A failed monsoon reaches the print three to six months late.
Food is nearly half the consumer basket, and most of that food grows under four months of rain. The weather inside the inflation number: the lag, the spikes, and the relief that arrives on the same delay as the pain.
The charts on this page draw the consumer price index on its 2012 base, the series that ran to December 2025. From January 2026 the NSO publishes the index on a new base, 2024 = 100. The idea on this page works the same on either ruler; the latest figures are in the Consumer prices guide.
The word, defined
First principlesIndia's inflation number has weather inside it. Food carries 45.9 per cent of the consumer basket by weight, and most of that food grows under a sky that delivers the bulk of the year's rain in four months. When the monsoon fails, the failure does not appear in the data the week it happens. It arrives later, at harvest, wearing the disguise of an onion.
The lag is mechanical. Rain falls, or fails, from June to September. The kharif harvest comes in from October, shortages reach the mandis over the following weeks, and the year-on-year print, which compares months twelve apart, keeps moving long after the skies have cleared. A bad monsoon in July is a food inflation story by winter.
The sky, visible in the data
Food y/y %, monthlyBelow is the food inflation rate for a decade, with the three episodes that teach the mechanism: the onion spike of December 2019, the vegetable spike of July 2023, and the deflation of 2025. One mechanism, three moments.
Rain in July, the print by winter: the lag, drawn once
This is the food inflation rate, the series that the basket's 45.9 per cent weight gives its power to. Read the three marked episodes as one mechanism. The shaded months are the 2019 monsoon; the spike marked beside them is its harvest, arriving at the December print. The 2023 spike was an erratic monsoon arriving almost in real time. The readings below zero in 2025 are the unwinding: a good harvest landing the winter after.
The 2019 monsoon was late, and the unseasonal autumn rain that followed it rotted stored onions across the growing belt. By December 2019 food inflation had reached –, the decade's highest reading, months after the rain had done its damage. Nothing about the December economy explained the December print. The explanation had fallen from the sky earlier that year.
The mirror image came six years on: after the spike of 2023 and a normal monsoon, food inflation fell through zero and bottomed at – in October 2025, closing the year at –. The same lag delivered relief that had once delivered pain.
Why weather becomes monetary news
The denominator under half the basketBecause food is 45.9 per cent of the basket, the sky gets a vote in monetary policy. A vegetable spike can push headline inflation through the ceiling of the RBI's band by itself; a winter glut can pull it below the floor. The central bank cannot make it rain. It can only decide how much of the shock to tolerate and how much to lean against with interest rates.
That is why forecasters in India read rainfall bulletins alongside bond yields, and why the companion entries on seasonality, index weights and inflation targeting belong on the same shelf as this one. The monsoon is the denominator under half the basket.
How to misread this concept
- i.
“Food inflation is the government's fault.”
Sometimes, in part: export bans, stock limits and import tariffs do move prices at the margin. But the first draft of every food print is written by the weather, months earlier. Blame and credit both arrive late.
- ii.
“One good monsoon fixes food prices.”
It fixes the winter after, once. The harvest that follows good rain lowers the level a single time; it does not stop the level from climbing again the summer after that. Food relief has an expiry date, and it is usually the next monsoon.
- iii.
“Food deflation is pure relief.”
Relief at the shop can be distress at the farm gate. When vegetables sell below the cost of growing them, the household budget smiles and the farm budget bleeds. The print records both as a single minus sign.