The Loneliness Epidemic

Man is by nature a social animal... (Aristotle, Politics, c. 350 B.C.)

Ah, look at all the lonely people... (The Beatles, Elanor Rigby, 1966.)

This newsletter was prompted by a factoid I overheard on someone else's car radio.

I've encountered statistics in a variety of settings, ranging from research reports to the backs of granola packages, but this is the first time I ever heard one while stopped at a red light.

The Kia next to my car had its windows down and the volume up, and in the midst of a rant about social media, the voice on the radio suddenly announced "1 in 4 people are lonely."

That was the factoid. Then the light turned green, and I was left wondering, once again, where does stuff like this come from?

Google quickly answered my question. CNN ran a story last week entitled "The loneliness epidemic: Nearly 1 in 4 adults feel lonely, new survey finds". The new survey had been conducted by Meta, in conjunction with Gallup, and published on October 24. Of course I had to look.

The "survey" turned out to consist of a single question: In general, how lonely do you feel? The answer options were very lonely, fairly lonely, a little lonely, not at all lonely.

Meta-Gallup posed this question to roughly a thousand adults in each of 142 countries. 24% of respondents described themselves as very or fairly lonely (hence CNN's "1 in 4 adults" reference). 49% of people reported being not lonely at all. Loneliness was slightly more prevalent among younger people as compared to older ones, and among women as compared to men.

24% of a 100,000+ sample is a lot of lonely people. Where do they all come from? To answer the Beatles' question, everywhere. According to a fuller report of the data released by Meta-Gallup yesterday, the percentages range from 5% (Vietnam) to 58% (Lesotho). The U.S. did fairly well, at 15%, but that's no cause for celebration, because it's still 1 in 6 people, and in other studies (using different measures) the percentages of lonely Americans are higher.

In spite of Meta-Gallup's large sample, I didn't feel particularly confident about data obtained via one rather leading question. I decided to take look at more substantial research. My questions were straightforward: Is there a loneliness epidemic? If so, how are people affected?

(In the Appendix I provide a definition of loneliness. Next week I'll discuss some of the most and least effective strategies that individuals as well as institutions can use to deal with this perennial social problem.)

A remark on loneliness statistics

When I first encountered the Meta-Gallup data, it struck me as vaguely ironic that researchers had been asking people all over the world how lonely they feel.

Meta's involvement adds a layer of irony, given that it owns Facebook, Instagram, and WhatsApp, three of the four most widely used social media platforms and, according to some experts, major contributors to the loneliness epidemic.

In spite of its limitations, the Meta-Gallup survey illustrates how the very technologies that connect people in historically unprecedented ways also allow us to know each other in historically unprecedented ways. How lonely were people in the 18th century? We'll never know. Back then we had no way to reach out to 100,000 people and ask.

Nowadays, the studies that explore loneliness rely on survey statistics that were developed in the early 20th century as a means of leveraging a new communication technology (telephones) as well as improvements to an existing one (the postal service). The statistics are more powerful now thanks to advances in computer technology, and they leverage a newer communication technology (the internet), but the underlying concepts are basically the same.

So too is the motivation behind the statistics. Simply put, we want to understand each other. In a mass society, statistics are indispensable to an understanding of the mass. Since we can't know the details of every individual's experiences, the hope is that by sacrificing depth for breadth, the stats can tell us a little something.

Is the U.S. experiencing a loneliness epidemic?

The term "epidemic" is often used for problems that weren't traditionally viewed as diseases. For instance, America is said to be experiencing an "epidemic of hate" (Kamala Harris), an "epidemic of fatherlessness" (Vivek Ramaswamy), and an "epidemic of narcissism" (Jean Twenge).

Calling these problems "epidemics" is a way of emphasizing their severity. The speaker may or may not have intended the term metaphorically, but it's illuminating to check how well it holds up as a metaphor.

Scientifically, the term "epidemic" is used when a disease spreads unusually rapidly among many individuals in a population. The disease may or may not be contagious. So, with this definition in mind, let's see whether the data on loneliness in the U.S. points to an epidemic…

Has loneliness been increasing?

This May, the U.S. Surgeon General Vivek Murthy released an advisory report, "Our Epidemic of Loneliness and Isolation" which shows that over the past two decades, Americans have become more socially isolated, in the sense of spending more time alone, maintaining fewer friendships, and so on. However, being alone is not the same as feeling lonely. Most of the data in this advisory only hints at that loneliness has increased.

It's hard to get a read on historical shifts in loneliness. Different studies have measured it in different ways and focused on different time periods. Particularly misleading are media reports on studies that only examine a one- or two-year time period, or discuss the impact of the pandemic.

The most credible data I've seen comes from a 2022 meta-analysis published in the prestigious review journal Psychological Bulletin. This study, led by Dr. Susanne Bücker at Witten/Herdecke University in Germany, looked at shifts in loneliness among 18 through 29-year-olds from 1976 through 2019.

Methodologically, this meta-analysis has many strengths, including a massive sample,(345 studies; 124,855 participants), reliance on a single measure of loneliness (the UCLA Loneliness Scale), and a careful approach to data analysis (e.g., efforts to counter publication bias).

One of the key findings was that from 1976 through 2019, loneliness in the U.S. increased, on average. The increase mainly occurred after 2000, and it was linear, meaning that loneliness each year tended to be slightly higher than in previous years. Comparable shifts were not observed in European and Asian samples.

How much more lonely have Americans gotten?

That's a tricky question, given the complexity of the statistics, but we can get some sense of the shift by considering how loneliness was measured.

Informally speaking, the UCLA Loneliness Scale is a pretty grim read. The original scale consists of 20 statements that reflect the worst aspects of loneliness (e.g., "I have nobody to talk to", "I feel isolated from others", "I am unhappy doing so many things alone"). Respondents are asked about each statement whether they never, rarely, sometimes, or often feel this way. These responses are coded as 0, 1, 2, or 3, respectively. Thus, one could score anywhere from 0 to 60 on the original scale.

Setting aside some technical details, here's a crude reconstruction of the findings: From 1976 through 2019, average scores increased by about 3 to 4 points on a 60-point scale, ending up at an average of slightly over 20.

In short, the increase was small but statistically significant – and, in my opinion, meaningful. A 3-point difference on this scale could mean, for instance, a difference between experiencing three of the items "often" rather than "sometimes".

Think of what this would mean for even just one item. This is the difference between sometimes having nobody to talk to versus often having nobody to talk to.

The meta-analysis doesn't tell us about people over 30. (Meta-Gallup's survey suggests that loneliness is less pronounced among older adults, a pattern that emerges in other studies as well, although loneliness increases again in one's 70s and 80s.) In addition, the meta-analysis doesn't tell us what's responsible for the mean differences.

Mean changes are consistent with many scenarios. For instance, it's possible that many young adults are slightly more lonely than those who came before them. Or it could be that just a few young adults are a lot more lonely. I reached out to the lead author, Dr. Bücker, for her views on this issue. Here's what she had to say:

"I would rather assume that on average, emerging adults from more recently born cohorts report slightly stronger feelings of loneliness than cohorts born earlier. We have found no evidence that these slightly rising trends over historical time are driven by smaller subgroups reporting greater increases in loneliness. In my opinion, the former interpretation is also supported by the fact that we have included a variety of nationally representative datasets, which are not expected to over-represent certain populations that may have a higher risk of (stronger increases in) loneliness."

In her reply, Dr. Bücker also commented on the use of the epidemic metaphor:

"The increase in loneliness (notably before the COVID-19 pandemic) is overall rather small and does not really justify the term "loneliness epidemic" from my point of view. In other age groups (e.g., cohort comparisons of older adults), one finds rather stable or even partially declining trends."

In other words, loneliness is increasing, particularly among young adults, but not to the extent that it should be thought of as an epidemic.

What's causing the increase in loneliness?

Everyone's favorite scapegoat is related in some way to digital technology. The internet. Smartphones. Social media platforms. Remote work. Even video games and online porn.

Here's a datapoint: In Bücker and colleagues' meta-analysis, loneliness levels in the U.S. did not increase significantly after 2012.

The time period 2012-2019 was singled out for analysis because 2012 is the year smartphones reached market saturation, and some researchers expected loneliness would subsequently increase. According to the data, it didn't. (Other researchers disagree, though many of them only examine a one- or two-year period.)

I don't find these debates especially helpful, because loneliness stems from many sources, and linking society-wide levels to the emergence of particular technologies is close to impossible. For one thing, American society has experienced a variety of changes since 1976 (or since 2012, or any other year you choose), and so any broad shift in loneliness could be attributed to a number of social changes.

Meanwhile, smartphones, social media, etc. do help some people feel more connected, at least some of the time. Although researchers debate whether or not these technologies have a net society-wide impact on loneliness, they all agree that individual differences are critical.

In sum, although we can say with some confidence that loneliness is increasing in the U.S., I'm not so confident we can describe the relative impact of various causes. Yes, social media exacerbates the problem, at least among some people, but the overall extent of influence is unclear.

Is loneliness contagious?

To echo Dr. Bücker, the "loneliness epidemic" metaphor doesn't work well, given the relatively slow rate at which loneliness has been increasing, as well as the fact that it may not be increasing among some age groups.

All the same, references to an "epidemic" highlight the widespread prevalence of the problem.

In addition, this metaphor captures the finding that loneliness can be contagious. A classic 2009 study led by John Cacioppo showed that spending time with lonely people tends to increase one's own sense of loneliness. Cacioppo and colleagues' data suggests that it's not that lonely people tend to hang out together, or that lonely people tend to live in environments that promote loneliness, but rather that a process of emotional contagion causes one person to "catch" loneliness from another one. This is a fascinating study that I'll discuss more next week.

How does loneliness affect people?

The epidemic metaphor also reflects the observation that many physical and mental health effects can be traced to extended periods of loneliness.

For more details on this, check out the Surgeon General's May advisory, or the website of the Campaign to End Loneliness, as both offer succinct but comprehensive reviews of the negative impacts of loneliness on cognitive and emotional functioning, cardiovascular disease, self-harm behaviors, and so on.

An important observation about these studies – one that experts have been making for at least two decades – is that loneliness undermines our health independently of, and more strongly than, social isolation does.

In other words, what affects peoples' health is determined more by how lonely they feel than by how much time they spend alone.

Is being lonely like smoking 15 cigarettes per day?

An often-quoted finding from the Surgeon General's report is that lack of social connectedness decreases one's lifespan as much as smoking 15 cigarettes per day.

This finding gets a lot of attention because it's dramatic, and because the Surgeon General himself keeps repeating it (e.g., in his two-page introduction to the report).

I have great admiration for the Surgeon General and for most of this report, and I'm persuaded that loneliness is unhealthy, but I wouldn't ride that 15-cigarette-per-day statistic too heavily. For one thing, the meta-analysis from which it was drawn relied on studies that didn't sufficiently control for all relevant variables, including social isolation. What may hurt you as badly as 15 cigarettes per day is not loneliness per se, but rather, for example, a degree of social anxiety that causes both loneliness and an aversion to medical professionals.

Why does loneliness undermine health?

We know that loneliness feels bad, but why would it damage our bodies and brains?

In a word, loneliness is stressful, and prolonged stress undermines our health. For instance, the hormonal changes arising from stress damage the cardiovascular system, increase inflammation, reduce immune system functioning, and disrupt sleep. This seems to explain why loneliness has been linked to an increased risk of problems ranging from heart attacks to dementia to colds. (The data for colds is interesting, because it shows that people who spend more time in face-to-face interactions with others are actually less likely to get sick, in spite of the greater exposure.)

Conclusion

Loneliness in the U.S. has been steadily increasing for decades, and it should be treated as a major social problem.

You can call it an epidemic, if you wish to emphasize that it's experienced by many Americans, contagious, and detrimental to mental and physical health.

However, increases in loneliness since in the 1970s are fairly small and mainly observed among young adults. In this respect you might say we're not experiencing an epidemic, but simply a problem that is continuing to increase, particularly among younger people.

Next week I'll address the issue of how we can deal with the loneliness epidemic (or, if you prefer, the "loneliness epidemic"). Some of the strategies may surprise you.

Thanks for reading!

---

Appendix: A definition of loneliness

Loneliness is the unpleasant feeling that one's social interactions are lacking in quantity and/or quality.

Researchers sometimes describe loneliness as a perceived mismatch between desired and actual social connectedness. This makes sense, as long as we add in the part about its unpleasantness. Humans are social animals, as Aristotle noted, and we've evolved a dependence on connections with others, in much the same way we depend on sleep or food. As Susan Pinker puts it in The Village Effect, we hunger for these connections.

Researchers distinguish between social isolation (being alone) and loneliness (feeling lonely). These constructs are related – when people are alone, they sometimes feel lonely – but they're not identical.

Social isolation is a neutral term that refers to the objective state of being alone. (If you're in this state and feel content, then you're experiencing solitude.)

Loneliness, in contrast, is a subjective state that can be experienced in any setting. You can feel lonely when you're alone or in the presence of others – which for some people feels even worse. As Louise Glück wrote,

It is terrible to be alone.

I don't mean to live alone –

to be alone, where no one hears you.

Robin Williams expressed the same idea, more prosaically, in the 2009 film World's Greatest Dad:

"I used to think the worst thing in life was to end up all alone. It’s not. The worst thing in life is to end up with people who make you feel all alone."

Not everyone gets the distinction: For instance, Washington DC has been described as the nation's "loneliest city", because the data show that a greater percentage of residents live alone (48.2%) than in other cities of at least 150,000 people. But this isn't loneliness data. It's simple data on whether people live with others. It hints at a greater degree of social isolation in DC, but we can't even be sure of that, because the DC residents who live alone may spend a lot of time interacting with people at work and socializing after hours.