India doesn't have a financing model. Nor do 94 other countries.
Every discussion of Indian health financing eventually reaches the same question: is India a tax-funded system or an insurance system, Beveridge or Bismarck? I built an atlas of 180 countries to answer it properly, and the finding is that the question does not work for most of the world. Ninety-five of the 180 have no single financing model. India is one of them, and so is over half the planet.
The useful question is not which model a country has. It is how it raises money, how it pools risk, who purchases care, and how providers get paid. Those four are separable, and countries mix them freely.
- 95 of 180 countries, 53 per cent, are mixed or segmented. No single mechanism describes them.
- The three textbook models between them cover 79 countries: 27 tax-funded, 27 public insurance, 25 social health insurance. Less than half the world.
- Out-of-pocket spending is not a model. Only 9 countries carry an out-of-pocket flag, and none is classified by it. Nor is donor dependence a model, though 29 countries are aid-sensitive.
- India sits in the largest group, with five pooling arrangements running side by side: central and state budgets, PM-JAY and state schemes, ESIC, employer and private insurance, and household payment.
- The framing is regional. Europe is 3 per cent mixed; Africa is 81 per cent. Beveridge and Bismarck describe the one region where the categories hold.
- Every row carries an evidence grade. 41 are A, 101 B, 38 C. The grades are published because a 180-country claim should show its own confidence.
02 — The distribution
What 180 countries actually look like
Those first three numbers are why the classification matters rather than being a filing exercise. Financial hardship from health costs is not a rare event affecting a minority; it is the experience of roughly a quarter of humanity. And it is produced by financing design, which is the thing the single-model question is too coarse to examine.
Classified using the SHA 2011 health-care financing schemes, the official framework, rather than the country labels that circulate in textbooks and consulting decks.
View the data behind this
| Architecture | Countries | Share |
|---|---|---|
| Mixed / segmented | 95 | 53% |
| Tax-funded public | 27 | 15% |
| Public / national insurance | 27 | 15% |
| Social health insurance | 25 | 14% |
| State-budget integrated | 3 | 2% |
| Mandatory private insurance | 2 | 1% |
| Medical savings + insurance | 1 | 1% |
Author's classification of 180 countries using the SHA 2011 financing-scheme categories. Archetype labels are a navigation aid, not official WHO designations.
03 — The method
Four questions instead of one label
WHO recommends looking at financing functions rather than country labels. Once you separate them, the single-model habit stops making sense, because a country can answer each question differently.
The test that matters cuts across all four: does the design move payment away from the moment of sickness and into prepaid, pooled financing? High out-of-pocket spending means the sick household is carrying the risk directly. That is a property of the whole arrangement, not of any one label.
And provider payment is its own layer, routinely confused with the others. Where the money came from and how the hospital gets paid are different questions. A tax-funded system can pay by DRG; an insurance system can pay by salary.
04 — The finding I did not expect
The single-model idea is a European artefact
Break the 180 countries down by region and the pattern is not subtle. Europe is the only place where clean models dominate. Everywhere else, mixed is the norm.
Which explains why the framing feels so natural. Beveridge and Bismarck are European names, describing European arrangements, in the one region where the categories actually hold. Then they were exported as a universal template to a world where 53 per cent of countries do not fit either.
So when an Indian policy discussion asks whether the country should follow the UK or Germany, it is importing a taxonomy from the least representative region on earth. India's real peer group is the 24 other mixed systems in Asia, and the 44 in Africa, not the two European countries that get named most often.
05 — India
India's actual row
Not a model. Five arrangements coexisting, by history rather than by design.
From the companion atlas. Grade B means the architecture is stable and cross-checked against a secondary source, not that every detail is primary-sourced.
Read that pooling line again. Central and state budgets, PM-JAY and state schemes, ESIC, employer and private insurance, and household out-of-pocket. That is five pools, each with its own eligibility rule, its own benefit definition and its own purchaser. Nobody designed that. It accumulated.
Which is exactly the four-Indias picture from Edition 68, arrived at from the opposite direction. That edition traced how a household experiences the segmentation. This one shows that the segmentation is India's actual classification, and that most of the world shares it.
06 — The consequences
Four things this reframing kills
07 — The flags
What gets flagged rather than classified
Eight conditions are recorded as flags on a country row instead of becoming its label, because each describes a pressure on financing rather than a structure.
The two largest flags are the same 29 countries seen twice, once for external financing mattering and once for aid sensitivity. That pairing is the real finding: where donor money matters most, the financing system is also most exposed to it changing. Out-of-pocket heavy applies to only 9 countries, which is worth remembering next time someone calls high out-of-pocket spending a model.
08 — The honest part
What I am less sure about
A 180-country claim invites a fair question: how would anyone verify it? So every row carries a grade rather than an implied uniform confidence.
The corrections I made while building it are published in the atlas itself, and they are the useful part. Egypt, Kenya, South Africa and Uzbekistan are marked as reforming rather than presented as finished designs. Cabo Verde, Djibouti and Algeria were moved off an over-simple social insurance label. Suriname was moved off a national insurance label. China is described as a public basic medical insurance architecture with multiple local pools, not as a single payer.
What the atlas does not do. It classifies architecture. It does not rank countries, does not score performance, and the archetype labels are a navigation aid rather than official WHO designations. If you need out-of-pocket or government-spending shares for a specific country, take them from the WHO Global Health Expenditure Database directly.
09 — So what
What to do with this
Stop benchmarking India against the wrong comparators
Britain and Germany are 15 and 14 per cent of the world respectively, and neither is segmented the way India is. The useful comparators are the other 94 mixed systems, particularly those that have made segmented pools interoperate rather than merge.
Argue about functions, not models
"Should India move to single payer" is unanswerable. "Should India's five pools share one identity, one benefit floor and one claims rail" is answerable, and it is the actual reform question.
Stop calling out-of-pocket spending a model
It is the outcome of a financing design, and treating it as the design makes the problem sound like a choice somebody made. Nine countries carry the flag. None is defined by it.
The single-model question survives because it is easy to ask and easy to answer wrongly. India is not moving toward Beveridge or Bismarck. It is one of ninety-five countries that will have to make segmentation work, because merging the pools is not on any realistic timetable. That is a less satisfying answer and a considerably more useful one.
Explore it yourself. The full atlas covers all 180 countries with revenue, pooling, mechanisms, flags, reform status and the evidence grade, filterable by region and architecture, with a CSV export. Open the financing atlas →
10 — Sources
Sources and evidence status
What is official, what is derived, and what is my classification. Nothing here is presented as an official designation unless it is one.
Before citing externally: the archetype labels are mine, not WHO's. 38 of the 180 rows are graded C and should be treated as indicative. Financing architecture changes with reform, so check the position at source before quoting any single country.