Are Individualistic Cultures Lonelier? The Question May Be Wrong

There is a tidy story about loneliness that almost everyone finds intuitive. Western societies prize independence. People move away for work, live alone, and treat self-reliance as a virtue. Collectivist societies keep families together and obligations strong. So loneliness should be a disease of individualism.

The large cross-national datasets broadly support this. They also contain a result that makes the story much harder to tell.

In 2023, a team led by Julie Aitken Schermer published an analysis in Current Research in Behavioral Sciences that split cultural orientation along a second axis, one that gets far less attention. The finding was that both individualism and collectivism predicted more loneliness in one form, and both predicted less loneliness in another.

The thing separating the two forms was not independence. It was hierarchy.

What the big surveys found first

The most widely cited cross-national figure comes from the BBC Loneliness Experiment, analysed by Manuela Barreto of the University of Exeter and colleagues, published in Personality and Individual Differences in 2021. It drew on 46,054 participants across 237 countries, islands and territories, which makes it one of the largest loneliness datasets assembled.

Three patterns came out. Loneliness increased with individualism at the country level. It decreased with age. And it was reported at higher levels by men than by women, a finding we examined separately in is there a male loneliness epidemic.

The first of those is the one that travelled. Individualistic countries, lonelier people.

An earlier and smaller study pointed the same direction. Luzia Heu, Martijn van Zomeren and Nina Hansen published work in Personality and Social Psychology Bulletin in 2019 (volume 45, pages 780 to 793) examining individualism, collectivism and loneliness across European countries. They found collectivism correlated negatively with loneliness. More collectivist orientation, less reported loneliness, and this held even when people's social networks were small.

Two studies, different scales, same direction. That looks like a settled question.

Why it is not settled

Heu and her colleagues noticed something in their own data that complicates their headline.

Collectivist contexts came with higher implied ideals about how socially embedded a person ought to be. Expectations about family contact, about obligation, about how much of your life should be shared. Those ideals appear protective when they are met. They also create a gap that can be fallen into, because a person who does not meet them has further to fall.

This is the part the tidy story cannot accommodate. If loneliness is the distance between the relationships you have and the relationships you expect, then a culture that raises expectations is not straightforwardly protective. It changes both terms in the subtraction.

The title the authors chose captures it: lonely alone, or lonely together.

The age result that gets left out

While we are here, the Barreto analysis contains a second finding that deserves more attention than it receives, and it is relevant to how much trust to place in any of this.

Loneliness decreased with age across that multinational sample. The youngest respondents reported the most.

This runs against the assumption built into most loneliness policy, which targets older adults. It is not an isolated result either; the pattern of young adults reporting high loneliness has appeared in enough datasets that it is no longer surprising to researchers, even though it still surprises most readers. We looked at the generational version of this in why loneliness hits Gen Z hardest.

Two caveats keep it from being conclusive. The BBC sample was self-selected and recruited online, which plausibly over-represents younger and more digitally engaged people, and may also attract people who are lonely enough to click on a survey about loneliness. And the Tapia-Muñoz inequality study, which I lean on later, sampled only adults aged 50 and over, so it cannot speak to the age curve at all.

The reason this matters for the cultural question is simple. If loneliness varies this much by age, then any comparison between two countries is also partly a comparison between two age structures. Cross-national rankings that do not handle this carefully are measuring demography alongside culture.

The reframe: horizontal and vertical

Schermer's 2023 study is, as far as I can find, the most direct test of whether individualism and collectivism are even the right variables.

The team, including Marija Branković, Đorđe Čekrlija, Kristi Baerg MacDonald, Joonha Park and Eva Papazova among others, surveyed 8,345 participants across 28 countries. Rather than treating culture as a single spectrum from individualist to collectivist, they used a four-way scheme that is standard in cross-cultural psychology but rarely applied to loneliness.

The second axis is horizontal versus vertical. Horizontal cultures treat people as broadly equal in status. Vertical cultures organise around rank, competition and position. Crossing the two axes gives four orientations rather than two.

What they found, controlling for age and sex:

• Horizontal collectivism predicted lower loneliness, in 19 country samples
• Horizontal individualism predicted lower loneliness, in 7 country samples
• Vertical individualism predicted higher loneliness, in 16 country samples
• Vertical collectivism predicted higher loneliness, in 6 country samples

Read down that list and the individualism and collectivism labels stop doing any work. Both appear in the protective rows. Both appear in the risk rows. What sorts them is horizontal against vertical.

On this evidence, valuing your independence is not the problem. Living somewhere that ranks you is.

A second line of evidence, from a different direction

If hierarchy is the operative variable, it should show up in structural measures of hierarchy too, not only in surveys of cultural values. Income inequality is the obvious candidate.

In 2022, Thamara Tapia-Muñoz, Ursula Staudinger, Kasim Allel, Andrew Steptoe, Claudia Miranda-Castillo, Josefa Medina and Esteban Calvo published a study in PLOS ONE covering 75,891 adults aged 50 and over in the United States and 16 European countries, drawing on three major ageing panels: the Health and Retirement Study, the English Longitudinal Study of Ageing, and SHARE.

They found that older adults in countries with higher income inequality were more likely to report loneliness, after adjusting for individual demographics and for GDP per capita. The effect they report is a standardised unit increase in the GINI coefficient raising the odds of loneliness by 53 percent (odds ratio 1.52, 95% confidence interval 1.17 to 1.97).

Loneliness prevalence in that sample ranged widely: 25.32 percent in the United States, 17.55 percent in England, and between 5.12 and 20.15 percent across the European countries.

Note what that adjustment for GDP does. It is not a finding that poorer countries are lonelier. The association is with how unequally resources are distributed, not how many there are. We looked at the individual-level version of this question in financial strain and social isolation; this is the same relationship measured at the level of a whole country.

Two methods, two literatures, two very different samples. Cultural values measured by questionnaire, and economic structure measured by GINI. Both point at rank.

What this does not establish

A good deal, and the limitations are serious enough that the synthesis above should be held loosely.

Everything here is cross-sectional. None of these designs can separate cause from effect. Unequal societies might produce lonely people. Alternatively, something else about those societies produces both. The Tapia-Muñoz authors say directly that their design prevents causal conclusions, and flag possible unmeasured confounding at both the individual and country level.

The explained variance is often tiny. Schermer's team report that these cultural dimensions accounted for between 0.6 percent and 27.7 percent of self-reported loneliness depending on the country sample. The low end of that range is close to nothing. Culture is not the main story for most individuals, whatever it does to national averages.

Cross-cultural measurement is genuinely hard. Loneliness scales are translated between languages where the key words carry different weight, and the stigma attached to admitting loneliness varies between societies. If people in some countries are more reluctant to report it, the national comparison is measuring reluctance as well as loneliness. This problem has no clean solution and it sits underneath every cross-national number in this post.

Sampling. The BBC Loneliness Experiment was a large self-selected online survey rather than a representative sample. Its size is impressive. Its representativeness is not guaranteed, and the researchers were careful on this point.

One study is not a reframe. The horizontal and vertical result is, so far, a single cross-sectional study. It is the most direct test available of whether the usual axis is the right one, and it needs replication before it displaces anything.

How to read a national loneliness ranking

A practical note, because these league tables circulate constantly and are mostly read wrongly.

When a chart shows that country A is lonelier than country B, at least four things are baked into that difference before culture gets a look in.

1. The instrument. Direct questions ("do you feel lonely?") and multi-item scales like the UCLA measure produce different prevalence figures from the same population, because the scales approach the subject obliquely and catch people who would not describe themselves as lonely out loud.
2. Stigma. Admitting loneliness is more socially costly in some places than others. A lower national figure can reflect greater reluctance rather than fewer lonely people.
3. Translation. The target word does not carry identical weight across languages. Some languages distinguish solitude from loneliness more sharply than English does; some do not distinguish them at all.
4. Who answered. Age structure, as above, plus whether the sample was representative or self-selected.

The 5.12 to 20.15 percent spread Tapia-Muñoz and colleagues report across European countries is a fourfold difference. Some of that is almost certainly real. Some of it is instrument and stigma. Nobody can currently say how the total divides, and anyone presenting such a ranking without that caveat is overselling it.

This is also why I find the horizontal and vertical result more persuasive than the raw national comparisons. It is a within-person association between an individual's own cultural orientation and their own loneliness, repeated across 28 countries, rather than a comparison of national averages. That design sidesteps some, though not all, of the problems above.

What it suggests, held loosely

If the pattern survives, the practical reading is not that collectivist societies should be envied or individualist ones reformed.

It is that the social question which predicts loneliness may be closer to "where do I rank here?" than to "how much does this place value the group?" A horizontal culture, individualist or collectivist, offers relationships between equals. A vertical one makes many relationships into comparisons, and comparison is corrosive to the kind of closeness that protects people.

That reading also fits something from a different corner of this literature. If loneliness functions partly as a signal about social safety, as the sleep evidence we reviewed in what loneliness does to sleep suggests, then living somewhere you are constantly ranked is exactly the condition that would keep such a signal switched on. Status insecurity and social insecurity are not the same thing, but they are close relatives.

I want to be careful not to overreach. Four studies, all correlational, do not establish that hierarchy causes loneliness. What they establish is narrower: that the individualism question, asked for decades, has produced an answer that does not survive contact with a better-specified model, and that two independent measurement traditions point somewhere else.

A closing thought

The appeal of the individualism story is that it offers a culprit and a cure. Value the group more, be less lonely.

The evidence does not support anything that neat. It suggests, tentatively, that what matters is whether the people around you are arranged as peers or as a ladder. That is a harder thing to change than a cultural attitude, and it sits in economics and institutions as much as in values.

For anyone building tools for connection, including YaraCircle, the uncomfortable implication is that design choices which introduce ranking, visible metrics, or status comparison may work against the thing they are meant to help with. That is a hypothesis rather than a finding, and someone should test it properly.