America’s Loneliness Map: What the Data Tells Us About Isolation

A measured public-health condition

In May 2023 the US Surgeon General issued *Our Epidemic of Loneliness and Isolation*, an advisory that formally elevated social disconnection to a public-health priority. The headline numbers in the advisory were stark: about half of US adults reported measurable loneliness in the years leading up to the report, and the mortality impact of chronic social disconnection was estimated to be comparable to smoking up to 15 cigarettes a day, drawing on the meta-analysis by Holt-Lunstad et al. (2015).

The framing matters because loneliness is now treated as a condition that can be measured, mapped, and intervened on, not a private feeling that policy avoids. The geography of that condition is uneven in ways that line up partly with what other indicators on this site track.

How loneliness is actually measured

There is no single national loneliness survey. The mapping work pulls from three main instruments.

• The UCLA Loneliness Scale and its short three-item form (Hughes et al., 2004) is the most-used standardized measure in the research literature.
• The CDC Behavioral Risk Factor Surveillance System (BRFSS) carries a "social and emotional support" item in many state samples, and is the source for most state-level estimates.
• Gallup, Pew Research, and Cigna/Evernorth run their own annual loneliness instruments. The Cigna US Loneliness Index is the most-cited proprietary version; Pew’s American Trends Panel is the most-cited public one.

Each instrument captures something slightly different. UCLA scales measure perceived isolation. The BRFSS support item measures available help, which is a different construct. Cigna measures self-reported loneliness on a composite scale. Cross-instrument comparisons in the Surgeon General advisory annex show that the prevalence figure shifts by 10 to 20 points depending on which instrument is used, which is a useful reminder when reading any single ranking.

What the geography shows

Three patterns are consistent across instruments and years.

First, young adults report the highest loneliness levels. The Harvard Survey on Loneliness found that 36 percent of all Americans, and 61 percent of young adults aged 18 to 25, reported "serious loneliness." This inverts a long-standing assumption that loneliness was primarily an aging-population issue.

Second, rural and low-density counties report higher chronic loneliness than dense urban counties, but the pattern flips for "any loneliness in the past week," which is higher in cities. The interpretation that has emerged from this is that cities produce more frequent, lower-intensity loneliness while rural counties produce less frequent but more sustained isolation. This is consistent with the social-capital findings from Chetty et al.’s 2022 Nature paper that we cover in our Geography of Opportunity Rankings piece.

Third, states that score low on the CDC BRFSS social-support item cluster in the South and parts of Appalachia. The states that score high are mostly in the upper Midwest, the Mountain West, and northern New England. These rankings correlate moderately with the geography of opportunity (about 0.45) and weakly with median household income (about 0.25), which means loneliness is not just a poverty story.

Why this matters beyond well-being

Loneliness is now treated as a risk factor for measurable health outcomes. The CDC Vital Signs report on social isolation in older adults documents elevated risk of cardiovascular disease, dementia, depression, and premature mortality associated with sustained social disconnection. The Cigna economic impact estimate puts the annual healthcare cost of loneliness-associated conditions at roughly $7 billion in additional Medicare spending.

The labor-market channel matters too. A 2024 Bureau of Labor Statistics analysis of the American Time Use Survey found that time spent socializing in person fell roughly 20 percent between 2003 and 2022, with most of the decline occurring before the COVID-19 pandemic. The pandemic accelerated a trend that was already underway, and remote work has not reversed it. AI-assisted communication tools, covered in our foundations explainer and LLM piece, have ambiguous effects on this trend: they make remote contact easier and they substitute for some kinds of in-person interaction.

What states and cities are doing

The interventions that have moved into practice fall into three categories.

• Built-environment policy that prioritizes "third places," walkable neighborhoods, and mixed-use zoning. The Strong Towns and Congress for the New Urbanism literature documents the design principles; the AARP Livability Index is the standard composite for community-design quality.
• Direct public-health programming, including the UK’s well-known social prescribing model and US analogues piloted in Vermont, Oregon, and Massachusetts. The Foundation for Social Connection tracks these pilots.
• Workforce and education interventions that build economic connectedness. The cross-class friendship findings from Chetty et al. (2022) suggest that mixed-income schools and youth organizations may be more cost-effective than individual mental-health treatment for the loneliness channel, although the evidence base is still thin.

The unifying thread is that loneliness is a structural outcome of how American life is organized: longer commutes, smaller household sizes, fewer civic memberships, more remote work, and weakened cross-class institutions. The conditions are different across counties, which is why a map is more useful than a national average. Our companion pieces on AI readiness by state and the American Dream Affordability Index cover related structural conditions that shape whether households can stay close to family, afford child care that includes social contact, and live in walkable neighborhoods.

How to read a loneliness map honestly

Three rules carry most of the weight.

The instrument determines the headline number. A map built on the UCLA short form will report higher prevalence than a map built on the BRFSS support item. Quoting the prevalence figure without naming the instrument is the most common error.

The age-adjusted view is the only fair cross-state comparison. States with younger populations will look lonelier on raw counts because young adults are the lonelier cohort. CDC publishes the age-adjusted version in its BRFSS prevalence data.

And small-sample variance is real. State-level loneliness estimates for small states carry wide confidence intervals and should not be ranked against neighbors with overlapping intervals.

The takeaway is that loneliness is now a measurable, mappable, and politically actionable indicator. The geography reflects deeper structural facts about American life, and the interventions that work look more like community-design and economic-connectedness investments than like individual mental-health treatment. The map is a starting point, not a verdict.

Sources

• US Surgeon General, *Our Epidemic of Loneliness and Isolation*, 2023.
• Holt-Lunstad et al., *Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review*, Perspectives on Psychological Science, 2015.
• Centers for Disease Control and Prevention, *Behavioral Risk Factor Surveillance System* and *Vital Signs: Social Isolation*.
• Cigna / Evernorth, *US Loneliness Index*.
• Harvard Making Caring Common, *Loneliness in America*.
• Chetty et al., *Social capital and economic mobility*, Nature 2022.
• Bureau of Labor Statistics, *American Time Use Survey*.
• AARP, *Livability Index*.