The Missing Economic Evidence on Loneliness

Cigna’s 2020 workplace survey reports that lonely employees miss work twice as often for illness and five times as often for stress-related reasons as their less-lonely colleagues, drawing on a survey instrument fielded across the US workforce. That is a real, quantified association. It is also nearly the only cost-relevant number in the loneliness literature that comes with a stated sample and method attached. Almost everything else that gets called an “economic cost of loneliness” in policy documents and press coverage is either a mortality risk ratio wearing a dollar sign, or an estimate with no visible study behind it at all.

This matters because the cost argument is doing real work. Governments and health systems do not fund screening programs, social prescribing schemes, or cabinet-level appointments on the strength of a hazard ratio alone. They fund them, in part, because someone has told them what inaction costs. The mortality evidence is genuinely strong. The economic evidence that sits next to it in policy documents is, on the whole, much weaker, and worth examining as its own question.

What the mortality literature actually measures

Julianne Holt-Lunstad’s 2010 meta-analysis, covering 148 studies and 308,849 participants, found that stronger social relationships were associated with a 50% greater likelihood of survival over follow-up periods, a magnitude the authors compared to established risk factors such as smoking. Her 2015 follow-up, disaggregating the construct, put the odds ratio for social isolation at 1.29, for loneliness at 1.26, and for living alone at 1.32, with the effects holding after adjustment for baseline health and appearing stronger in samples under 65.

These are risk ratios for mortality. They are not costs. Converting a hazard ratio into a dollar figure requires a chain of further assumptions: how many excess deaths the ratio implies in a given population, what those deaths would have cost to avert through medical care, what value is placed on the years of life lost, and how isolation’s effect on morbidity — as opposed to mortality — feeds into health spending before death occurs. None of the sources reviewed here perform that conversion. The National Academies’ 2020 consensus report on social isolation in older adults, which estimates that roughly one quarter of adults 65 and older are socially isolated, calls on the health care system to routinely assess isolation and loneliness. It does not attach a cost estimate to the quarter of the older population it identifies as at risk, and the clinical commentary that followed in the American Journal of Geriatric Psychiatry focuses on how screening might be implemented in practice, not on what isolation currently costs payers or systems.

That gap is not an oversight. It reflects how hard the underlying calculation actually is.

The absenteeism figure and its limits

The Cigna number is worth taking seriously precisely because it is the kind of evidence a cost argument should be built from: a defined outcome (missed workdays), a defined comparison group (lonely versus non-lonely employees), and a survey instrument applied consistently across a large sample. The same 2020 report finds that 61% of U.S. adults report sometimes or always feeling lonely, up seven points from the prior year, and that 73% of workers aged 18 to 22 report loneliness, rising above 80% among employed Gen Z workers specifically.

But absenteeism is not the same thing as cost, and self-reported missed days are not validated against payroll or productivity records. The survey does not report wage data, does not adjust for occupation or income, and does not attempt to translate missed days into lost output or replacement-hiring costs. It establishes that loneliness and absenteeism travel together in a large sample. It does not establish, and does not claim to establish, what that association is worth in dollars to an employer or an economy. Treating “twice as often” and “five times as often” as inputs to a cost model requires assumptions the survey itself does not supply.

Policy responses have outrun the cost evidence

The UK’s 2018 loneliness strategy, the first national strategy of its kind, is instructive on this point. It embedded loneliness measurement into the Office for National Statistics — a genuine methodological contribution, since it means the UK now has a standardized government time series where most countries have only ad hoc academic surveys — and it funded social prescribing, in which general practitioners refer patients to community activities rather than, or alongside, medical treatment. The strategy document frames this as a response to loneliness’s health and social consequences. It does not rest that funding decision on a published cost-of-loneliness estimate for the UK, because no such rigorously constructed estimate exists among the sources reviewed here. The funding decision preceded, rather than followed from, a settled cost figure.

Vivek Murthy’s 2020 book makes the broader case that loneliness should be treated as a public health issue rather than a private misfortune, and the argument leans heavily on the mortality and morbidity literature — the Holt-Lunstad meta-analyses recur throughout. It is a persuasive argument on those terms. It is not, and does not present itself as, an economic costing exercise, and it should not be read as one when it is cited in support of specific budget figures.

Why the calculation is hard, not just unattempted

Part of the difficulty is that isolation and loneliness are not single conditions with a single treatment pathway, the way a cost-of-illness study for diabetes or hypertension can trace a fairly standard sequence of diagnosis, treatment, and complication. Isolation is a structural property of a person’s network — how many contacts, how frequent, how reciprocal. Loneliness is the subjective experience of insufficient connection, and the two predict outcomes independently: a person can have a small network and not feel lonely, or a large one and feel isolated within it. Any economic model has to decide which construct it is costing, using which instrument, and that decision changes both the prevalence base and the plausible causal pathway to health spending. A model built on the National Academies’ one-quarter isolation estimate for older adults is answering a different question than one built on the AARP finding, using the 20-item UCLA Loneliness Scale, that a third of adults 45 and older report being lonely.

What a defensible cost estimate would require

A study capable of supporting the figures now circulating in policy discussion would need, at minimum, a longitudinal cohort with validated measures of both isolation and loneliness at baseline, linked administrative health records covering utilization and spending over a defined follow-up period, and a comparison against otherwise similar individuals who differ on social connection but not on the confounders — chronic illness, income, mobility — that independently drive both loneliness and health spending. That is a demanding design, and it is not what most of the widely cited figures rest on. Until that work exists, the honest position is that the mortality case for taking isolation and loneliness seriously is strong, and the dollar case, however often repeated, is mostly an inference stacked on top of it rather than a finding in its own right.