What's the Recipe for Loneliness?

Is it lack of religion? Little education? A low income?

I am going to be in Kansas City for a speaking engagement in late February. My schedule affords me the opportunity to speak somewhere else on the night of Thursday, February 27th. If you have a group that would like to host me, I can offer you a really good rate. And you won’t have to pay for travel.

If you are interested, just use this Speaker Request Form.

---

In 2023, the Surgeon General of the United States issued a report with the title, “Our Epidemic of Loneliness and Isolation.” (PDF). Dr. Vivek H. Murthy wrote in the introduction that “Loneliness is far more than just a bad feeling—it harms both individual and societal health.”

There’s a nice graph on page 13 of that report that is worth emphasizing here. Between 2003 and 2020:

• Social isolation increased 24 hours a month for the average person.

• Social engagement with friends dropped by 20 hours.

• Companionship dropped by 14 hours per month.

That’s a trend that is worrying, of course. And that's what I wanted to explore in this post. I wanted to really try and sort through what factors are leading to less socialization. Let me start by highlighting two of my favorite bugaboos: education and religious attendance.

My data is the Household Pulse Survey, which is fielded by the United States Census Bureau. This data was collected about six months ago, so this is about as good as it gets.

This is the relationship between the highest level of education obtained and self-reported annual religious attendance.

I like this graph because it helps us navigate the overall relationship between these two variables. It’s fair to say that there’s a weak-ish connection between the two. Among folks who earned no more than a high school diploma, the share who never attend church is about 58%, while those who attend once a month is much lower at 22%. To me, there’s a clear line of demarcation in the graph between those who attended college but didn’t earn a degree and those who obtained an associates degree.

There’s some evidence here that college educated folks are more likely to be attending religious services than those with lower levels of education. Among people surveyed who earned a graduate degree, 29% indicate that they attend religious services at least once a month. That’s seven points higher than those high school graduates.

I have written about this phenomenon so much but - there’s a generally positive relationship between education and religious attendance. In the linked post, I dig through a bunch of datasets to get you to this conclusion - there is absolutely no evidence that increased education leads to decreased religious attendance. At worst, it’s a flat relationship. In many cases, the line points upward: increased education leads to increased religious engagement.

But let me pivot from education for just a bit. I will return to it, don’t worry. I wanted to poke around the relationship between one type of social activity (religious attendance) and other types like talking on the phone and socializing with friends/family.

Well, this is a pretty strong relationship. Among people who chat on the phone less than once a week, about three quarters of them say they never attend church services. As phone usage climbs, so does religious attendance in a fairly linear fashion. Among those who say that they chat on the phone with friends or family at least five times a week, just 45% report never attending religious services, while about a third are at church on a monthly basis.

Clearly one type of social activity is deeply related to another in this case. But, you folks who are paying attention carefully will quickly point out that this may actually be a bad measure because old folks are more likely to chat on the phone compared to younger folks. And we know that religion is much stronger among Boomers than it is among Generation Z. So maybe this is just masking an age effect.

So, let’s switch to a question about socially gathering with friends or family.

Okay, I think it’s fair to say that the same general pattern from the prior graph shows up here, too. Among folks who rarely hang out with other people, they are also a lot less likely to be going to a house of worship this weekend. Among those who socialize less than once a week, about two-thirds never attend religious services. You see the same staircase pattern as the previous graph, too. As socializing increases so does religious attendance. Among the most social folks in this graph, they are nearly as likely to be monthly attenders of church compared to never attenders (36% vs 43%). That gap for the non-social group is 49 points, for comparison.

So, I think that there’s a pretty compelling case that one type of socialization is deeply intertwined with other types. But let me bring education back into the fold now. I wanted to figure out if those with a high level of education were more social on these dimensions.

I already discussed the pattern of education and religious attendance above, but I wanted to replicate it here. Those who are the least likely to be never attenders are those with graduate degrees. While folks who stopped their education after high school had the highest share of never attenders.

But what about those other two metrics related to talking on the phone and socializing with folks? I think that middle graph - socializing in person should be a huge alarm bell for anyone interested in the social fabric of America. Among those with a high school diploma, 44% said they socialized less than once per week. Among those with a graduate degree it was only 32%. And you can see the same general pattern with talking on the phone, too. Although the shift across the educational spectrum is lower at just five points.

But here’s a theory that could make all this moot - many people who don’t socialize also report that they aren’t lonely. So if they aren’t feeling bad about their level of socialization then this is much ado about nothing. Luckily, the survey asked how often the respondent felt lonely.

To me, this is where the rubber meets the road. The people who are the most likely to report feeling lonely “usually” or “always” are those at the bottom end of the educational spectrum. Among folks who didn’t finish high school, 20% chose one of these two options. For people with graduate degrees only 7% were usually or always lonely.

Consider this - among high school graduates, 55% say that they never or rarely felt lonely. Among those with an advanced degree that figure was a whopping 67%. Looking at the last couple of graphs in totality, a pretty terrifying picture appears. Folks with low education don’t socialize that much. They also say that they are lonely at much higher rates than people with college degrees.

However, I had a nagging question in mind - is it really education that makes people less lonely? We know that education is correlated with a bunch of other factors like income. So maybe education is really just masking the income variable when it comes to loneliness. Let’s sort that out in this last graph.

Well, now I am convinced that this isn’t about education that much at all. It’s really about money. It’s as clear as can be - as income increases the likelihood of feeling never or rarely lonely goes up dramatically. Among high school graduates who make less than $25,000 per year only 40% are never/rarely lonely. Among high school grads who make at least $200,000 per year, it’s 74%. That same pattern is there among all levels of education - increased income leads to decreased loneliness.

And what is really stunning is how little education matters when you control for income. Look at the far right bar for each level of education, that’s folks who make at least $200K per year. That group is easily the least lonely of any other income level, regardless of education. Rich people are much less lonely than poor people. That’s the only conclusion that makes any sense.

Listen, all these relationships are a mess to untangle from a methodological perspective. Education leads to higher income. Higher incomes leads to less loneliness. But education also leads to more socialization leading to less loneliness. You see what I mean?

As I’ve written before - dropping out begets dropping out. We should be especially concerned with those folks at the bottom end of the socioeconomic spectrum. They seem to be falling through the cracks of civilized society, which will make it harder for them to move their way up the economic ladder. And if I’ve learned anything in doing all of this work it’s the poisonous nature of despair. When people feel hopeless and helpless, we all need to start worrying.

Code for this post can be found here.