One hundred is a place to stand.
Why statisticians abolish the rupee and measure everything against a single year set to 100. What the base year claims, what it does not, and why the fixed basket is both the index's strength and its blind spot.
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 principlesPrices are messy to compare. Rice sells by the kilo, bus fares by the trip, school fees by the term. To make one number out of three hundred such prices, statisticians perform a small act of violence: they abolish the rupee. In its place they build an index, a scale with no unit at all, anchored to a single year set equal to 100.
The choice of base year is exactly that: a choice. Nothing about 2012 was normal, average or specially blessed. It was simply the year the statisticians picked to stand on, helped by a fresh household spending survey that told them what the basket should contain. Any other year would serve; the ruler would just be marked differently.
The fixedness is the strength and the blind spot. Strength, because the index measures prices and nothing else. Blind spot, because households are not statues: when onions spike, people buy fewer onions and more of something else, and the frozen basket never sees the switch. The index answers its own question precisely, which is not always the question you meant to ask.
Same decade, two rulers
Index level, monthlyBelow is the consumer price level drawn twice: once as published, on the 2012 base, and once with every reading divided by the January 2016 value and multiplied by 100. The second line is the first line with a different ruler held against it.
Two rulers, one decade: the shape never moves
The two lines are the same series. The blue one is the published index, measured from 2012; the grey one is the identical data rescaled so that January 2016 reads 100. They rise together, flatten together and end the decade with the same percentage gain. Only the labels differ, because only the ruler differs.
That is the whole lesson. A level is a measurement, and every measurement needs its ruler named. The change is the fact; the level is the fact photographed from a chosen spot.
On the published ruler the decade ends at –. On the rebased ruler it ends at –. Both lines report the same rise since January 2016: –. Two different numbers, one fact.
Whoever quotes you a level without naming its base is selling you a ruler as if it were the world.
The arithmetic of a new ruler
A worked example, Jan 2016 → Dec 2025Rebasing is one division and one multiplication. Divide every reading by the base-month reading, –, and multiply by 100. Nothing is gained and nothing is lost: the new series reports the same changes as the old, month for month. This is why pure rebasing can never move an inflation rate; only the labels move.
One honesty note. When India revises an official base year, more changes than the zero mark: the basket and the weights are updated too, and then the rates themselves can shift. That is a real revision. The exercise on this page is the pure kind, arithmetic only, and it proves the point: the base is a choice of ruler, not a fact about the world.
How to misread this concept
- i.
“100 is the average price.”
It is not a price at all; it is a marker. The 100 means “we started measuring here”, and it could have been 1, or 1,000, with nothing else changed. The unit was abolished on purpose.
- ii.
“The base year is when things were normal.”
No year is normal; some years are merely chosen. 2012 was picked because a fresh spending survey existed for it, not because life was calm then. A base year is scaffolding, and scaffolding is not the building.
- iii.
“Two indices can be compared by their levels.”
Only if they share a base. An index at – on a 2012 base and another at – on a 2016 base can be reporting the very same decade, as the chart above shows. Compare the percentage changes, or rebase one series to the other. The levels are ruler talk.