What the 72% Figure in Gallup and Meta's Global Connection Survey Actually Measured

The number circulating since October is 72%. Gallup and Meta’s “State of Social Connections” report states that 72% of people surveyed across 142 countries feel very or fairly socially connected — roughly 3.2 billion people in the countries covered. It has already been cited as evidence that most of the world is doing fine on connection even as institutions from the World Health Organization to the U.S. Surgeon General treat loneliness as a public health priority. The two claims are not actually in tension, but reconciling them requires looking at what the 72% figure was built from, who paid for it, and what it leaves out.

One question, asked once

The connectedness figure comes from a single survey item: participants were asked how socially connected they feel, and could answer very, fairly, not very, or not at all. That is the entire measurement. There is no sub-scale, no attempt to distinguish structural isolation from subjective loneliness, no repeated administration to check stability over time. Gallup and Meta surveyed roughly 1,000 people aged 15 and older in each of the 142 countries, between June 2022 and February 2023, and one four-point question generated the headline statistic.

This is a legitimate way to ask a question. It is not a legitimate substitute for the instruments the loneliness literature actually relies on. The UCLA Loneliness Scale, used in AARP’s 2018 survey of adults 45 and older, is a 20-item instrument designed to detect loneliness indirectly, because direct self-report is known to undercount it — people are reluctant to describe themselves as lonely even when their answers to component questions indicate they are. AARP’s survey, using that scale, found that one in three older U.S. adults were lonely. A single direct question, of the kind Gallup and Meta used, tends to produce lower prevalence than a validated multi-item scale precisely because it invites a more flattering self-assessment. Comparing the 72% figure to loneliness rates measured with proper instruments is not comparing like with like.

Gallup’s own companion number complicates the story

The same Gallup dataset produced a second widely cited figure, published two weeks later: 24% of people worldwide report feeling very or fairly lonely, rising to 27% among adults aged 19 to 29 and falling to 17% among those 65 and older. If 72% feel connected and 24% feel lonely, the two do not sum to 100%, and Gallup’s own follow-up analysis says explicitly that connectedness and loneliness are not simple inverses of one another. A person can report feeling connected on the single-item measure while separately reporting loneliness, and a meaningful share evidently do.

This matters for how the 72% figure gets used. It cannot be read as “72% are not lonely,” because the survey did not ask a question designed to establish that. It measured a subjective global assessment of connectedness, which coexists, for some fraction of respondents, with a subjective assessment of loneliness. The instrument mismatch is not a flaw specific to Gallup’s methodology; it reflects the broader problem the field has struggled with for years. A 2023 review in BMC Public Health mapping the state of loneliness and isolation research names inconsistent measurement across studies as a central barrier to comparing findings, and this pair of figures from a single survey wave is a compact illustration of exactly that problem.

Who funded the question, and what it does not cover

Meta co-sponsored the survey. That does not automatically invalidate the data, and Gallup’s fieldwork and sampling methodology are the same it applies elsewhere. But a survey about social connectedness funded in part by a company whose core product is social media platforms warrants the same scrutiny any industry-funded research would get in a different domain. The report does not test whether social media use predicts the connectedness score one way or the other; it is a general population survey, not an evaluation of platform effects. Readers who encounter the 72% figure in coverage suggesting technology has not damaged social connection should note that the survey was not designed to answer that question at all.

There is also a coverage gap worth stating plainly: the survey excludes China, meaning it covers roughly 77% of the world’s adult population rather than the whole of it. A statistic described as “global” is, more precisely, global minus the largest single national population. That is not a minor rounding issue for a figure being used in international policy discussion.

Where this lands in the policy conversation

The timing is not incidental. The World Health Organization launched its Commission on Social Connection in November 2023, co-chaired by Vivek Murthy and Chido Mpemba, explicitly framing loneliness as a global public health priority rather than a concern specific to wealthy countries. The U.S. Surgeon General’s 2023 advisory, issued months earlier, put the U.S. figure at roughly half of adults experiencing loneliness and drew the comparison — now widely repeated — that the mortality risk associated with social disconnection is comparable to smoking up to fifteen cigarettes a day. Against that backdrop, a headline claiming 72% global connectedness reads, to an unwary audience, like a counter-narrative: perhaps the epidemic framing is overstated.

It is not a counter-narrative, because it is not measuring the same thing. The Surgeon General’s advisory and the WHO Commission are concerned with loneliness and isolation as health risk factors, established through decades of work using validated scales and mortality data. Gallup and Meta’s connectedness item is a one-time snapshot of self-assessed connectedness using a different instrument entirely, and its own companion loneliness figure — 24% worldwide — sits comfortably within the range other loneliness surveys report, once instrument differences are accounted for. Policy staff citing the 72% figure to argue against expanded intervention, or citing the 24% figure to argue for it, are both making a bigger claim than one survey wave can support on its own.

What would make this figure usable

A single cross-sectional item cannot establish whether global connectedness is rising, falling, or stable, because there is no earlier wave to compare it against. The most useful next step would be repeating the same instrument on a fixed schedule, ideally alongside a validated loneliness scale rather than a single parallel item, so that researchers could see how the two move together or apart over time and across the same populations rather than treating each Gallup release as a stand-alone fact. Adding China to the sampling frame would also matter more than most other refinements, given its share of the world’s population.

Until then, the figure has a narrow, legitimate use: it is one data point about how people in 142 countries answered one direct question about connection in a nine-month window in 2022 and 2023. It is not evidence that the loneliness concerns driving national strategies in the UK, Japan, and the United States are overstated, and it is not evidence that they are understated either. It is evidence of what happens when a single self-report item stands in for a construct the rest of the field has spent considerable effort trying to measure properly.