India’s new inflation basket changes what a food shock means

India’s new consumer price index is a better starting point for monetary policy because it reflects more recent spending patterns and measures prices that the previous basket missed. The first release on the 2024 base puts January inflation at 2.75%. The last December estimate on the 2012 base was 1.33%. Treating the difference as an acceleration of 1.42 percentage points within a single unchanged basket would overstate what the two headlines establish. Both the month and the measurement system changed. The useful question is how much price pressure has changed for comparable items, and how the new basket will translate future shocks into the headline the RBI targets. MoSPI’s first CPI release on the 2024 base · MoSPI’s December CPI release · Reuters, 12 February

The decline in the food weight has two distinct components

The Ministry of Statistics and Programme Implementation updates expenditure weights using the 2023–24 Household Consumption Expenditure Survey. It also adopts the international COICOP 2018 classification, dividing consumption into 12 categories. The published weight of food and beverages is 36.75%, compared with 45.86% under the old classification. Part of the 9.11 percentage point difference comes from classifying expenditure differently. For example, restaurant and accommodation services now have a separate division. To isolate the effect of the updated spending pattern, hold the classification fixed. MoSPI’s first CPI release on the 2024 base

The ministry provides both bridges. Under the old classification, applying the updated weights takes food and beverages from 45.86% to 40.10%, a 5.76 percentage point decline. Under the new classification, applying the older expenditure weights would put the category at around 42.62%, compared with 36.75% using the updated survey. Both comparisons point to a meaningful reduction in the food share under consistent definitions. The total difference between the published headline weights combines that underlying change with the classification change. The distinction matters when judging whether families have shifted consumption away from food or whether a type of expenditure was simply moved to another division. MoSPI’s first CPI release on the 2024 base

To see why it matters for policy, suppose prices for the items in the old food and beverages group rise 10% together while all other prices stay flat, starting from base-period relative prices. Applying weights of 45.86% and 40.10% to that same shock gives contributions of about 4.59 and 4.01 percentage points, respectively. The example holds the category and price shock fixed and changes only the weight. The approximate difference is 0.58 percentage points. It is an illustration rather than a decomposition of January inflation. Actual contributions depend on the detailed indices, relative prices and which parts of the basket are moving. MoSPI’s first CPI release on the 2024 base

Rural housing changes the coverage of the index

The old CPI had no separately measured rural housing group. The new system includes rural house rent, with a weight of about 2.46% in the rural basket. That figure is smaller than the 11.76% for the entire rural division covering housing, water, electricity, gas and other fuels. The larger number should not be described as rent. The distinction is material if a region’s utility tariffs move while actual rents remain unchanged. The new release puts January housing inflation at 2.05% nationally, with rural and urban readings of 2.39% and 1.92%. These are observations under the new measurement system, which provides a way to follow the two segments separately. MoSPI’s first CPI release on the 2024 base

A national average is useful for a single monetary policy rate, but spending shares vary substantially by household. A low-income family that buys much of its food and has little scope to substitute when prices rise may experience a much larger squeeze than the revised headline suggests. A renting household in a fast-growing town can face an increase where the national rent index moves slowly. A change in CPI weights therefore changes the aggregate measurement of purchasing power. It does not redistribute income across these households or replace the need for separate evidence on food affordability, wages and rent burdens. That matters if the index is used to assess welfare payments as well as interest rates. MoSPI’s first CPI release on the 2024 base

Coverage also expands beyond housing. The number of weighted items rises from 299 to 358, and the ministry now uses online prices for some purchases alongside its market surveys. For categories such as air travel or streaming subscriptions, this can give a better view of how the service is actually bought. It also changes the measurement questions. A promotional fare can be unavailable for most dates, and a subscription can change its features along with its price. A meaningful comparison needs stable specifications, collection rules that reflect how consumers buy and an explanation of how temporary discounts and unavailable prices are handled. MoSPI’s first CPI release on the 2024 base

The overlap can help, if it is used for the right purpose

Both CPI vintages are available for 2025 at the general level. The ministry uses the ratio of the average new to the average old index in that overlap to rescale earlier all-India series. Multiplying an old index by a constant preserves growth rates within the portion transformed that way. It does not recreate what the historical inflation rate would have been if today’s basket and classification had been in use. The ministry limits direct linkage to the general index because the classifications differ. An analyst estimating food pass-through or the persistence of services inflation should avoid mechanically joining category histories across the break. MoSPI’s first CPI release on the 2024 base

There is a reasonable objection to putting too much emphasis on the break. The RBI still has to take decisions in real time, and a new index can be operationally valuable from its first release. A reading of 2.75% under the new definition is a useful estimate for January. The mistake would be to infer all of the change in underlying pressure from the jump relative to the last old-series headline. The overlap and the detailed items provide an alternative. Compare movements within the new basket through the recent months, check their breadth and use genuinely comparable price series for areas where the aggregate groups were recast. MoSPI’s first CPI release on the 2024 base · Reuters, 12 February

The policy consequences run in both directions. A smaller food weight can reduce the headline response to a temporary vegetable shortage. It can also make slow-moving categories more influential when their prices rise persistently. The RBI’s usual interest-rate channel works on spending and financial conditions with a lag. It has limited ability to increase the current supply of vegetables or rented homes. A more representative headline makes the target more useful, while the policy response still depends on the source and persistence of the pressure. Measures of inflation excluding volatile categories should be reconstructed from clearly documented definitions before their old thresholds are reused. MoSPI’s first CPI release on the 2024 base

January’s rural and urban headline rates, 2.73% and 2.77%, are close. Their food inflation rates, 1.96% and 2.44%, are further apart. This is a reminder that two similar headline numbers can conceal different sources of pressure. A useful release would let readers trace the difference to the categories and items that contributed most, rather than asking them to infer household conditions from the national average. A district with rising staples prices and a city with rising rents can call for different responses from local policymakers even when the national policy rate remains appropriate. MoSPI’s first CPI release on the 2024 base

What to look for in the next readings

The case for treating January as evidence of wider inflation would strengthen if price increases persist across several large categories, with independent support from wage, rent and selling-price data. If the rise is concentrated in a few food items or reflects the month used as a comparison, a broader tightening response would be less well supported. MoSPI can help by keeping both 2025 vintages accessible and publishing detailed bridges with clear category definitions. Researchers should document which version of the data they use when they assess RBI decisions or estimate inflation persistence. The new basket will earn confidence by making those comparisons more informative and reproducible. MoSPI’s first CPI release on the 2024 base