Your inflation and the number

Chapter 2 built the number. This chapter is about why you have never felt it.

The month the onion did it

From the September 2026 release, the five items with the highest inflation at the all-India combined level:

Item	Weight in the basket	Inflation, Aug 2026	Inflation, Sep 2026
Onion	0.7006%	48.78%	108.09%
Silver jewellery	0.3127%	106.81%	87.80%
Ginger	0.2556%	73.82%	62.85%
Garlic	0.3738%	43.60%	54.62%
Lemon	0.1367%	15.15%	36.41%

And the five lowest, from the same table, in the same month:

Item	Weight	Inflation, Aug 2026	Inflation, Sep 2026
Potato	0.7549%	−13.15%	−11.50%
Tomato	0.4961%	−31.09%	−8.62%
Motor car and jeep	0.4665%	−6.73%	−4.51%
Lady's finger	0.2488%	−5.41%	−4.48%
Coconut	0.1931%	−3.00%	−3.99%

Onion was up 108.09% and potato was down 11.50% in the same month, in the same basket, measured by the same field staff. A spread of nearly 120 percentage points between two vegetables that sit next to each other in the same shop.

The headline for that month was 5.54%. It is the average of those and about three hundred other things, and it describes the experience of nobody in particular.

The arithmetic of a weight

Now the part that makes the headline legible, and it is one multiplication.

An item's contribution to the headline is its inflation times its weight. Onion rose 108.09% carrying a weight of 0.7006%, so it contributed:

108.09% × 0.7006% = 0.757 percentage points

Headline inflation rose from 4.83% in August to 5.54% in September — a jump of 0.71 points.

So onion alone contributed more than the entire month-on-month increase in the national inflation rate. Everything else, net, slightly offset it.

Potato went the other way: −11.50% at a 0.7549% weight contributed −0.087 points, dragging the headline down by about a twelfth of a point while sitting in the same shopping bag.

Two lessons, and the second is the useful one.

A small weight with a huge move can run the headline. Onion, potato and tomato together are 1.95% of the basket, and in September 2026 they decided the direction of a number that moves national policy.

Which means a monthly print is often about vegetables. When the headline jumps, the first question is not "what is happening to the economy". It is "which item moved, and how much of the basket is it?" The release prints both columns so you can answer it in a minute.

Three official inflations, published together

The release does not give one number. It gives a family of them, and the gaps are the information.

September 2026	Headline	Food (CFPI)
Rural	6.08%	8.06%
Urban	4.88%	7.18%
Combined	5.54%	7.74%

Rural inflation ran 1.20 points above urban. The reason is structural rather than mysterious: food is a larger share of rural household spending, and food inflation at 7.74% was running 2.20 points above the headline. Put a bigger weight on the fastest-rising part of the basket and you get a bigger number. The same month, the same prices, a different basket.

This is also the clearest answer to "why does the government's number feel wrong". If you are a rural household, the government published a number for you, and it was 6.08%. It is in the same table as the one the headlines used.

So what is your inflation?

Your personal inflation rate is the same calculation with your weights. Three predictable reasons yours will differ from 5.54%:

Your food share. The single biggest driver of the gap. Food was 7.74% against a 5.54% headline, so a household spending more than the basket's share on food had higher inflation than the headline, mechanically, with nothing else changing.

Whether you rent or own, and when you bought. Housing moves slowly and in steps. A renter whose landlord has not raised rent this year has zero inflation on a large part of their spending, while someone whose rent just moved 10% has a shock the index smooths across everybody.

Your stage of life. The basket contains education services and health in fixed proportions. A household with two children in school and a household with none are buying measurably different inflation rates, and neither is the published one.

None of this makes the headline wrong. It makes it an average, which is what it says it is. The error is not in the statistic; it is in reading a national average as a personal statement.

Where to look when you disagree with it

The release publishes division-level inflation for all twelve divisions, item-level rates for the biggest movers, and state-level figures. In September 2026 the five large states with the highest headline inflation were Tamil Nadu at 6.81%, Odisha 6.70%, Madhya Pradesh 6.68%, Telangana 6.64% and Andhra Pradesh 6.29% — against a national 5.54%.

A reader in Tamil Nadu who thought the official number felt low was right, and the official number agreeing with them was published the same day in the same document.

The state and item detail goes further on MOSPI's eSankhyiki portal, with the caution chapter 2 quoted: some state-level item indices rest on thin samples and should be used carefully.

The point

The headline is a weighted average and the weights decide it: in September 2026 onion rose 108.09% at a weight of 0.7006% and contributed 0.757 points, which is more than the entire 0.71-point rise in the national rate that month. Potato fell 11.50% in the same basket. Rural inflation was 6.08% against urban 4.88% because food is a larger share of rural spending and food was running at 7.74%. Your number differs from the headline because your basket does, and the release publishes enough detail for you to find out by how much.