What’s actually inside India’s inflation basket?

When you hear “CPI inflation is 1.3%” (December 2025, Combined, MoSPI), you are hearing one number that merges the price tag of every roti, every litre of mustard oil, every school fee paid that month. That number is a giant weighted average. Statisticians at the National Statistical Office (NSO) track the prices of exactly 299 items, from rice to gold, from LPG to a doctor’s consultation, across markets and urban centres. Then they multiply each price change by a fixed weight that reflects how much India’s households, on average, spend on that item. The weights come from the Household Consumption Expenditure Survey (HCES).

Add up those weighted price changes across all items, and you arrive at the Consumer Price Index (CPI). The CPI’s percentage change over the year gives us the inflation rate.

The 2012-base CPI (the series that produced the December 2025 figure) grouped everything into six baskets:

Stacked bar chart: [India’s CPI Basket (2012 weights)](visual)

That 45.9% chunk for food makes the Indian CPI uniquely sensitive to the monsoon, to mandi prices, to the onion and tomato. But this basket is from a world where Indians spent nearly half their money on food. That has changed, and in February 2026 the basket changed with it.

Yes. On 12 February 2026, MoSPI launched a new CPI series with base year 2024, built on the HCES of 2023–24. The headline inflation it first reported for January 2026 was 2.75%.

The biggest change: food and beverages’ weight fell from 45.86% to 36.75%. This is Engel’s law made visible: as households get less poor, they spend a smaller share of their budget on food and more on services. Housing, water, electricity and fuels rose to 17.66%, transport and communication to 12.41%, health to 6.1%, clothing and footwear to 6.38%. The number of items in the basket grew from 299 to 358; it added e‑commerce and online prices, telecom data plans, and OTT streaming subscriptions. The old six groups were replaced by 12 COICOP divisions, including new ones like restaurants and accommodation (3.35%) and recreation (1.52%).

So the basket you see in today’s headline is fundamentally different from the one that generated the detailed month‑by‑month breakdowns available for 2025. Those monthly figures (vegetable inflation, contribution to headline, etc.) still come from the 2012‑base series and must be read as recent history under the old basket. Over time the new series will fill up with its own history; for now, the old series shows us the mechanics of how a price index works.

Weight is the share of the basket. Contribution is how much a group actually adds to the current inflation number. They are connected but not the same. The contribution of a group in a given month is roughly its weight multiplied by its own inflation rate. So a heavy group with near‑zero inflation can have a tiny contribution; a lighter group with a sharp price jump can have an outsized role.

Look at December 2025’s figures (2012 basket):

Food, with the largest weight, actually pulled the headline down because food prices fell year‑on‑year. The miscellaneous group, at 28.3% of the basket, contributed more than the entire 1.3% headline because services inflation was running at 6.2%. Weight tells you what matters over the long run; contribution tells you what matters right now.

The food and beverages group is itself a collection of sub‑groups that behave as if they live on different planets. Cereals are stable and heavy; vegetables are light and manic. In December 2025 (2012 base):

Cereals (index 197.4) and milk (192.3) moved more slowly and predictably. The cereal index rose from 92.6 in January 2011 to 197.4 in December 2025, a steady climb. But vegetables crashed: the tomato index touched 259.6, yet overall vegetable inflation plunged because a year earlier prices had been even higher. That single sub‑group dragged the entire food CPI into negative territory. When vegetables collapse, food inflation looks subdued even if your dal and cooking oil are still getting costlier.

A few things in the food basket give inflation its soap‑opera quality: vegetables (especially tomato, onion, potato), pulses, edible oils, eggs, and sugar. Their prices can double or halve in a season because supply is weather‑dependent and demand can’t easily adjust in the short run. December 2025’s item‑level numbers (2012 base):

Tomato prices rose despite the overall vegetable deflation; onion collapsed by almost half. A sabziwala in Azadpur mandi sees this weekly. When a storm flattens the Nashik onion crop, the index spikes and the evening news runs a “onion tears” headline. But these items combined are a tiny fraction of the overall basket. Their volatility causes the food inflation number to swing wildly month‑to‑month, but it seldom persists. The headline grab is real, but the long‑term anchor for your kitchen budget is still the steady creep of cereal, milk, and oil.

The “miscellaneous” group is a grab‑bag of services that don’t make headlines but relentlessly push living costs upward. In December 2025 (2012 base) its own inflation was 6.2%, far above the headline 1.3%. Within it:

Personal care inflation surged to 28.2%, the highest across the basket. Look at the absolute index levels. A doctor’s fee index stood at 223.6 in December 2025; medicine at 203.0; tuition fees at 204.7. These haven’t surged in one year, they have compounded. A school in Patna that raised fees by 10% five years back had already embedded that jump in the index; this year’s 3.4% is on top of a much higher base. Transport and communication inflation was only 0.8% because cheaper data packs and stable fuel offset rising petrol (index 140.2) and diesel (189.1). But the stickiest items, health and education, keep rising at a rate that won’t reverse in a good monsoon. That is why, even when food prices fall, the “core” inflation, which excludes food and fuel, was 4.6% in December 2025.

A group with a small weight can dominate the headline if its own inflation is high enough, while a heavy group with flat prices sits quiet. In December 2025, the miscellaneous group’s 28.3% weight produced a contribution of +1.75 percentage points to a headline that was only +1.33 points. Food, at 45.9% weight, contributed -0.85 points. The difference is a textbook case from index‑number methodology: contribution ≈ weight × own‑inflation. So when vegetables deflate by 18.5%, their weight is multiplied by a large negative number, dragging the headline down; when services inflate by 6.2%, their smaller weight gets multiplied by a positive number large enough to drive the headline upward.

That means the headline can be a tug‑of‑war. A family that spends more on dal, oil, and school fees than on tomatoes and onions will feel a higher personal inflation than the official 1.3% suggests. A rural household that grows its own vegetables but buys diesel and health services will have a different number again (rural CPI was 0.8% in December 2025, urban 2.0%). The basket is an average, not a mirror.

Don’t stop at the headline. Look at which groups and items are actually driving it. In December 2025:

• Headline: 1.3%
• Core (excluding food and fuel): 4.6%
• Food and beverages: -1.9%
• Miscellaneous (mostly services): 6.2%

If the headline is low only because vegetable prices have crashed, but health, education, and rent keep rising at 3–6%, then the average middle‑class urban household is experiencing something closer to core inflation than to the headline. The 1.3% figure is true for the statistical average household, but no actual household is exactly average.

Ask yourself: “Which items in this basket do I actually buy, and how much do I weigh each of them in my own monthly spending?” If you are a young renter in Bengaluru, the 10% housing weight and the 2.9% housing inflation may undershoot your reality; if you are a farmer in Vidarbha, the food weight may overstate your exposure to market prices. The MoSPI basket is one answer to a complex question; your own life is the sharper one.