Verified by 
"Married adults report better health, but singles are catching up," proclaimed one of the many headlines touting the latest marital status study to make it into the media spotlight.
I'll give you my bottom line about this study first. Then I'll explain in greater detail.
BOTTOM LINE
Here's what the study really did show:
1. For people who had always been single, their health improved continuously from 1972 until the end of the study, 2003. This improvement was evident in all groups - men and women, Blacks and whites. (This was accurately reported.)
2. The other part of the headline, that married people are still the healthiest, was misleading in at least four ways:
A. The study does NOT show that differences between the currently-married and the always-single occurred BECAUSE the currently-married people got married. Instead, the data were analyzed in a way that gave the currently-married people an advantage it did not give the single people. (That is standard practice in cottage industry of marriage studies.)
B. Even allowing for the approach that makes marriage look better than it really is, the differences in health between the currently-married and the always-single are tiny. By the last year of the study (2003), the probability of reporting good health was about .928 for the currently married and about .926 for the always-single. If it is easier to think in terms of percentages, that means that about 92.8% of currently-married people said their health was good or excellent, compared to about 92.6% of always-single people. So the headlines saying that married people are still the healthiest are touting a difference of about two-tenths of one percent. And remember, that advantage did NOT come from getting married (as I explain below). Is that 0.2% what you imagined when you read headlines claiming that married people are the healthiest?
C. For the African-Americans, by 2003, there was no difference at all between the health of the currently-married and the always-single.
D. Women who had always been single were healthier than men who were currently married. In 2003, the likelihood that always-single women would report good or excellent health was about 92.8. For currently-married men, it was about 91.8. (This wasn't noted in any of the media reports I found.)
THE MORE DETAILED DISCUSSION
The authors crunched data collected over 32 years (from 1972 through 2003) from more than a million people, so this is a study worth taking seriously. Importantly, the same people were NOT followed that whole time. Each year, a different set of people participated. They described their overall perceptions of their own health, on a scale ranging from poor to excellent. Those reports were then linked to their marital status.
The key question was: How has the health of the different groups changed over the 32 years? So, for example, if you compared the health of the currently-married participants in 1972 to the health of the currently-married participants in 1973 and every other year up to the final one (2003), what would the change over time look like?
For the people who had always been single (or "never-married," as they are more often called), the answer was clear. Whether male or female, Black or white, their health steadily improved over time. The same could not be said for any of the other marital status groups (currently married, divorced, separated, or widowed).
Sounds good for singles, except when you keep in mind the headline. Sure, singles are doing better now than they were a few years ago, but they are still not as good as married people. The Washington Post headline was even more stark: "Married Folks Still the Healthiest." Singles did not even get the Most Improved Player Award in that story title.
The currently married people actually were the healthiest in most (though not all) of the analyses, so what's my problem? (You just know that I have one.) It is the implication that they are the healthiest because they got married (e.g., "marriage benefits health"). Again, sadly, the Washington Post was one of the worst offenders, recruiting an "expert" not involved in the study to comment, "This study provides confirmation that marriage does tend to make people healthier." Of course, it does nothing of the sort.
In my last post, I reprinted a section from the "Science and the Single Person" chapter of my book, Singled Out, to explain why studies like these do not and cannot demonstrate that getting married makes people healthier (or happier or anything else). I'll do a quick recap here, but you may want to go back to that post (or to the book) for more details.
Here's the hypothetical example I like to use. Suppose a drug company did a study in which they let people decide for themselves whether to take the new drug (rather than randomly assigning them to a drug condition or a placebo condition). They also let people quit taking the drug whenever they wanted to. Some people started taking the drug and hated it, so they were removed from the Drug group and set aside into a different group (No Drug - intolerable). Others started taking the drug but eventually lost access to it (No Drug - withdrawn). The drug company removed them from the key Drug condition and set them aside, too.
So now the drug company takes the data from only those people who started on the drug and stayed on it. It compares how good those people felt to how good everyone else felt - the people who never did take the drug, those who hated it and stopped taking it, and those who were cut off from the drug. Their conclusion? Our drug works! Yeah!!! Everyone should take our drug and then they will feel so much better.
That's the logic of all of these studies of marital status that are not longitudinal (i.e., that do not follow the same people over a number of years as they get married, get unmarried, or stay single). You can claim that getting married is good for you, as long as you do not count the people who got married and found it not so good at all.
A previous study further illustrates how currently married people can end up with the best health scores even though getting married did NOT make them healthier. This is a study that was longitudinal: 10,000 people were followed for more than 4 years. They found that the married people who had more health problems early in the study were nearly twice as likely to divorce by the end of the study. See how this works? The married people who have health problems are more likely to get divorced. So now they are taken out of the married group and put into the unmarried group. The people who are left in the married group now have fewer health problems than the others. But that's not because they got married. Getting married did not cause them to get healthier. Getting unhealthy seemed to motivate some of the married people to get divorced.
I'm not saying that getting married never results in getting healthy. I am saying that there are lots of ways to account for results from the recent study and others like it. Typically, in our matrimaniacal culture, interpretations that make married people look better are favored over others, even when there is no good scientific reason to do so.
The marriage-centered point of view is evident not just in the framing of the results, but in the speculations that are offered to account for them. Let me illustrate with another finding. Over time, the health of people in the divorced, separated, and especially the widowed group got even worse. So, from 1972 to 2003, the health of people who had always been single looks better and better, while the health of the previously married looks worse. By the last year of the study (2003), the people who had always been single had a health advantage over the previously married that was greater than it had ever been before.
Why are the always-single people healthier than the previously-married? The authors one possible explanation that is called a "stress" or "crisis" or "loss" hypothesis: "the never-married are relatively immune to any apparent disadvantage associated with the stress of marital dissolution." Translation: People who have always been single have not dealt with a marital relationship that has ended in death or divorce, so of course they are better off.
Do you see any problem with that? Can you think of any other explanation for why the always-single might fare better than the previously-married? I think this issue is so important that I am going to highlight it in a future post. So for now, think about your answers, and I'll get back to you.
The question raised by the recent findings is not just why the always-single are healthier than the previously-married, but why the difference between them has grown larger between 1972 and 2003. Why would the dissolution of a marriage be harder on people in recent years than in decades past? The authors say more research is needed, and they are right. Here's one possible hypothesis: People who got married or unmarried in 1972 were not so caught up in the Soul Mate mentality that became so prevalent later. They did not expect their spouse to be their everything - to fulfill all of their wishes and hopes and dreams. They continued to value other important people, such as friends and relatives. People from more recent times who have bought into the myth of the Soul Mate and The One, and who expect the world of that one person, and who have relegated everyone else to the back burner, are going to be crushed when that one person is gone.
There are a few other points from the study that I'd like to clarify or underscore.
1. What is the magnitude of the difference between the currently-married and the always-single? If you can access the original article (unfortunately, it is behind a pay wall), look at the graphs on the bottom of p. 247. The BIGGEST difference was for men in 1972. For the always-single men, the likelihood that they would report good or excellent health was about 89%. For the currently married men, it was about 91%. For the women, the difference was smaller than that (a few tenths of 1%) for all 32 years. (How can such a small difference be statistically meaningful? One reason is that there are more than a million people in the dataset.)
2. You may have read that "the gap between the married and the never-married is closing, especially for men." I want to clarify the "especially for men" part. Again, if you can, look at the graphs on the bottom of p. 247. The gap is closing more for men in part because it started out bigger. So in 1972, there is a bigger difference between the married men and the always-single men than there is in 2003. For women, there is a very small difference between the currently-married and the always-single at every point in time. Looking at the last year on the graphs (2003), the difference in health between the currently-married and the always-single appears to be about the same for the men as for the women.
3. Women who had always been single were healthier than men who had always been single for all 32 years. (Not a big difference.)
4. Women who had always been single were healthier than men who were currently-married; from the graphs, it appears that this was true for all 32 years as well. (Again, none of these differences are big.)
5. At the start of the study (in 1972), if you control for family income, all of the categories of unmarried people (always-single, divorced, separated, and widowed) were healthier than the currently-married. That suggests that if the currently-married people were healthier than the currently-unmarried people, it may have been because the married people had more money. If the unmarried had as much money as the married people did, they would probably have been healthier than the married people. (No, that wasn't reported anywhere either.)
The ways of presenting studies like this one are not unique to these authors. The kinds of explanations they entertain, such as the loss hypothesis, are the standard ones. The authors are, in fact, better than most. One of the points they were making in the article is that it is important to look separately at different categories of unmarried people, rather than glomming them all together. That's commendable.
The authors also end their paper in a way that is rare for scholars of marriage:
"the self-rated health status of the never-married has improved for all race and gender groups examined, and it is more similar to the married for men now than ever before, which suggests that encouraging marriage in order to promote health may be misguided. In fact, getting married increases one's risk for eventual marital dissolution, and marital dissolution seems to be worse for self-rated health now than at any point in the past three decades."
[ANSWER to the question from the last post: At the end of the last post, I predicted that you would already be able to figure out what was wrong with the first sentence of one of the press releases about the study. The sentence was: "For years, researchers have known that adults who have swapped rings say they are healthier than their never-married peers are." Of course, that's not true. People who are divorced and widowed have "swapped rings," but they are typically less healthy than people who have always been single.]
