Why Young Adults Report the Most Loneliness, and Why That Is Not Settled

Gallup’s 2023 survey of 142 countries found that 27% of adults aged 19 to 29 report feeling lonely, against 17% of adults 65 and older. That gap — younger people lonelier than older people — shows up almost everywhere researchers look for it, and it inverts what most people assume about who is isolated. The stereotype of loneliness is an elderly widow living alone. The data says something closer to the opposite.

Cigna’s 2020 workplace survey found 73% of workers aged 18 to 22 reporting loneliness, against much lower rates among older cohorts. Harvard’s Making Caring Common project put serious loneliness among 18-to-25-year-olds at 61% in early 2021, compared with 36% for American adults overall. AARP’s 2018 survey of adults 45 and older, using the 20-item UCLA Loneliness Scale, found one in three lonely — high, but still below what the younger-cohort surveys report using comparable instruments. The pattern holds pre-pandemic, mid-pandemic, and in the most recent global data available: Gallup’s October 2024 update again found loneliness concentrated among people under 35.

What is far less settled is why. At least four explanations circulate, and they are not interchangeable — each implies a different intervention, and the evidence supporting each is uneven.

The phone explanation

The most publicized argument, from Jonathan Haidt’s 2024 book, holds that the shift from play-based to phone-based childhood in the early 2010s displaced the unsupervised, physical, peer interaction that used to build social competence and belonging. Haidt draws heavily on Jean Twenge’s generational survey work and argues the timing — a sharp rise in adolescent anxiety and loneliness coinciding with smartphone and social media adoption — is causal rather than coincidental.

The Anxious Generation’s own companion evidence review, a collaborative document maintained by Haidt and Twenge, is unusually candid about the state of the argument: it includes dissenting studies alongside supportive ones. That is worth taking at face value. The correlational timing is real. Whether phones caused the shift, accelerated a trend already underway, or are one factor among several competing for the same explanatory space is not resolved by timing alone. Critics dispute the strength of the causal claim, and that dispute is part of the record, not a footnote to it.

The life-stage explanation

A less dramatic but better-evidenced account treats elevated youth loneliness as partly structural rather than technological. Young adulthood is the period of maximum residential and relational turnover — leaving school, moving cities, changing jobs, forming and dissolving early relationships — all of which disrupt the standing, low-effort social contact that accumulates over a settled adult life. Gallup’s global analysis notes explicitly that connectedness and loneliness are not simple inverses: a young adult can report a large network and still score high on loneliness, because the instrument is capturing something about subjective fit, not just headcount.

This explanation predicts that loneliness should decline with age as people settle into stable networks — which is exactly the shape of the AARP and Gallup data. It does not require phones as a mechanism at all, and it fits equally well in countries with very different smartphone adoption curves, which the phone-specific account has more trouble with.

The friendship-erosion explanation

A third account points not at technology or life stage but at a longer decline in the raw stock of close friendships across the whole population, which happens to land hardest on the young. The American Enterprise Institute’s 2021 survey found 12% of Americans report no close friends at all, up from 3% in 1990 — a fourfold increase — and that the share of men with six or more close friends fell from 55% to 27% over the same period. WBUR’s 2024 reporting on American Time Use Survey data adds a detail that matters for adjudicating between explanations: face-to-face socializing was already declining before the pandemic, and continued declining after it, which weakens any account that treats COVID-19 disruption or recent-vintage smartphone apps as the primary driver. Something was already moving before either the pandemic or, in Haidt’s telling, the critical years of phone-based childhood.

If friendship stock has been eroding for three decades, younger cohorts would be the ones to inherit a thinner starting baseline — never having built the informal civic and social habits described in the earlier literature on associational decline. That is consistent with the age pattern without requiring a phone-specific mechanism, though it does not rule one out either.

The workplace variant

A fourth, narrower strand concerns work rather than adolescence. Gallup’s 2024 State of the Global Workplace survey found loneliness higher among employees under 35, and higher again among fully remote workers (25%) than hybrid (21%) or fully on-site (16%) employees. Younger workers are disproportionately represented in remote and early-career roles with less established workplace social capital, which offers a fifth-order explanation specific to the workplace slice of the age gap, distinct from adolescent smartphone use or generational friendship decline.

What would actually distinguish these

None of these four accounts is well-designed to rule out the others, because almost all the supporting data is cross-sectional. A cross-sectional survey — Gallup’s, Cigna’s, AARP’s — captures loneliness and its correlates at one moment and cannot show which came first, or whether a third factor drives both. The AEI friendship data and the WBUR time-use reporting are closer to time series, which is why they are useful for the timing argument against a phone-specific 2012 inflection point, but neither was designed to isolate mechanism.

A study capable of separating these explanations would need to track the same individuals across the transition from adolescence into settled adulthood, measuring smartphone and social media use, residential and relationship stability, and network size and quality simultaneously and repeatedly, then testing which trajectory predicts the decline in loneliness that is visible cross-sectionally between 25-year-olds and 55-year-olds. It would also need to be run in countries with different smartphone adoption timelines and different baseline rates of civic and associational life, since a mechanism that is genuinely about phones should not depend on the collapsing club membership Robert Putnam documented decades before smartphones existed.

Until that kind of design exists, the honest summary is narrower than any of the four accounts on its own: young adults are reliably the loneliest age group in survey after survey, across countries and instruments, and at least four plausible mechanisms compete to explain it, with evidence sufficient to rule out none of them and to confirm all of them fully.