Three Countries, Three Loneliness Numbers

The United Kingdom’s official loneliness strategy, published in 2018, does not report a single national loneliness rate. Instead it committed the government to embedding a loneliness question into the Office for National Statistics’ ongoing surveys, so that the country could track change over time using a consistent instrument. Japan, which created its own loneliness ministry in February 2021, has taken a different route, drawing on academic surveys such as JACSIS rather than a single standardized national question. The United States has no federal loneliness measure at all; its most-cited figures come from private surveys — AARP, Harvard’s Making Caring Common project, Cigna — each built on a different instrument, asked of a different population, at a different moment.

The result is that when a policymaker or journalist says “loneliness in Country X is Y%,” the number is doing far less work than it appears to. It reflects a specific question, a specific age range, and a specific moment, not a stable fact about the population.

Same word, different questions

The AARP Foundation’s 2018 survey of adults 45 and older used the 20-item UCLA Loneliness Scale, the closest thing the loneliness literature has to a standard instrument, and found that one in three respondents qualified as lonely. Because it used an established academic scale rather than a bespoke question, the AARP figure is unusually comparable across other UCLA-scale studies. That is the exception, not the rule.

Harvard’s Making Caring Common survey, conducted in the fall of 2020 and published in February 2021, found that 36% of Americans reported “serious loneliness” — a figure built on the survey’s own composite measure, not the UCLA scale, and drawn from a period of acute pandemic disruption rather than a stable baseline. Cigna’s 2020 workplace report, using yet another instrument, put the share of U.S. adults who sometimes or always feel lonely at 61%, up seven points from the prior year. None of these three U.S. figures — AARP’s one-third, Harvard’s 36%, Cigna’s 61% — can be reconciled into a single “American loneliness rate,” because none of them measured the same thing in the same people. Reporting them side by side as if they tracked a trend is a category error the source data does not support.

What the UK and Japan are actually building

The UK strategy’s more interesting move is procedural rather than empirical: rather than commissioning one big survey and treating its number as definitive, it built loneliness measurement into an existing statistical infrastructure that already tracks other social indicators over time. That approach trades a splashy headline figure for something more useful to a health system — a consistent time series.

Japan’s ministerial appointment came three years after the UK’s strategy, in February 2021, and the two governments held what appears to be the first bilateral ministerial meeting on loneliness policy later that year, in June 2021. The joint statement from that meeting signals convergence on loneliness as a policy category, but it does not indicate that the two governments settled on comparable measurement. Japan’s own evidence base includes longitudinal work such as the JACSIS study, which tracked isolation and loneliness across the 2020–2021 pandemic period using repeated measurement of the same population — a design that can speak to change within Japan, but was never built to be stacked against the UK’s ONS figures or a U.S. survey headline.

This is the general problem with cross-national loneliness comparison: even governments that are explicitly coordinating on policy are not coordinating on instrument. A ministry can be created without a metric being standardized.

Isolation and loneliness get folded together, differently, in each country

A second source of mismatch runs beneath the instrument problem. Some of these efforts are measuring loneliness, the subjective feeling of lacking wanted connection; others are measuring social isolation, the structural fact of having few contacts, regardless of how a person feels about it. The AARP study’s discussion of network size and diversity as predictors treats isolation as an input to loneliness, not the same thing. Cigna’s workplace figures are about subjective feeling. National strategies tend to use “loneliness” as an umbrella term covering both, which makes their headline numbers even harder to compare, since a country with high structural isolation but strong subjective resilience could report a low loneliness figure and a high isolation one simultaneously.

What would actually make these numbers comparable

A cross-national loneliness comparison worth trusting would need the same validated instrument — the UCLA scale is the obvious existing candidate — fielded to comparable age-stratified samples in each country, at a specified and shared point in time, with isolation and loneliness reported as separate figures rather than folded into one word. Nothing meeting that description currently exists across the UK, Japan, and the United States. Until it does, the honest response to “which country is lonelier” is that the question cannot be answered from the data that exists — only “which country’s loneliness went up more, on its own instrument, over its own tracked period” can be, and even that comparison should stop at the water’s edge of the survey that produced it.