IESET.
Conditions › Conditions favoring intervention

Health insurance adverse selection

Voluntary private health insurance markets face the Akerlof (1970) lemons problem in an acute form: insurers cannot perfectly price individual risk, so premiums set at the average reflect the expected loss of a mixed pool; low-risk individuals then exit, raising the average and the premium, iterating toward market unravelling. The problem is compounded by moral hazard, costly screening, and the fact that the sickest individuals — the ones insurance is most valuable to — are the least insurable. These features mean purely voluntary private health insurance markets systematically fail to cover the population efficiently.

confidence: highConditions favoring interventionentry added 2026-07-18health_insurance_adverse_selection

Institutional features that make the model work

›Mandate or automatic enrolment
Individual mandate (Netherlands Zorgverzekeringswet 2006, Switzerland LAMal 1996, pre-2017 ACA) or payroll auto-enrolment (Germany GKV, France) forces low-risk participation and prevents unravelling. Singapore's MediSave/MediShield layered system makes participation near-universal via CPF mandatory contributions.
›Risk equalisation across insurers
Where multiple private/non-profit insurers compete, a central risk-equalisation pool transfers funds from insurers with low-risk enrolees to those with high-risk enrolees, neutralising the incentive to cream-skim. Netherlands and Germany both run sophisticated risk-adjustment formulas.
›Public monopsony as alternative
Single-payer systems (Canada Medicare, UK NHS, Nordic systems, Taiwan NHI) sidestep adverse selection by enrolling the whole population by default. Achieve cost control through administrative pricing and global budgets.
›Community rating with subsidies
Forbid risk-based premium variation beyond limited factors (age, smoker status) and subsidise low-income enrolees on a sliding scale so the mandate is affordable across the income distribution.
›Competitive delivery with social financing
Delivery (hospitals, physicians, pharma) can remain private and competitive even when financing is public or mandated. The market-failure argument is about the insurance contract, not about medical services generally.

Supporting cases

netherlands_zorgverzekeringswet_2006

Mandated private insurance with risk equalisation and subsidies. Universal coverage, strong choice, OECD-average per-capita spending, high satisfaction. Shows that regulated private insurance can work when the mandate and risk-equalisation infrastructure are built properly.

germany_bismarck_system

Over a century of mandatory sickness-fund enrolment with competing non-profit funds and strong risk-equalisation. Universal coverage at ~12% of GDP with high quality.

singapore_cpf_medisave_medishield

Mandatory CPF contributions feed individual medical-saving accounts (MediSave), catastrophic insurance (MediShield), and subsidy scheme for the poor (MediFund). Among lowest health spending in the OECD-comparable set at strong outcomes.

taiwan_national_health_insurance_1995

Single-payer system established 1995, near-universal coverage at ~6-7% of GDP, high patient satisfaction, strong outcomes. Built in a middle-income country within five years — an existence proof of rapid implementation.

Disconfirming cases

us_pre_aca_individual_market

Pre-2014 US individual insurance market showed textbook adverse selection: medical underwriting, pre-existing condition exclusions, rescissions, high uninsured rates (~15%). Illustrates the unregulated failure mode.

us_post_aca_without_mandate

The 2017 removal of the individual mandate penalty did not fully unravel ACA exchanges but did worsen the risk pool and raise premiums, as ex-ante models predicted. Partial natural experiment on the mandate's importance.

us_overall_cost_vs_oecd_peers

US spends ~17% of GDP on health with worse life-expectancy and coverage outcomes than OECD peers spending 10-12%. The hybrid captured regulated system underperforms all the coherent alternatives — Bismarckian, single-payer, and mandated-private.

What this condition is NOT

  • A general argument that state provision of medical services is efficient
  • A claim that all healthcare sub-sectors exhibit this market failure — medical services, pharmaceuticals, and devices have different efficiency properties from the insurance contract
  • A claim that single-payer is always superior to mandated-private — both work
  • A vindication of the US status quo as a market system — it is a captured hybrid, not a market
  • A claim that adverse selection is the only mechanism causing high US costs — provider pricing, administrative overhead, and pharmaceutical pricing also contribute

Policy implications

The choice is between (a) single-payer or (b) mandated private insurance with risk equalisation and subsidies. Both work where implemented cleanly; voluntary private markets do not. The US underperformance on cost and outcomes is not evidence that "markets fail in healthcare" in a general sense — it is evidence that a captured regulatory-hybrid system that combines the worst features of both poles underperforms coherent systems at either pole. Reform should aim at institutional coherence, not ideological purity.

Framework position

Health insurance is a canonical case where the market equilibrium is demonstrably worse than achievable institutional alternatives. Adverse selection is a real mechanism, confirmed repeatedly in the US individual market and reproduced in natural experiments on mandate removal. The framework endorses two families of institutional solution — single-payer public monopsony or mandated private insurance with risk equalisation — and treats the choice between them as a matter of local political and administrative fit rather than a first-order efficiency question. It rejects the framing that US underperformance is evidence for markets' failure in healthcare generally, because the US system is not a market: it is a regulated hybrid whose poor outcomes reflect institutional capture and design failure layered on top of the underlying adverse-selection problem.