The Gallup Numbers That Complicate the Loneliness Epidemic

The 2023 U.S. Surgeon General advisory opens with a stark claim: roughly half of American adults report experiencing loneliness, and the health risk of disconnection rivals smoking up to 15 cigarettes a day. That figure has become the anchor statistic of a policy movement, cited by government strategies, health systems, and newsroom coverage alike.

Gallup’s global survey of social connection, conducted with Meta across 142 countries between June 2022 and February 2023, tells a different story at first glance. Surveying roughly 1,000 people aged 15 and older in each country, it found that 72% of respondents felt very or fairly socially connected — about 3.2 billion people across the countries surveyed, which exclude China and cover roughly 77% of the world’s adult population. A companion release put the loneliness figure at 24% globally, and a follow-up survey in 2024 found just over one in five adults reported feeling lonely a lot of the previous day. Nearly half of respondents, 49%, said they did not feel lonely at all.

These are not small or marginal surveys. They are among the largest connection-specific datasets ever assembled. And they sit awkwardly next to the epidemic framing that dominates the American policy conversation.

Two questions, not one contradiction

The apparent conflict mostly dissolves once the instruments are separated. Gallup asked people how connected they felt and, separately, how often they felt lonely. Connectedness and loneliness, Gallup’s own analysis notes, are not simple inverses of one another — a person can report feeling both. The Surgeon General’s advisory, meanwhile, is built substantially on U.S.-specific survey traditions, including the Harvard Making Caring Common report that found 36% of Americans reporting serious loneliness, rising to 61% among adults aged 18 to 25, and Cigna’s 2020 workplace index, which put self-reported loneliness at 61% of U.S. adults on its own scale. These are not competing measurements of the same fixed quantity. They are different instruments, asking different questions, of different populations, at different moments, and arriving at different numbers almost by design.

This is the instrument mismatch that recurs throughout the loneliness literature, and it is worth naming plainly here rather than trying to average the figures into one number. A survey that asks “how often do you feel lonely” using a frequency scale will produce a different prevalence estimate than one using the UCLA Loneliness Scale, which AARP’s 2018 survey of adults 45 and older employed to find that one in three midlife and older Americans qualify as lonely. Gallup’s global instrument is closer to a single-item frequency question than to a validated multi-item scale. None of this makes any of these figures wrong. It makes them not directly comparable, and the practice of citing them interchangeably — as though “24% globally lonely” and “half of Americans lonely” describe the same thing at different scales — is itself a source of confusion that a 2023 BMC Public Health review flagged as a structural problem for the field: inconsistent measurement makes it difficult to compare findings across studies, let alone across countries.

What the Gallup data does support

It would be a mistake to read Gallup’s 72% connectedness figure as evidence that the epidemic framing is overstated. The same dataset that produced the reassuring headline also found that young adults aged 19 to 29 report the highest loneliness rate of any age group, at 27%, while adults 65 and older report the lowest, at 17% — a pattern that inverts the assumption, common in aging-focused policy work, that isolation is chiefly an old-age problem. That finding is consistent with the Harvard survey’s 61% figure for young adults and with the broader literature’s growing attention to adolescents and young adults as the demographic where disconnection concentrates, not the demographic where it is most visible.

What the global figures do complicate is the flattening of “epidemic” into a claim that most people, most of the time, are lonely. Gallup’s own framing is that a substantial global majority feels connected, and that loneliness, while common enough to be a serious public health concern, is not the modal experience. Both things can be true simultaneously: a condition affecting roughly a quarter of the world’s population, concentrated disproportionately among younger adults, is a legitimate public health priority without requiring that it describe the majority experience. The Surgeon General’s advisory does not actually claim that most Americans are lonely most of the time — it claims that roughly half report experiencing loneliness, which is a claim about incidence over some recall period, not a claim about a constant state. The rhetorical shorthand that has grown up around the advisory, in policy discourse, sometimes drifts past what the underlying figure supports.

What would settle this

A genuinely comparable cross-national dataset would use a single validated instrument — the UCLA scale, or an equivalent short-form version — administered with consistent recall periods, across representative samples, repeated over time. Gallup’s single-item measure has the virtue of scale and geographic reach that the academic literature generally lacks; the UCLA-based national surveys have the virtue of comparability to decades of prior research but rarely reach beyond single countries. Neither currently does both.

Until that exists, the honest reading of the evidence is that global connectedness looks considerably better than the U.S.-centric loneliness narrative implies, while the age gradient — young adults reporting more loneliness than older adults nearly everywhere it has been measured — looks robust across nearly every instrument used to date, from Gallup’s frequency question to Harvard’s survey to AARP’s UCLA-based scale. That gradient, not the aggregate prevalence figure, may be the more durable finding underneath all this measurement noise.