Genes and Marriage: Their Claims, My Qualms

Anyone who wants to claim that getting married makes people happier or healthier or less depressed or anything else (and there are multitudes of people who want to do so) are up against a hard fact: Causality is impossible to prove. If married and single people differ, we can’t know for sure whether they differ because the married people are married, or because married and single people differ in some other important way. (For example, maybe married people have different kinds of personalities, or different access to resources through income or education, and those differences, rather than marriage, account for the ways the two groups differ. The possible ways in which married and single people might differ, other than in their marital status, are endless).

One way social scientists have tried to improve on the study of the implications of getting married is to study the same people over time (longitudinal research) rather than comparing groups of people at one point in time (cross-sectional research). If you follow people as they transition from being single to being married, and find that people who got married are lastingly happier or healthier than they were when they were single, then that’s better evidence for the supposed benefits of marrying than the cross-sectional alternative, though it is still not definitive.

Despite the widespread assumption among laypersons and social scientists that getting married results in better happiness, health, and so forth, the results of longitudinal studies are often less than compelling. (I reviewed many of them in a chapter in this book. I discussed studies published since then in blog posts such as this one and this one.) A recent article proposes a different approach to the causality issue.

What follows is a lengthy critique of the new study of the supposed benefits of getting married. There are five sections:

I A Relatively New Approach to the Marriage Question: Behavioral Genetics

II The Study and the Findings

III The Supposed “Marriage Benefit”: Two More Reasons Why I Don’t Buy It

IV What Does It All Mean? The Ideology of Marriage and Family

V Implications the Authors Draw from their Findings: Let’s Have More Government Spending on Promoting Marriage?

I A Relatively New Approach to the Marriage Question: Behavioral Genetics

The authors of a recent study (reference is below) took a different approach, involving behavioral genetics. They had access to a nationally representative sample of young American adults that included pairs of siblings who differed in their genetic relatedness. The 1,613 relevant pairs included monozygotic (identical) twins, dizygotic twins, full biological siblings, and half siblings, as well as cousins and genetically unrelated siblings. This allowed them to look at the importance of biological components, “shared environmental components” (for example, what siblings share when they grow up in the same family), and “nonshared environmental components” (what two siblings do not share even if they are identical twins – for example, maybe one marries and the other does not).

With those three components and the appropriate statistical models, the authors can estimate answers to the question of whether getting married really does result in any benefits to health or well-being, or whether any differences between married and single people were likely already present even before anyone got married, or whether there are actually no real differences at all between the groups. As the authors appropriately note, the conclusions are still not definitive. We can’t randomly assign people to get married or stay single, so we are looking for other ways to understand the implications of marrying. The authors’ idea to take a behavioral genetics approach is a promising one. They also have a great dataset to work with (though limited in important ways).

II The Study and the Findings

The title of the journal article is “Accounting for the physical and mental health benefits of entry into marriage.” In fact, though, as the authors acknowledge, they never directly compare married people to single people. Instead, they make two other comparisons: (1) They compare people who are currently married or in a marriage-like relationship (cohabitation) to those who are single; so, this is a coupled vs. single comparison; and (2) they compare married people to cohabiting people.

Those of you who are readers of Singled Out or of this blog have probably already noticed that the coupled people are the currently coupled. I bet you are already raising your red flags.

The authors actually acknowledged what they were up to, in their own way. They admitted that they excluded people who got married and then got divorced. We have seen this before. Social scientists do this unapologetically. Still, it amazes me every time. Let’s stop for a moment and consider what is happening:

In a study of the purported benefits of getting married, the authors excluded anyone who got married and then got divorced. So they are going to see whether getting married results in getting healthier or less depressed, but they are going to exclude anyone whose marriage was so unhappy that the couples refused to stay in it.

The authors think they did a good thing, because their alternative to including the divorced people was to mix them in with the people who stayed single. They realized that including the divorced with the always-single could make the single group seem to be doing less well than they really are. But it never seemed to occur to them that if they are going to study the implications of getting married, they need to include everyone who ever got married, and not only those who got married and chose to stay married. As the wonderful Eleanore Wells quipped on a recent radio interview the two of us did with an NPR station, “I doubt they were getting divorced because they couldn’t stand all that happiness” (my paraphrase).

The authors looked at 6 ways the marital status groups might differ:

• Physical health (participants rated their overall health)
• Cigarette use
• Antisocial behavior (theft, burglary, selling drugs, writing a bad check, etc.)
• Depressive symptoms
• Thinking seriously about suicide in the past 12 months (yes or no)
• Alcohol use (frequency of drinking, frequency of heaving drinking, etc.)

First, the results of the comparisons between the married people and the cohabiting ones: There was only one way, out of the 6, that the two groups differed that could not be attributed to selection effects (i.e., they already differed even before they got married or started cohabiting). The officially married committed fewer antisocial acts than the cohabiting couples.

Now let’s see how the couples (married plus cohabiting) differ or do not differ from the singles, once selection effects are set aside:

• Getting coupled did not result in any better physical health.
• Getting coupled did not result in any less cigarette smoking.
• Getting coupled did not result in any less anti-social behavior.
• Getting coupled did result in fewer depressive symptoms. The difference, though, was small. (For those of you who know statistical jargon, the difference between coupled and single MZ twins was just .13 SDs. The rule of thumb is that .3 is a small effect, so this is smaller than small.) The results of other studies also add important cautions and qualifications; see, for example, here and here and here.
• Getting coupled did result in fewer thoughts about suicide. We’re talking about suicidal thoughts, not actual suicides. Check out this discussion of marital status and actual suicides: Are married people less likely to kill themselves?
• Getting married did result in less drinking. Because routine drinking was included along with heavy drinking, we don’t know whether singles differed from couples in anything more than social drinking.

So, even after analyzing the data in a way that gave coupled people an unfair advantage (by excluding anyone who got married and then got divorced), this is all the authors could come up with in support of the supposed benefits of entering marriage.

I’m just getting started. There are other ways in which these supposed benefits, which the authors think they have established in a quasi-causal way, deserve even more skepticism.

III The Supposed “Marriage Benefit”: Two More Reasons Why I Don’t Buy It

Just a Honeymoon Effect?

Longitudinal studies, following adults as they transition from being single to getting married, sometimes show that getting married has no positive implications at all, or that any initial benefits decrease over time until the married people look the same as they did when they were single. (See, for example, Marriage and happiness: 18 long-term studies, and this study, described here, which showed that between four and six years after marrying or entering a cohabiting relationship, the coupled people were not any less depressed, they were not any happier, they were not any healthier, and they had no higher self-esteem. Instead, the couples remained more withdrawn from friends and family and neighbors.)

So it matters whether we are talking about marriages/partnerships in their initial years or longer-term unions. The participants in this behavioral genetics study were, on the average, 29 years old. The range was 24 to 34. Between 2001 and 2009, when participants entered the study, the median age at which Americans first married was between 27 and 28 for men, and between 25 and 26 for women. That suggests that on the average, these marriages were in their first few years. Any purported benefits could well be honeymoon effects that will disappear over time.

What Else Might “Cause” These Differences Other than Marriage?

Using their behavioral genetics approach, the authors have ruled out other explanations, such as selection, for any differences they found between the coupled and single people. What’s left is marriage. So if married people look better than single people in any way, it is because marriage “caused” that benefit.

Or is it?

There are alternative explanations. In the United States, marriage comes with more than 1,000 federal benefits and protections. Those freebies are not intrinsic to marriage – American lawmakers decided to add them to the marital package. Sometimes cohabitors in civil unions have access to special benefits and protections, too. Who does not benefit? Single people.

Because of those laws, and for other reasons as well, it costs more to be single than to be married. Maybe when single people are feeling down, they do not have the same access to mental health resources – they can’t afford the therapist, and they cannot be added to someone else’s health care plan at a reduced rate the way some married people can be added to their spouse’s plan.

Then, of course, there is all the singlism – the stereotyping and stigmatizing of single people, and the discrimination against them. How would single people fare if their lives were valued and appreciated as much as married people’s are?

Let’s consider once again the ways that single people did not differ from coupled people in this research, once red herrings such as selection were set aside. They were no less healthy, and they were no more likely to engage in antisocial behaviors or to smoke. If single people – who tolerate all sorts of singlism, who expenses are much greater than those of married people, and who are left out of the 1,000+ federal benefits that only go to married people – are doing just as well as coupled people in all of these ways, I think they are more than equal. I think they are actually more resilient than coupled people.

IV What Does It All Mean? The Ideology of Marriage and Family

The authors’ discussion of what their results mean is particularly telling, I think, with regard to the unacknowledged power of the Ideology of Marriage and Family. (For more on that, see here and here.)

Some of the problems are the typical ones that occur throughout the journal article. The authors have little doubt that marriage is beneficial, and so they repeat claims that actually are not well supported. For example, they believe that marriage protects against loneliness and social isolation. For what the literature on loneliness really does say, check out the chapter on singles in this book. Also, a growing literature on “greedy marriage” shows that it is single people, more so than married ones, who are likely to maintain ties and exchange support with parents, siblings, friends, and neighbors.

When the authors discuss the supposed superiority of marriage over cohabitation (remember, of the six measures, they found one that differed between the two groups), they reprise the popular nagging-wife hypothesis, though without the insensitive wording: “It is also possible that husbands and wives have or assume ‘permission’ to monitor their partner’s behavior more closely, fostering greater engagement in prosocial activities and less engagement in antisocial ones.”

There is something I always find interesting about authors who reach for this nagging/monitoring hypothesis to suggest some way in which married people will be better off: They almost never mention the research showing that getting married seems to result in getting fatter.

The authors also add an explanation that I don’t think I’ve seen before: Young adults who marry instead of cohabiting are better at delaying gratification. They take the long-term perspective as they think about having kids and buying houses and so they stay away from all that anti-social behavior. You may know the “delay of gratification” research as the marshmallow studies – some kids can put off eating the one marshmallow right in front of them in exchange for getting two marshmallows if they wait. The cohabitors, if I am interpreting the authors correctly, just can’t wait for the two marshmallows like the married people can.

Really, though, the authors do not want to be too hard on the cohabitors. They spend a paragraph describing ways in which the experiences of cohabiting and marriage vary – some relationships are better than others, they acknowledge: “In short, marriages and cohabiting relationships are both heterogeneous…”

If the Ideology of Marriage and Family were not so powerful, I think the authors may have devoted the same kind of attention to the experiences of single life. You know, some experiences are better than others – single people are heterogeneous. But no, there is no discussion of single people or single life whatsoever. Either we singles are all the same or we are not worth thinking about or – my best guess – it truly never occurred to the authors to take single life seriously.

If the authors were to take single life seriously, they might wonder why the single people did just as well as the coupled people on half of the measures. They might wonder how singles manage to do so well in the face of so much singlism and matrimania and all of the marital status discrimination that is written right into the law. Maybe they would think about what is good about single life and what is not so good about married life, instead of considering only what is potentially bad about single life and good about coupled life. (See, for example, Singled Out.)

The authors are so sure that getting married is good for health and well-being that they admit they were surprised to find that getting married did not result in any better physical health. They should not have been. As far back as 2005, in that special issue of Psychological Inquiry on singles in society and in science, Karen Rook and Laura Zettel wrote a brief review that burst the marriage-makes-you-healthy bubble. Still, the authors are not about to abandon their belief in the health benefits of marriage. Instead, they propose that “the possible physical health benefits of marriage may accrue over the life course.” See? They didn’t find any benefits but the benefits really are there, just waiting to emerge.

V Implications the Authors Draw from their Findings: Let’s Have More Government Spending on Promoting Marriage?

When the authors get to the part where they spell out the implications of their work, they say this:

“The past decade has witnessed legislation supporting marriage promotion initiatives…Our findings…are critical to the rationale behind such efforts that assume causation, not just correlation.”

I guess they are not explicitly saying that, based on their work, the government should spend more money promoting marriage, but they sure are not expressing any caution about the idea. These programs, of dubious effectiveness, are arguably ideologically motivated initiatives that redirect funding away from other programs that really do work in achieving goals such as the reduction of poverty.

Reference: Horn, E. E., Xu, Y., Beam, C. R., Turkheimer, E., & Emery, R. E. (2013). Accounting for the physical and mental health benefits of entry into marriage: A genetically informed study of selection and causation. Journal of Family Psychology, 27, 30-41.