Medical inflation in India is routinely quoted at 12 to 14 percent a year and treated as the rate at which the price of healthcare is rising. It is not a price measurement at all.

The two most cited sources sit on the same calendar and disagree by an order of magnitude. Aon’s 2026 Global Medical Trend Rates Report, published on 11 December 2025, puts India’s 2026 medical trend at 11.5 percent, down from 13.0 percent for 2025, against a global average of 9.8 percent. The CPI release for July 2026 on base 2024=100 puts year-on-year inflation in the Health group at 1.34 percent. What follows concerns one component of that gap: what each number counts, and why the difference is a definitional artefact rather than evidence of concealed price inflation. It is not an argument that Indian healthcare is getting cheaper, or that households should hold less cover.

A Trend Rate Counts Spend Per Member, Not Price Per Procedure

The definition is worth stating precisely, because it is published and rarely read. Aon defines a medical trend rate as the annual percentage increase in medical plan costs per employee, across both insured and self-insured plans. The numerator is total plan spend. The denominator is covered employees. Nothing in the construction holds treatment volume constant.

The figure therefore moves on five things at once: the unit price of care, how often members claim, the severity and case mix of what they claim for, changes to benefit design, and the age profile of the covered group. Aon attributes India’s step down from 13.0 to 11.5 percent to moderated utilisation levels. Utilisation is a volume variable. A number that falls because people went to hospital less often is not reporting that hospitals charged less.

CPI Prices a Fixed Basket. A Trend Rate Prices a Moving One.

The natural reading of the gap is that CPI understates what hospitals charge. The more useful reading is that the two indices measure different quantities.

India’s CPI was rebased to 2024=100 in February 2026, with weights drawn from the Household Consumption Expenditure Survey 2023-24, and the Health group carries a weight of 6.100 percent of the combined basket. Within a series, that basket and those weights are held fixed. The index answers one question: what does the same basket cost now against a year ago. Quantity is constant by construction.

A trend rate answers a different question: what did the plan spend this year against last. Quantity is free to move, and in an ageing, wealthier, more insured population it moves upward on its own.

So this is not an honest measure against a dishonest one. It is a fixed-quantity price index against a variable-quantity spend measure - a difference of kind, not degree, and enough to account for most of a ten-point gap without anyone misreporting anything.

The Two Numbers Sit Ten Points Apart on the Same Calendar

What each published “medical inflation” figure for India actually counts. The blue bars are fixed-basket price inflation from the CPI release for July 2026 on base 2024=100; the gold bars are Aon’s projected medical plan cost per employee. They are not the same measurement, and the axis is the only thing they share.

The row that matters for insurance is inpatient care services, at 1.57 percent. That is the line a hospitalisation policy reimburses, and it is running below headline CPI of 4.45 percent. Whatever produces an 11.5 percent trend rate, the surveyed price of inpatient care is not the bulk of it.

The Headline Figure Is a Forecast, and It Has Been Revised Down

The published numbers are also not observations. Aon’s report is built on insights from more than 100 of its offices that broker, administer or advise on employer-sponsored medical plans, and reflects those professionals’ expectations from their interactions with clients and carriers. It is a projection of next year’s plan costs, of the kind an actuary would call a pricing assumption.

Run the available estimates together and they do not converge. Aon has India at 13.0 percent for 2025 and 11.5 percent for 2026 - decelerating. WTW’s 2026 survey has Asia Pacific medical inflation at 14 percent for 2026 against 13.2 percent for 2025 - accelerating, and regional rather than Indian. Milliman, in The Actuary India in April 2026, records a trend of 12 percent for 2024 and 13 percent projected for 2025, against general inflation of 4.2 percent.

The estimates span roughly 11.5 to 14 percent for overlapping periods and disagree about which way the line is pointing. The 14 percent that circulates most widely in Indian financial planning is the top of that band, and it is usually quoted without its year, its survey, or its definition attached. Any single headline figure is a choice of survey presented as a measurement. The range, and the disagreement about direction inside it, is the finding.

A Retail Buyer Is Being Sold a Group Plan’s Number

A second distinction is lost when the figure crosses from a broker report into a financial plan: the population changes. Aon and WTW measure employer-sponsored plans - group cover, a working-age employee base, network rates negotiated by a corporate buyer. The number then gets quoted to a 55-year-old buying an individual retail policy, whose premium moves on filed age bands and portfolio repricing, and whose pool has neither the age profile nor the buying power of the group surveyed.

Joanne Buckle, FIA, Principal and Consulting Actuary at Milliman, with Rachin Aggarwal and Heena Arora, has argued for exactly what is missing here: a practical, data-led method for measuring medical inflation in India, and a credible published index that would reduce the standing conflict between insurers and hospitals over what costs are doing. Their case rests on a premise worth sitting with. India has no agreed measure of medical inflation. It has broker forecasts of employer plan spend, and a fixed-basket price index that was never built to price an insurance book.

What This Means for Advisers

None of this settles whether healthcare is becoming less affordable for Indian households. It is, and the strongest case against the framing here is that CPI’s health basket may miss where the money goes: out-of-pocket spending is 39.4 percent of total health costs in India, private tertiary pricing is the segment least likely to be captured in a fixed sample of outlets, and a fixed basket lags a shift toward more expensive procedures by construction. The narrower claim is about what the quoted number measures, not about whether costs are rising.

Three things follow for anyone building a plan around it.

Source the escalation assumption before using it. A medical inflation input should carry a survey name, a year and a definition, in the same way an equity return assumption carries an index and a period.

Separate price escalation from utilisation when projecting a client’s own costs. Only the first belongs in a lifetime projection at a constant rate; the second is a function of the client’s age curve and should be modelled as such.

And model retail premium escalation off the insurer’s filed age bands and its history of portfolio revisions, not off a group trend rate. The two reprice on different mechanisms and different schedules.

Healthcare costs are rising. The figure in circulation is a forecast of someone else’s plan. What actually reprices a retail policy is the insurer’s own book, and nobody publishes that number.

4/7 - Rethinking Health Insurance

Back to Writing