How '61% of Young Adults Are Lonely' Became a Policy Fact

In February 2021, Harvard’s Graduate School of Education, through its Making Caring Common project, published a survey reporting that 61% of young adults aged 18 to 25 experienced serious loneliness, alongside 36% of Americans overall and 51% of mothers with young children. The figure travelled fast. It appeared in press coverage within days, was cited in advocacy materials through 2021 and 2022, and by May 2023 had become one of the headline statistics behind the U.S. Surgeon General’s national advisory on loneliness and isolation, which frames young adults as a priority population.

That is a reasonable use of a survey finding. It is also worth tracing carefully, because the number’s route from data collection to federal policy document illustrates a pattern the Center sees often: a single instrument, administered once, generating a figure that gets repeated as though it were a stable epidemiological fact rather than one reading from one scale at one moment.

What the 61% figure actually measures

The Harvard report’s “serious loneliness” figure came from an online survey fielded during the pandemic, capturing a moment when university closures, remote instruction, and reduced in-person contact were unusually severe for the 18-25 age group specifically. The report itself makes this context explicit: 43% of young adults said their loneliness had increased since the pandemic began, and about half of lonely young adults reported that no one had spent more than a few minutes recently asking how they were doing in a way that felt genuine. That second detail describes something closer to a felt absence of meaningful contact than a technical isolation measure. It is a snapshot of a specific cohort in a specific disrupted period, not a baseline rate that can be assumed to hold in 2023.

Compare it with the instrument used in AARP Foundation’s 2018 survey of adults 45 and older, which found one in three reporting loneliness using the 20-item UCLA Loneliness Scale, a validated academic instrument with decades of use across populations. That comparability is exactly what lets researchers place the AARP figure alongside other UCLA-scale studies. The Harvard young-adult figure does not have that same anchor. Without knowing precisely how “serious loneliness” was operationalized in the survey instrument and how it maps onto UCLA-scale categories, the 61% cannot be directly compared to the loneliness rates reported by AARP, by Cigna’s 2020 workplace survey (which found 73% of workers aged 18-22 reporting loneliness, using yet another framing), or to the roughly 50% figure the Surgeon General’s advisory cites for U.S. adults generally.

This is the instrument mismatch problem in a fairly pure form: three different surveys, three different questions, three different numbers, all describing themselves as measuring “loneliness” among young people, and none straightforwardly reconcilable with the others.

How the number moved into policy

The Surgeon General’s advisory does not present the 61% figure as the sole basis for treating young adults as a priority group. It draws on a wider evidence base, including the American Enterprise Institute’s 2021 survey finding that 12% of Americans report no close friends at all, up from 3% in 1990, with the decline concentrated among men. That trend line, tracked across three decades with a consistent instrument, is a stronger piece of evidence than a single pandemic-era snapshot, because it demonstrates change over time rather than a level at one moment.

But the Harvard figure did the rhetorical work the trend line could not: it gave a specific, quotable percentage that fit neatly into a single sentence about a single population. That is precisely the property that makes a number attractive to press coverage and advocacy framing, and precisely the property that should make a reader ask what instrument produced it and when.

The advisory’s own headline claim, that roughly half of U.S. adults report experiencing loneliness, is presented with less epidemiological precision than the mortality comparisons it draws alongside it, such as the claim that the health risk of social disconnection is comparable to smoking up to 15 cigarettes daily, a framing built on the meta-analytic mortality work rather than a single prevalence survey. The advisory is on firmer ground when it argues about consequences than when it states a single youth loneliness rate as though it were settled.

Why this matters for how the figure gets used going forward

None of this means the underlying concern is wrong. There is a reasonably convergent signal across independent sources, using different instruments, that loneliness among young adults rose during the pandemic and that something distinct is happening with this age group relative to prior decades: fewer close friendships, per the AEI trend data, and disproportionately high self-reported loneliness during acute disruption, per Harvard and Cigna. Three sources using three instruments converging on “young adults are affected more than other groups” is a stronger claim than any one of them individually stating a precise rate.

The problem is treating 61% specifically as a durable population parameter rather than as one reading, taken under unusual conditions, using an instrument not designed for comparability with the rest of the literature. A 2023 review in BMC Public Health surveying the state of loneliness and isolation research names inconsistent measurement across studies as a central barrier to comparing findings, which is exactly the failure mode on display when a single pandemic-year figure gets folded into a federal advisory as though it described a stable rate three years later, in 2023, when conditions producing that number, closed campuses, remote instruction, curtailed in-person contact, no longer hold in the same form.

What would settle this more usefully is not a single fresh survey but a young-adult-specific tracking instrument, administered with a validated scale at repeated intervals before, during, and after the acute pandemic period, reported alongside network-size and contact-frequency measures rather than loneliness alone. Isolation and loneliness are not the same thing: a young adult can have a large network and still report loneliness, or a small network and not. The Harvard figure captures a subjective state at a moment of maximal disruption. It says less than its afterlife in policy documents implies about where things stand now.