People without health insurance choose not to have it

The claim

The choice framing holds that uninsurance reflects voluntary opt-outs: young healthy people who calculate they do not need coverage, people who prefer to spend money elsewhere, or individuals who simply have not gotten around to enrolling. The implication is that the uninsured have access to coverage but exercise their freedom not to take it. This framing was used to argue against the Affordable Care Act’s individual mandate and against Medicaid expansion.

Survey data, natural experiments, and policy outcomes directly refute this framing for the overwhelming majority of the uninsured. The verdict is refuted.

The mechanism

Cost is the structural barrier, documented by self-report and structural analysis. KFF surveys consistently find that more than 75% of uninsured adults cite cost as the primary reason they lack coverage. This is not a hidden variable requiring sophisticated analysis — it is what people say when asked. The follow-up question is whether cost is an unavoidable structural reality or a reflection of prioritization. The answer is structural: employer-sponsored insurance has declined from covering 69% of the non-elderly population in 2000 to 54% in 2022, as employers have reduced or eliminated health benefits. Simultaneously, individual market premiums have risen faster than wages for most of this period.

The coverage gap is a direct policy artifact. The ACA created a coverage architecture with an explicit structural gap. In states that did not expand Medicaid, adults who earn too much to qualify for Medicaid (above ~100% of the federal poverty level) but too little to qualify for ACA marketplace subsidies (below 100% FPL) fall into a coverage gap. As of 2024, approximately 2.2 million Americans live in this gap — they are not uninsured by choice; they are uninsured because the policy architecture left them there. Many of these individuals are working adults in low-wage jobs.

Undocumented immigrants face categorical exclusion. An estimated 4–5 million undocumented immigrants are uninsured not because they choose to be, but because they are legally barred from enrolling in Medicaid and ACA marketplace plans regardless of their income or ability to pay. This is an explicit structural exclusion that cannot be characterized as a choice.

The Oregon Medicaid lottery as a natural experiment. In 2008, Oregon had more people eligible for Medicaid than it had slots, so it ran a lottery to allocate enrollment. This created a randomized control trial: lottery winners (Medicaid enrollees) vs. lottery losers (waitlisted, effectively uninsured). Finkelstein et al. (2012) found that winning the Medicaid lottery significantly reduced financial strain, increased rates of depression screening and treatment, and improved self-reported health. The lottery design rules out adverse selection — both groups were equally motivated to enroll. When structural access was provided randomly, people took it.

Medicaid expansion as policy experiment. The ACA’s Medicaid expansion, implemented at different times in different states, created a natural experiment across states. States that expanded Medicaid saw uninsurance rates drop by approximately 10 percentage points among low-income adults. Analyses by Sommers and colleagues found mortality reductions in expansion states. If uninsurance were primarily a choice, we would expect modest effects from simply making Medicaid available — the large, consistent coverage increases are inconsistent with the choice hypothesis.

Employer-sponsored coverage is not a choice. Most Americans receive insurance through employment, and whether an employer offers coverage — and at what employee cost-sharing level — is not within employees’ control. Workers at small businesses, in part-time or gig arrangements, or in low-wage industries are disproportionately uninsured because their employers do not offer coverage. This is a structural characteristic of the labor market, not a reflection of employee preference.

Who benefits

Employers who do not offer coverage benefit from the choice framing because it attributes absence of insurance to worker preference rather than employer omission. States that refused Medicaid expansion benefit because they avoid the modest state cost-share while federal funds go unspent — their framing of uninsurance as a choice provides political cover for this decision. Insurance companies benefit from maintaining a market in which the poor and sick are more often excluded, improving their risk pool. Politicians who oppose universal coverage benefit from framing the status quo as the result of individual choices rather than policy architecture.

The data

Comparators

United States — expansion vs. non-expansion states. The clearest domestic comparison is between states that expanded Medicaid under the ACA and states that did not. Among low-income adults, expansion states saw roughly 10 percentage-point reductions in uninsurance. Workers in non-expansion states with the same income, same employment status, and same stated desire for coverage remained uninsured because of where they lived. Geographic accident, not personal choice, explains the difference.

Canada. Canada’s single-payer system achieves near-universal coverage at lower per-capita cost than the US system. Canadian uninsurance rates are effectively zero for core medical services. Canadians did not become dramatically more responsible about choosing coverage when Medicare was introduced in the 1960s — the structural barrier (cost and access) was removed, and enrollment followed automatically.

Germany. Germany’s social health insurance system requires employer and employee co-contributions. Coverage is near-universal because enrollment is mandatory and cost is shared. The coverage rate is not the result of Germans having stronger preferences for insurance — it is the result of structural architecture that does not leave coverage to individual initiative.

The counter

A genuine “young invincibles” segment exists — young, healthy adults who are offered coverage and decline it, calculating (often correctly for the short term) that their expected medical costs are lower than premium costs. The ACA individual mandate was designed to address this actuarial problem. This segment is real and has some effect on risk pool dynamics.

However, this group represents a small fraction of the uninsured. Most of the 26 million uninsured are not young healthy opt-outs — they are adults who lack access to affordable employer-sponsored coverage, who fall in policy-created gaps, or who face prices that genuinely exceed their means. The choice framing generalizes from the minority to the majority in a way that is empirically unsupported and politically convenient.

References

Finkelstein, A., Taubman, S., Wright, B., Bernstein, M., Gruber, J., Newhouse, J. P., Allen, H., Baicker, K., & Oregon Health Study Group. (2012). The Oregon health insurance experiment: Evidence from the first year. Quarterly Journal of Economics, 127(3), 1057–1106.

Kaiser Family Foundation. (2022). Key facts about the uninsured population. KFF. https://www.kff.org/uninsured/issue-brief/key-facts-about-the-uninsured-population/

Kaiser Family Foundation. (2022). Employer health benefits survey 2022. KFF.

Sommers, B. D., Baicker, K., & Epstein, A. M. (2012). Mortality and access to care among adults after state Medicaid expansions. New England Journal of Medicine, 367(11), 1025–1034.

US Census Bureau. (2022). Health insurance coverage in the United States: 2022. Current Population Reports. https://www.census.gov/library/publications/2023/demo/p60-281.html