There is another big problem, too, as I have been arguing since writing Singled Out and even before. The group of people who are currently married does not include all of the people who ever got married. Divorced and widowed people are separated out of the currently-married group. So if currently married people are happier than other people, you cannot say that if the unmarried would only get married, they would be happier, too. The divorced and widowed people did get married. If you want to understand the implications of getting married, their experiences have to be included.
The real kicker is that even when marriage is given the utterly unfair and methodologically indefensible advantage of a design in which only the currently married are compared to others, there is still very little difference in happiness, and sometimes the people who did get married and then divorced (or were widowed) are less happy than those who stayed single. The results from the nationally representative sample that I described in Singled Out, for example, were (on a 1 to 4 scale, with 4 indicated the greatest happiness): 3.3, currently married; 3.2, always-single; 2.9, divorced; 2.9, widowed.
Better Ways to Study the Implications of Marital Status for Happiness, Health, and Everything Else
If you really wanted to know, using the scientific gold standard, whether marrying makes people happier, you would have to randomly assign people to get married or stay single and see what happens. Of course, it is not possible to do that.
The next best thing is to study the same people over the course of their adult lives, and see how their happiness or satisfaction with life changes as they experience various life events. If you want to know the implications of getting married (or, say, getting divorced) for people’s happiness, then start asking them about their happiness or satisfaction before the event ever happened, and continue asking them (maybe once a year, though more often might be even better) how they feel long after the event occurred.
In the Journal of Personality and Social Psychology, a group of four authors published a statistical analysis and summary (a meta-analysis) of 18 such studies of people who got married and 8 of people who got divorced.
For one of the events, they found that people (on the average) felt a little worse just after the event occurred, then, over time, they reported feeling better and better every time they were asked.
For the other event, people may have felt a bit better right after the event than they had before, though it depended on the particular question you asked. Then, over time, they either felt no differently, or they reported feeling even worse. (Again, the particular question matters, though all of the questions have something to do with happiness or life satisfaction or satisfaction with a partner.)
So which of the results describes the implications of getting married and which describes the implications of getting divorced?
It was the people who got divorced who felt worse at first, but then felt better and better over time. The people who used to be single and then got married (well, some of the people who used to be single and then got married – more on that in a moment) felt either a little bit better at first (or their feelings/appraisals did not change or they got a bit worse), and then, over time, their feelings/appraisals either stayed about the same or got worse of time. (If you can access the paper, the relevant graphs are Figures 3 and 4.)
The authors realize that you could look at those timelines of well-being and suggest that: (1) getting divorced makes you happier over time; and (2) getting married does not make you happier and may even make you less happy.
They don’t like those interpretations. Taking the marriage findings first, they suggest that people were already becoming happier than usual before they married, in anticipation of the wedding. So when married people start reporting lower satisfaction after the marriage than they did before, they are just going back to the level of satisfaction they felt before a wedding was in the picture.
I don’t object to that interpretation. It is entirely possible. As the authors note, you would need to study satisfaction for enough years before the wedding to be more certain that this explanation is a good one.
For divorce, the thinking is similar. Levels of happiness were probably already heading down for people headed to divorce, and so getting divorced only makes people happier relative to how increasingly miserable they were feeling, year after year, when they were married. Again, I buy that as plausible, and related research suggests as much.
The 18 Long-Tem Studies of the Implications of Marrying: Some Specifics
The 18 key studies of the implications of marrying for well-being were all prospective studies. That means that people were asked about their happiness or satisfaction starting before they got married and continuing for a while afterwards. On the average, people started reporting their satisfaction about 4 months before they married, and continued doing so repeatedly. The average number of times they reported their satisfaction was about 5. Some of the research has been ongoing for more than a decade.
In at least 11 of the 18 studies, the people in the marriage group included only those who got married and stayed married all through the study. This is important. The cumulative results of the 18 studies don’t really tell us about the implications of getting married; instead, they tell us about the implications mostly only for those who get married and stay married. For those who marry and then divorce or become widowed, the implications may be very different.
The authors of the 18 studies asked about well-being in at least one of three different ways:
- Happiness. I’m calling this happiness, but the authors of the meta-analysis use the term “affective well-being.” The participants in the studies were sometimes asked about happiness and sometimes asked about unpleasant feelings such as a depressed mood (which is different from clinical depression).
- Life satisfaction. Participants are asked how satisfied they are with their lives. The authors called this “cognitive well-being.”
- Relationship satisfaction. Participants are asked how satisfied they are with their relationship with their partner.
The first question the authors of the meta-analysis answered was: How did the participants’ happiness or satisfaction change from just before they got married to just after? (Remember, “just before” was, on the average, 4 months before the wedding. Just after was the first time they were asked after the wedding.) The second question was: How did happiness or satisfaction change over time after the wedding?
Here’s what they found:
- For happiness, there was no difference in happiness from just before the wedding until just after. Over time, on the average, happiness did not change. Participants did not get either happier or less happy as the years of their marriage marched on.
- Satisfaction with life did increase from just before the wedding to just after. But then it decreased continually over time.
- Compared to life satisfaction, relationship satisfaction decreased from just before the wedding to just after. As time went on, relationship satisfaction continued to decrease at about the same rate as overall life satisfaction.
Here’s what did not happen: Except for that initial short-lived honeymoon effect for life satisfaction, getting married did not result in getting happier or more satisfied. In fact, for life satisfaction and relationship satisfaction, the trajectories over time headed in the less satisfied direction.
What is really remarkable about the combined findings of the 18 studies is that the designs were biased in favor of making marriage look good. At least 11 of the studies included only those people who got married and stayed married.
There was one sentence in the results section of the meta-analysis about how the results were different for those studies which included people who had separated, rather than tossing them out of the marriage group: “These samples did not differ in the initial reaction; however, the rate of adaptation was significantly less negative in samples without any separations.”
Translation: Negative adaptation means that people were getting less satisfied over time. If you take out the people who got separated and just look at the people who got married and stayed married, then the decrease in happiness is not as striking. That’s another way of saying what I’ve been saying all along: If you just look at the people who got married and stayed married, you are skimming off the top. You cannot generalize from just those people to offer blanket advice such as, Get married and you will be get happier (as Dan Buettner, author of The Blue Zones, actually did in the February/March 2013 issue of the AARP Magazine). Even the skimmed people did not get happier and stay happier.
After 18 Failures to Show that Getting Married Increases Happiness, They Are Still Insisting that It Does
Too many social scientists simply are not going to give up on the claim that getting married makes you happier. Harvard Magazine recently reported that Dan Gilbert, Harvard professor and bestselling author of Stumbling on Happiness, delighted an audience by asking them “how many believed getting married led to happiness” and then proclaiming “you’re right!” to the people who raised their hands.
There were no references in the magazine, but maybe Gilbert was referring to the latest attempt to salvage the myth of marital bliss. It was a study of people’s life satisfaction over time, similar to the ones I described above. (It probably was one of the 18 studies, though the specific studies were not listed in the article.)
In the same type of analyses conducted for the 18 studies, participants’ reports of their life satisfaction were tracked starting before they married and continuing for years afterwards. Only those who got married and stayed married throughout the study were included in the analyses.
The results were the same as for the 18 studies. Participants reported an increase in life satisfaction around the year of the wedding (compared to before the wedding), but, as the authors noted, “this effect was short-lived.” Over time, the participants went back to feeling as satisfied or as unsatisfied as they were with their lives before they got married.
So how did the authors find a way to make getting married look like a boon to happiness? First, they looked at normative changes in life satisfaction over the course of the adult years. Setting aside considerations of marital status, the study showed (as have other studies) that life satisfaction decreases over time. Then they looked specifically at the people who stayed single, and found that their life satisfaction showed some decrease over time.
(Some specifics: At the time of the marriage, those who got married and stayed married reported life satisfaction that was a half of one point, on a 7-point scale, higher than the matched single people. In the years afterwards, those who married and stayed married averaged .28 of one point on a 7-point scale greater life satisfaction than those who stayed single. About the “matching”: For each person who got married and stayed married, the authors tried to find a single person who was as similar as possible in age, sex, education and income. They didn’t say when they assessed income. The matching was not totally successful. For example, the single people, on the average, were four years older than those who got married and stayed married.)
Here’s what the authors said about their results: “…although our previous analyses showed that people were no more happier after marriage than before marriage, these results suggest that married people are indeed happier than they would have been if they did not get married.”
That interpretation was repeated in the press. The study may also be the basis for claims by people such as Dan Gilbert that getting married makes you happier.
Do You See What Is Wrong with the Study and the Claims about the Results? If So, You Are Way Ahead of Most Journalists, Professors, and Bestselling Authors
I think there are at least two major problems with the claims made about what this study suggests. The authors of the study acknowledge only one of them. There are also a number of alternative interpretations that go unmentioned.
See if you can figure out the flaws in the study and in the claims made about the results. I’ll tell you what I think they are in a future post. [UPDATE. Here it is: Every time you hear that getting married will make you happier, read this.]
I’m guessing that many of you can identify problems and generate alternative interpretations, even though you may have no training in research methods. If you can, you are doing better than the authors of the study, the reviewers and editors who critiqued the study before it was published, the journalists who wrote about it, and highly successful people such as Harvard professor Dan Gilbert and “happiness expert” Dan Buettner.
[Note: My latest elsewhere: Is the lean-in conversation going to leave out single women?]