A base year is a ruler's zero mark.
Every Indian statistic carries a year in its name, and that year does two quiet jobs: it sets the zero against which everything is measured, and it freezes the weights that decide what counts. This page is about what changes when the year is moved, and what can never change.
The zero mark
First principlesRead any Indian statistic closely and a year is attached to it. The consumer price index is published on base 2024 = 100. GDP is measured at constant 2022-23 prices. The year is not decoration. It is the reference point every other reading is measured against, and the source of the weights that decide what matters in the number.
The base year also supplies the weights. The CPI basket's shares come from a household spending survey: the 2012 base took them from the survey of 2011-12, when food was a little under half of the basket; the 2024 base takes them from the newer spending surveys of this decade. Those shares are a photograph of how India spent around the base year, frozen into the index. Every print after is that photograph repriced.
Why bases grow old
Drift, slowly and one-wayA base begins to age the moment it is set. The 2011-12 spending survey behind the old CPI predated cheap mobile data, food delivery, smartphones in every pocket and the goods-and-services tax. By 2025 an index weighted on that India undercounted what the present India bought and overweighted what it had stopped buying. The drift is slow, but it runs one way: the ruler stays calibrated to a country that no longer quite exists.
So bases are revised, roughly once a decade. In 2015 the CPI was re-anchored to 2012 and GDP to 2011-12, and in 2026 to 2024 and 2022-23, each time with new weights, new items and new surveys underneath. The exercise is maintenance, not makeover. A statistic that never changed its base would grow more precise and less true with every passing year.
What a revision changes, and what it cannot
The map, not the territoryA new base changes the weights, the levels and sometimes the growth story. When GDP moved to the 2011–12 base in 2015, growth rates for years already past were restated, and some of them rose on paper without a single extra factory being built. That is not forgery. It is the same past measured with a better ruler, and the honest way to handle it is to publish the new series and say what changed.
What a revision cannot change is the past itself. The wheat harvested in 2012 does not grow because the ruler moved. What changes is the measurement: which goods count, at what weights, at whose prices. A base revision is a better map of the same territory. The old map was not a lie. It was old.
How to misread this concept
- i.
“A new base year is the government moving the goalposts.”
Base revisions happen under governments of every party, on roughly a ten-year cycle, because the economy changes under all of them. The methods are published and the new weights are public. Judge a revision by its stated method, not by whether you preferred the old number.
- ii.
“The base year is when things were normal.”
No normality is implied. 2024 is not a golden age; it is a zero mark. The index would work exactly as well anchored to any other year, because a base is a convention, not a verdict on the times.
- iii.
“Base revisions rewrite history.”
They rewrite the measurement of history. When the 2011–12 base restated earlier growth, no year was relived and no harvest grew. The ruler changed; the past did not.