,
16 tweets,
7 min read
Read on Twitter
Last weekend, @profpauldolan spoke at the Hay festival in the UK, and some of his remarks were picked up by the BBC, the Guardian, and the Independent, and then repeated dozens of times in outlets across the world, including US reporting from FOX News to local TV stations. 1/ 
I’m no “happiness expert” and don’t have strong ideological feelings about whether everyone should be getting married or not, but I have done a ton of research with the American Time Use Survey (ATUS), which he said he based his statements on. And the claims felt weird to me. 2/
First of all, there’s this statement: that when a married woman’s spouse is not “in the room,” she’s “fucking miserable.” I know that this information isn’t included in the ATUS, so I reached out to him. He has since retracted this statement and will correct it in his book. 3/
The Guardian corrected that statement, but didn’t touch the others he made. So I got a copy of his book and looked at the evidence he presents to back these up. 4/
Obviously, he’s taking a bit of license in speaking about men “calming down” when they get married, but the evidence does support significant observed differences between married and unmarried men in terms of health, life satisfaction, and so on. 5/
I can't tell where the claim that women “die sooner” when they get married originates. In the book, he puts it this way, going from there being evidence of benefits to marriage to saying there “really do not appear to be any health-related reasons to marry if you are a woman.” 6/
The citation in that second paragraph crucially does not say that there are no benefits to women marrying, only that they are *not as large as benefits to men*. An older article he cited earlier claims that unmarried women have 50% higher mortality rates than married women. 7/
Next, the claim that “healthiest and happiest population subgroup are women who never married or had children.” The ATUS lacks data on *ever* having children, but I can compare never/ever married with and without children in the household. This doesn’t back up his claim. 8/
I don’t think that a mean of a 0-to-6 happiness scale is the most meaningful measure, but I use it to be consistent with the data underlying 9 figures in Prof. Dolan’s book. I can more generally look at happiness over the life course by gender and marital status, as I do here. 9/
My overarching observation is that across most categories, people seem to have about the same level of happiness. But ATUS data suggest:
❌ married women aren’t “fucking miserable”
❌ women who never married or had children aren’t the “happiest population subgroup”
10/
❌ married women aren’t “fucking miserable”
❌ women who never married or had children aren’t the “happiest population subgroup”
10/
And there is evidence that marriage is related to health benefits for women (though it’s by no means convincing), contradicting the claim that married women "die sooner." 11/
So there does not appear to be evidence supporting *any* of the dramatic claims in the press. While one has been retracted, I believe that all of them should be retracted and corrected. And I would be glad to walk @profpauldolan or any journalist through any of this. 12/12
Some more thoughts. First, this figure shows a five-year moving average happiness level for each category by age. One observation is that the lines for married men and women look similar, while unmarried men tend to be a little less happy (on average) than unmarried women. 13/
Marriage decisions are related to other factors, so I'm not sure how useful this comparison is. But also, I think marriage proponents (like @NickWolfinger and @WilcoxNMP) would agree that marriage has changed a lot over the past decades. Like these huge shifts in prevalence. 14/
Given such changes, we probably want to avoid comparing all unmarried women of all ages to all married women of all ages. But even if we deal properly with age, marriage is still endogenous. So are married people happier because they're married, or for some other reason? 15/
To wrap this up: I didn't explicitly state this above, but the problem with the "spouse in the room" statement was that it was based on misreading a key variable, as I explained here. In an email, Prof. Dolan said he believes I'm correct about this. 16/16
