and relationship stability.  New York: Plenum.;  Stanley, S. M., Rhoades, G. K., & Whitton, S. W. (2010). Commitment: Functions, formation, and the securing of romantic attachment. Journal of Family Theory and Review, 2, 243-257.
[vii] Sawhill, I. V. (2014). Generation unbound: Drifting into sex and parenthood without marriage. Washington D.C.: Brookings Institution Press. 
[viii] Stanley,S. M., Rhoades, G. K., & Markman, H. J. (2006). Sliding vs. Deciding:Inertia and the premarital cohabitation effect. Family Relations, 55, 499-509.

Wednesday, September 3, 2014

Which Door? Thoughts on The Consequential Choices of Researchers and Other Humans


In the frenzy of all the attention regarding the National Marriage Project report, co-authored by Galena Rhoades and me, and recently released, we wrote a conceptual piece posted at The Institute for Family Studies on issues that arise in how researchers think about causality and personal choice, and how statistical approaches intersect with what researchers do and what they conclude.

It's somewhat more technical than most things I post, but the piece is more conceptual than statistical if you are interested in the subject.

Which Door? Thoughts on The Consequential Choices of Researchers and Other Humans