The government recently announced a revised Consumer Price Index basket. Inflation is the percentage change of the Consumer Price Index over time. Inflation statistics determine how we interpret growth, assess macro stability, and protect real incomes.
The Consumer Price Index (CPI) is the nominal anchor of macroeconomic policy. This makes the new basket an important development, one with implications for everything built around the CPI, such as monetary policy, pension indexation, and fiscal arithmetic. The benefits of this revision will be fully realised when the authorities publish a properly backdated series so that the new CPI can be compared with earlier data.
What has changed?
The basis for measuring inflation is the cost of a “basket” of goods and services that a typical household consumes (food, rent, mobile phone, laptop, fuel, transport, health, education, etc). The expenditure on these items is measured through a household survey. Each item is given a weight based on how much people spend on it. The base year is one which reflects the year of the survey. Once the basket is fixed, one can measure how much the basket costs every quarter, or every year, to know how much more expensive (or not) the same goods and services are.
It should be no surprise that the “basket” of a typical household changes. Twenty years ago, smartphones were perhaps not as common, and ten years from now, subscriptions to the latest AI models will be ubiquitous. A basket based on older consumption data would overweight items that matter less today and underweight items that now take up a larger share of household budgets. This means that we need a change not only in the items in the basket, but also the weight of each item.
For example, as a household grows richer, the share of expenditure on food declines, as the household is now spending on other goods— education, healthcare, digital goods, etc. A good statistical system, therefore, has to continuously update the consumption basket to ensure that it correctly represents what households consume.
The new CPI basket reflects these changes. For example, in the new basket, the weight of food has declined modestly, while items such as healthcare, education, transport, communication, and recreation have gained weight. There have also been methodological improvements in price collection, including greater use of digital systems and updated sampling strategies.
These are welcome changes. They ensure that when the RBI targets CPI inflation, it will be targeting the lived consumption experience of households. The changes also ensure that payments such as dearness allowance or pensions, that are linked to the CPI, will be more closely based on the true consumption baskets of households.
Also read: Why tax disability pension of soldiers who keep serving? It punishes loyalty
The importance of comparability
Updating the consumption basket is an important step. But it does create another problem—how do we compare inflation across time? If the basket of yesterday is different from the basket of today, the change in its price does not convey much information. Comparing the price of the “same” basket over time is fundamental to economic reasoning. A new basket may help us measure inflation for a few periods from now, but it doesn’t tell us how today compares to yesterday. And given that baskets will have to be over time, how does one ensure continuity?
As of now, the ministry has given us a linking factor used to connect India’s old CPI series (base 2012=100) to the new CPI series (base 2024=100). The linking factor scales the old CPI series so that it aligns with the new one at the point of overlap. With this, one can rescale the old series to understand what it would have looked like, had we used the new consumption basket. But providing a single linking factor has some disadvantages.
It assumes that the relationship between the old and new series in that year is representative of earlier years, which may not always be true. If the previous year was an abnormal year, then a one-year linking may not be representative of previous years. Further, since the weights of the components have changed (as food now has a smaller weight in the basket), a single linking factor cannot account for the changing influence of different groups on inflation trends.
There are ways to solve for this. For example, one could publish both series for a number of years. One could rebuild the old series entirely, possibly for a decade, instead of just scaling it. In doing so, it will be in line with best practices in other jurisdictions. For example, in the UK, the Office for National Statistics (ONS) updates the CPI basket frequently and revises weights using detailed expenditure data. It publishes continuous historical series and detailed methodological notes that convey how historical comparability is preserved.
The Australian Bureau of Statistics (ABS) has moved toward more frequent reweighting, increasingly using national accounts data to ensure that the CPI reflects current spending patterns. The ABS maintains historical series and links them carefully, with adequate explanations on the impact of revisions. The Ministry of Finance should institutionalise a system to revise more frequently in smaller steps so that large discontinuities can be avoided.
The new CPI basket is a step in the right direction. It recognises that India of today is not the India of a decade ago. While we update the series, preserving continuity and comparability is equally important.
Renuka Sane is managing director at TrustBridge, which works on improving the rule of law for better economic outcomes for India. She tweets @resanering. Views are personal.
(Edited by Saptak Datta)