Four Indias, one shared bill
India's health-insurance debate argues about one number: how much cover to add. That is the wrong question. India does not run one health system that needs more of something. It runs four systems under one name, and each one breaks at a different point in the same six-step journey from staying well to follow-up. So a single national fix lands on the wrong break for three of the four. Except one. There is one bill, though, that all four pay directly. Getting at it is cheaper than any new pool or premium, but even that single bill needs a different fix in each India. I named these four Indias in Edition 54. This one is about what to do with them.
↑ smaller, better served · larger, thinner cover ↓
Self-pay and premium insurance. Access complete, coordination absent. Sets the prices and holds the data the other three tiers are measured against.
Employer cover, ESIC, CGHS. Best access, worst continuity. Money stops at the first visit and again after discharge; the record does not survive a job change.
The missing middle. Own pocket and borrowing, no pool at all. One admission is paid for with savings, a loan or a sold asset.
PM-JAY and state schemes. Free once admitted, paying at every step before it.
01 — The wrong axis
The debate measures the wrong thing
Every reform conversation runs along one line: more insurance, or less. But cover is not a single gap you can widen. NFHS-6 puts household health-cover at 60.2%, up from 41%, which sounds like steady progress until you see that the 60% is four different situations wearing one number. The affluent household with a policy that excludes outpatient care, the salaried worker whose cover vanishes at a job change, the shopkeeper with nothing, and the scheme-enrolled family that pays for everything short of admission, all get counted the same way. The number hides the problem instead of measuring it.
02 — Four different breaks
Same chain, four different places it snaps
Follow any household through six steps, staying well, first visit, tests, treatment, bill paid, follow-up, and watch three things try to flow alongside them: the patient, the money, and the medical record. Each tier breaks somewhere different. India A carries all the way to discharge, then the record stops and nobody owns the follow-up. India B1 has the best access but the money stops at the first visit, because outpatient care is excluded, and again after discharge. India B2 breaks almost everywhere, and the hospital bill itself gets paid out of savings, a loan or a sold asset. India C is free once admitted and paying at every step before it.
View the data behind this chart
| Tier | Steps covered | Where it stops |
|---|---|---|
| India A | 5 of 6 | Carries to discharge, then follow-up drops |
| India B1 | 1 of 6 | Stops at the first visit, outpatient excluded |
| India B2 | 0.5 of 6 | Stops almost everywhere except the admission |
| India C | 1.5 of 6 | Free once admitted, pays every step before |
Author's assessment based on scheme design and NFHS-6 coverage data.
03 — Why one fix can't fit all
The favourite fixes each help one tier and miss three
This is why the popular fixes disappoint. Expand hospital insurance and you serve India C's admission and India A's surgery, while B1 and B2 keep paying cash for the outpatient care that never gets covered. Build tertiary hospitals and you help whoever can already reach them. Even the money that is spent lands in the wrong place: out-of-pocket is still 43.4% of all health spending, and for India C, non-medical costs like travel, lodging and food are 23.6% of what a covered hospital stay still costs the family. And the schemes themselves are a state lottery, household cover under PM-JAY-type schemes runs from 21% in Bihar to 90% in Chhattisgarh. "India is covered" is a sentence about Chhattisgarh.
View the data behind this chart
| State | Households with any health cover |
|---|---|
| Chhattisgarh | 90.5% |
| Rajasthan | 87.5% |
| Andhra Pradesh | 80.0% |
| India (average) | 60.2% |
| Maharashtra | 35.0% |
| Uttar Pradesh | 24.0% |
| Bihar | 21.1% |
Source: NFHS-6 (2023-24) India and State Fact Sheets, IIPS.
04 — The one bill they all pay
Medicines are the shared exposure. The fix is not shared.
Here is the one thing that does cut across all four. Medicines are 60.3% of everything Indian households pay out of pocket for outpatient care. The affluent family buying a monthly refill and the scheme-covered family whose health centre has run out are both paying at the same counter. It is the largest single line item in the part of the system no admission-based scheme touches.
India already built most of the answer. Jan Aushadhi generic stores sell at 50 to 80% below branded equivalents and sit in 776 of 784 districts. Yet they hold about 1% of a roughly ₹1.5 lakh crore medicine market. The shops are there. The prescriptions are not.
But this is where a tidy story would lie to you. The same bill has four different causes.
India A and B1: habit and margin, not software
Doctors here do not write brand names because their screen lacks a generic option. They write them out of habit, trade margins, and real doubt about whether every batch of a cheap generic is the same medicine. A price shown at the point of writing shifts the default, which matters. It does not settle the trust question, and a household saving two hundred rupees may not walk to a different shop anyway.
India B2: substitution is an economics problem
Let a chemist swap the brand and they will, for the version that pays them best. That is usually a private branded generic, not the unbranded one. Substitution without margin rules moves the label and leaves the bill roughly where it was.
India C: the prescription is already generic
Public doctors mostly prescribe by molecule already. The reason a scheme-covered family still pays is that the health centre is out of stock, so they buy from the private shop outside the gate. For this tier the fix is not the prescription at all. It is procurement and reliable supply.
View the data behind this chart
| Component | Share of outpatient OOP |
|---|---|
| Medicines | 60.3% |
| Consultations, tests and other | 39.7% |
Source: JAMA Network Open cross-sectional study of Indian out-of-pocket expenditure.
So the honest version is this. Medicines are the one exposure all four Indias share, which makes them the right place to start. But molecule-name prescribing on its own is a fix for India A and B1. B2 needs substitution rules that make the cheap option the profitable one. C needs public procurement that does not run dry. Same bill, three different levers, and none of them costs what a new insurance pool would.
05 — So what
Stop widening one number. Fix the joins.
For policy: prescribe by molecule, by default
Three moves, not one. Mandate molecule-name prescribing with the price shown as the doctor writes it, which bites in the private market. Set substitution rules so the cheapest quality-assured option is also the profitable one for the chemist, or the swap just moves to a pricier generic. And fix public procurement so health centres do not run dry, because a stockout sends a scheme-covered family straight to the shop outside the gate. Underneath all three sits the same unfinished job: batch-level quality assurance, until that is credible, no amount of price signalling will move a sceptical doctor.
For insurers and financing: cover the chain, not the admission
In three of the four tiers the break is outpatient: the first visit, the test, the monthly refill, not the hospital stay. And medicines are only 60% of that. The other 40% is consults and diagnostics, and for a missing-middle family managing diabetes or a cardiac condition, repeat testing is its own slow catastrophe. Cheaper medicines reduce the bill. Only a pool makes outpatient care actually covered.
For digital health: the rails already decide who wins
Generic substitution, a portable record that survives a job change, and a referral that pays the informal first provider are all rail problems, not hospital problems. Whoever gets the molecule, the record and the referral to travel is building the layer all four Indias share.
Find your own tier. Two companion tools go with this edition: a six-question check that tells you which of the four Indias you are actually in, and the full value chain tier by tier. Which India are you in? → · The value chain, tier by tier →
As Edition 54 put it, "pan-India" is usually a lie the average tells. India runs four health systems, and the standard reflex, add more insurance, aims at the one break three of them do not have. There is no single line that fixes all four. There is one bill they all pay, and three different reasons it stays high. Start with the medicine, but stop pretending one lever moves four systems. India already built the shops. It forgot to write the scripts, stock the shelves, and make the cheap box worth selling.
06 — Sources
Sources and evidence status
What is officially published, and what is my analysis. Nothing above is presented as fact without one of these tags.
Before citing externally: The four-tier frame is an analytical heuristic. Tier sizes shift with the definition used, so the argument does not rest on them.