And Statistics

One of the side effects of regularly listening to the 538 Politics Podcast, especially their "Good or Bad Use of Polling" segments, is that I've become better able to understand data when it's being presented in the media. Because few things are really what they're cracked up to be. Take the following story from Axios:

Married people are happier than single adults: survey

Brought to you by the wedding-industrial complex. Not really. It's actually brought to you by a conservative think-tank, that's looking "to strengthen marriage and family life." Because headlines need to be punchy, Axios somewhat misstates things.

Gallup, the organization that conducted the surveys, doesn't set out to directly measure happiness. What they're measuring are people's self-reported positioning on the "Cantril Self-Anchoring Scale," which asks respondents to place themselves on a 0-10 scale of worst to best possible lives for themselves 1) in the present and 2) in five years time. Using these scores, Gallup places people into one of three buckets: Thriving, Struggling or Suffering.

What the Institute for Family Studies is touting, is the fact that the Gallup data show a correlation between being married, and attaining a "Thriving" rating in the Gallup groupings. Their point, which is to be expected, given their mission is that "In America at least, it looks like well-being ebbs and flows with marriage." The column does, however make an important caveat: "Marital status is not randomly assigned." And this is important, because the article is at pains to point out "that marital status is a stronger predictor of well-being for American adults than education, race, age, and gender." While it notes that the data can't prove a causal relationship, the piece maintains that marriage is an overall good. Which makes sense for an organization that's looking to boost marriage.

Looking at the actual data, what stood out for me was 2017. With the exception of divorcées, this is the high-water mark for all groups' thriving percentages. And 2018 was a bit worse for all groups. 2020 was a disaster for everyone, and only the never married cohort continued their recovery into 2023. Divorcées were on track to take over the second-place spot before the pandemic derailed everything, and they've barely recovered any ground since 2020. All groups are much lower in the rankings as of last year than they were in 2017, and married couples are on a downward trend. The Institute for Family Studies article speaks to none of this. Which is unfortunate, because there is likely to be insight there.

Instead, it shifts, and focuses on the fact that "People living in metropolitan areas with higher rates of marriage enjoy higher subjective wellbeing." To demonstrate this, it take the metropolitan areas with the ten highest and ten lowest rates of marriage, and plots their Gallup "Thriving" rates, along with their rates of "deaths of despair," defined here as "deaths from suicide, drug or alcohol poisoning, or overdose." (As an aside, "deaths of despair" has become something of a buzzword, and like most buzzwords, it's not strictly accurate. After all, drug and alcohol poisoning/overdose deaths may also happen at parties.) But to do this, it relies on an analysis done for The Brookings Institution. That analysis shows something that the Institute for Family Studies alludes to; the deaths of despair rates for the top and bottom ten metro areas by "Thriving" rates. Provo-Orem, Utah holds the number one spot on both lists, having the highest percentage of married households and of people who say that they are thriving. But, notably, other than that, none of the top ten metro areas for "Thriving" show up on the top ten list for marriage rates, and there is no overlap at all in the lists for the bottom 10 metros. Brookings, rather than focusing on rates of marriage when evaluating whether people are likely to self-report as thriving, points to demographics and location, noting that different ethnic groups have different "Top Ten" lists. It doesn't mention marriage at all as a factor.

In the end, I found myself looking for a data point that wasn't there. Mainly because I don't believe that Gallup asks about it in their survey. Is there a correlation between whether people want to to be married or not, and whether they report themselves to be thriving? Policies designed to help people who want to be married become married would look significantly different than policies designed with marriage for everyone in mind.

Coverage of the benefits of marriage, especially when promoted by marriage boosters, tend to view the institution as something that's good for everyone, regardless of other concerns. "Married people are happier than single adults" and "Married People Are Living Their Best Lives" being just two examples. But "People who have the relationships they want are happier than other adults" and "People Who Have the Relationships They Want Are Living Their Best Lives" are just as likely to be true.