Is Marriage Toxic to Women? No, Misleading Reporting Is

Don’t count on getting married as a cure for depression

Posted Apr 13, 2010

A post by PT blogger Diana Kirschner that shot to the top of the most-read list was titled, "Is marriage toxic to women? Part II. Massive Meta-Analysis Says Marriage Reduces Depression." I didn't have time at the moment to go through all of her claims and describe what the original scientific reports really did say. So in the meantime, I wrote the post, "Really? Marriage Reduces Depression?" and asked Living Single readers to take a look at the evidence and see if they could tell whether Kirschner's claim was justified. A number of smart comments were posted; take a look at them here.

Now, I've finally made some time to look into the research myself, so I'll tell you what I found. I better hurry up, though, as Kirschner has already written the next several parts to her series. As a preview, consider this opening paragraph of her Part 3:

"Before we launch into Part Three I want to address the concern that single readers have expressed. That somehow I am condemning them to Hell if they don't get married. No! No! And No! First, there has been a global sea change where now more people are living together than are married. You'll want to check out my series on living together. Plus, some studies have shown that singles with strong social support are basically as happy as married women in good marriages. So relax. Marriage isn't for everyone and being single is way better in a lot of ways than being in a terrible marriage." (The original includes a footnote and links.)

I like one of the sentences in this paragraph, and you can probably guess which one. I also bet that Living Single readers can also see how Kirschner is in other ways practicing singlism, though no doubt unwittingly. She is very respectful of the people who comment, so I don't question her intentions. But much of this paragraph (and her series) is a great case study in how singlism gets perpetrated even by people who do not at all mean to be prejudiced against single people.

How My Approach Differs from Kirschner's

Kirschner draws primarily from a review of studies of marital status. The strength of a review paper is that it is based on multiple studies. Apparently, though, Kirschner either did not read the original studies, or did not think there was anything wrong with the way they were summarized in the review paper. I went to the journals to read the original research articles, and also checked for other relevant papers not included in the review.

A few points of clarification:

• In the title of her post, Kirschner describes the review as a meta-analysis, but it isn't. A meta-analysis is a quantitative review - typically an exhaustive one, and the paper she draws from is not.

• In a footnote, Kirschner says there were 70 studies covered in the review paper. The section on depression, though (which is the one relevant to her post) reviews seven studies. Four are analyses of the same dataset, and two are totally irrelevant to the question of whether people who marry become less depressed. (They are studies of divorce. The implications of getting unmarried are entirely different from the implications of getting married.) The seventh study was by Horwitz, White, and Howell-White. (Full references are in the review paper.) Bottom line: The relevant section of this supposedly "massive" review was based on two unique datasets.

Kirschner Claim #1

Kirschner: "In studies of those continuously married versus, those single or living together, researchers found that entry into marriage significantly reduces depressive symptoms in women (and men)."

My analysis: Kirschner's statement confuses two different findings. Studies of the continuously married do not tell us about the implications of entry into marriage. In analyses comparing the continuously married to one or more groups of continuously unmarried, no one entered into marriage during the relevant time period (about 5 years in the study in question). Entry into marriage is examined by following people as they transition from single to married.

I think she is referring to the first bullet point on the bottom of p. 30 of the review paper. The comparison is actually between those who got married during the time interval in question, and those who either (a) were already married and stayed that way, or (b) were unmarried and stayed that way.

Here's what the research shows.

Horwitz, White, and Howell-White compared those who got married and stayed that way to those who stayed single the whole time. They found that the men who got married did become less depressed but the women did not. This journal article includes one of my favorite lines of all time. Explaining why they did not include in the got-married group the people who married but then got divorced, the authors said:

"We do not include this group because clearly they are not deriving any benefits from marriage."

Pause for a moment to absorb that statement. In a study conducted to learn about the implications of getting married for depression, the authors deliberately set aside the men and women who got married but clearly were not deriving any benefits from marriage. The study was published in what is perhaps the premier journal for marriage research, the Journal of Marriage and Family. Even with that advantage given to the got-married group, the women in that group who got married still enjoyed no decrease in depression.

The four studies of one dataset were also marred by the same flaw: The only people included in the group who got married were those who got married and stayed married. Another more recent study of the same data was not included in the review. I described it in great detail in this post. Here I'll just share my conclusion.

If you get married, you may end up less depressed:

• if you start out among the 20% most depressed people to begin with,
• if you don't get divorced,
• if you end up in a marriage that is happier than most,
• if no one asks how you feel after the first few years, and
• no one compares the marital relationship to any other relationship that offers companionship and emotional support.

Compare that to the sweeping subtitle of the Kirschner post: "Massive meta-analysis says marriage reduces depression."

Kirschner Claim #2

Kirschner: "Those who were single and stable reported increases in depression over a five-year period while the married women did not."

My analysis: I think Kirschner is referring to the comparison of the continuously married with the continuously unmarried. (The second bullet point on p. 31 of the review paper.) The problem again (as Alan pointed out in the comments section) is that the married group does not include all of the people who got married, but only those who got married and stayed married the whole time. The divorced are set aside. Also, as usual, the unmarried group is not treated in a parallel fashion. All people who stayed unmarried are included, whether they wanted to be single or not.

With those unfair practices in mind, let's see what the results really did show.

Horwitz, White, and Howell-White found that the levels of depression dropped among the people who stayed single during the entire study - just the opposite of what Kirschner claimed about that group.

The studies based on the same dataset showed that any increase in depression over the course of the study was greater for those who were divorced the whole time than for those who stayed single. The analyses reported by Marks and Lambert showed an increase in depression for the men who stayed single, but not for the women. Typically, for depression and other heath variables as well, the risk lies not in staying single, but in getting married and then unmarried.

Kirschner makes two more claims in her post, and they, too, are misrepresentations of what the research literature really does show. You can probably already figure out why from what I've said so far here, or from other posts to this blog, or from my Single with Attitude book.

Living Single readers, please spread the word about what the scientific research really does show. Depression matters, and our understanding of it should be based on an accurate representation of the science.