Telangana illustrates the scale of the shift. Its inflation print of 4.92 per cent in January 2026 — now the highest in the country — marks a sharp jump from its 13th position in December 2025 under the 2012 series. Kerala has slipped from the top spot to second at 3.67 per cent, Tamil Nadu has risen from sixth to third at 3.36 per cent, and Rajasthan has surged from 26th to fourth at 3.17 per cent.

 


At the bottom, Manipur recorded the lowest inflation of 0.12 per cent in January, compared with its 11th position in December, when inflation stood at 2.08 per cent.

 


What has changed is not just the ranks, but the height of the peaks and depth of the troughs. The highest state inflation rate itself is lower than before, falling to 4.92 per cent (Telangana) from 9.49 per cent (Kerala) in December 2025. And — crucially — no state is in deflation in January 2026. That is a sharp change from December 2025, when nine states were in negative territory.

 


The Expert Group’s report provides the rationale for this re-ordering. First, state-level inflation is now being calculated using new weights derived from the Household Consumption Expenditure Survey (HCES) 2023-24, which capture very different consumption patterns by state.

 


States such as Telangana, Tamil Nadu and Karnataka now have much higher shares of services, housing, telecom, fuel and online spending than a decade ago, and those categories have also seen relatively faster price increases. With the revised CPI assigning greater weight to these faster-rising categories, the headline inflation numbers for these states have correspondingly moved higher.

 


Second, the market frame has been rebuilt. The new CPI expands urban coverage in large, fast-urbanising states and adds many more markets in peri-urban and smaller towns. That matters for states such as Telangana and Rajasthan, where earlier CPI rounds were effectively “anchored” in a smaller set of cities and markets.

 


Third, the series has been aligned to the Classification of Individual Consumption According to Purpose (COICOP) 2018, and the item structure has been re-engineered. Some items have been split (for example, different types of clothing, education levels and telecom services), others clubbed, and many “other” categories cleaned up and reallocated.

 


P C Mohanan, member and former acting chairman, National Statistical Commission, noted that at the individual state level, the effective sample size per item becomes much smaller and, because of that thinness, item-level and very fine subgroup indices at the state level can be noisy.

 


“So, that is where some of the outliers and abnormality is likely to come. Because the size of the sample at the state level for individual items may not be very large,” he reckoned.

 


He pointed to the case of Kerala, where inflation almost touched 10 per cent in December and is now under 4 per cent. “This anomaly is because the weighing diagram at the state level may not be as stable or as robust as the national weighting diagram,” he added.