Why your inflation isn't the official number

In December 2025, headline inflation was 1.3%, but your wallet probably told a different story. That’s because the Consumer Price Index (CPI) doesn’t track your household. It follows a fixed set of 358 goods and services, from rice to rent, from cinema tickets to chemotherapy, with weights that mirror the spending of an average Indian. If your spending doesn’t match those weights, your personal inflation will differ.

That same month, while food prices overall dropped -1.9% (vegetable prices crashed -18.5% and onion prices fell -48.1%), other items surged. Gold shot up 68.7%. Your doctor’s fee, captured by an index that rose from 117 to 223.6 since 2014, still climbed. If your spending tilts toward these non-food items, you feel pressure even when the headline is low. The average hides the extremes.

The CPI basket is no one’s actual shopping list. It’s built from the Household Consumption Expenditure Survey (HCES) of 2023-24, which asks thousands of families what they buy. MoSPI, the statistics office, then averages the responses to create a “representative” household. On 12 February 2026, India launched a new CPI series with 2024 as the base year, using 358 items, expanding coverage to include e-commerce prices, OTT subscriptions and telecom data plans.

The biggest shift from the old 2012 basket: food’s weight fell from 45.86% to 36.75%, a textbook case of Engel’s law. As people get richer, they spend a smaller share on food. Housing, water, electricity and fuels now account for 17.66%, transport and communication 12.41%, and health 6.1%. But these are still averages. A family paying ₹35,000 monthly rent in Mumbai might spend 40% of its budget on housing, not 17.66%. A daily-wager in Bihar spending 60% on food is equally unrepresentative. The basket is a compromise. The moment you deviate from it, your inflation diverges.

India publishes separate CPI for rural and urban areas because their spending baskets differ. In December 2025, rural inflation was 0.8% while urban inflation was 2%, a gap of 1.2 percentage points. Rural households devote a larger share to food, so they benefited more from food deflation (rural food inflation was -2.3% versus urban -1.1%). Urban households spend more on services, education and transport, which have been rising. For instance, the health index stood at 204.8, and education at 196.1, both pushing urban inflation above rural.

Even within food, the composition differs: rural families eat more cereals and pulses, while urban diets include more processed foods and restaurant meals. The basket weights reflect these differences. So when the news says “inflation fell,” it might have fallen harder for the villager than for the city dweller, or the other way around.

In December 2025, Kerala’s retail inflation hit 9.5% while neighbouring Telangana managed just 1.8%, a spread wider than the all-India headline itself (1.3%). States diverge because their consumption baskets and local conditions differ. Kerala has a long history of higher inflation, partly because of high demand for services, heavy gold consumption (gold inflation was 68.7%), and remittance-driven spending from the Gulf. Telangana’s inflation is tempered by lower food price pressures and a different mix of spending.

State-level CPI data shows that “India’s inflation” is a statistical abstraction. When you live in Thiruvananthapuram, paying for a doctor’s visit and watching gold prices soar, your inflation is closer to 9%. In Hyderabad, with its own food supply chains and lower service costs, it’s closer to 1.8%. The national number smooths these regional realities.

This is Engel’s law in action. The poor allocate a much larger share of their budget to food. When food prices rise, their true inflation is higher than the headline; when they fall, they get more relief than the average. In December 2025, with food deflation at -1.9%, a poor household might have seen their expenditure actually shrink, even as the rich faced rising tuition fees (3.4%), doctor’s fees (index up to 223.6), and petrol costs.

For the better-off, services dominate. The “miscellaneous” category, which includes health, education, transport and recreation, contributed 1.75 percentage points to the headline inflation of 1.3%, more than offsetting food’s negative contribution of -0.85 points. Meanwhile, core inflation, which strips out volatile food and fuel, was 4.6% in December 2025, far above the headline. That core figure better captures the steady price rises that salaried urban families face. So while the average Indian saw near-flat prices, the professional in a metro felt the persistent creep of fees, fuel and rent.

The headline is built on spending shares that may not match yours. If you’re sending two children to private school, tuition fees are probably a large chunk of your budget, and the tuition index hit 204.7, up from 111.9 in 2014. If you have a chronic condition, doctor’s fees and medicine (index 203) push your costs up. If you commute by car, petrol and diesel indices are at 140.2 and 189.1. And if you rent, housing inflation was 2.9% nationally, but in cities it can be much higher.

Now imagine a household that spends 30% on rent, 20% on education, 15% on health and only 20% on food, with the rest scattered across transport and other items. Using December 2025 component inflation rates: housing 2.9%, education 3.4%, health 3.4%, transport and communication 0.8%, and food -1.9%. A rough weighted average gives about 2.3%, already higher than 1.3%. Tweak the composition further towards education or health, and personal inflation easily exceeds 4%. The CPI basket weights these categories at 17.66%, 3.33%, 6.1%, 12.41% and 36.75%, a very different distribution. Your inflation is the weighted average of your actual spending, not the nation’s.

It is not wrong; it is a faithful average. The CPI methodology is sound: price collectors visit shops, hospitals, schools across India every month, tracking the same 358 items. The problem is that an average hides the spread. When onion prices crash and gold soars, the average looks moderate because the two cancel out. But no single household buys exactly the average basket.

A fixed basket also lags reality. When tomato prices double, people switch to cheaper vegetables, but the basket still tracks tomato at the old weight. The 2024 update improved timeliness by adding online prices, but it still assumes a fixed spending pattern between revisions. So the official number is essential for interest-rate decisions, but as a personal gauge, it is nearly always off.

1. Identify your three biggest spending categories. For most Indians, that’s food, transport and housing. 2. Track the component indices MoSPI releases monthly, like “CPI, Vegetables,” “CPI, Petrol,” or “CPI, Health.” They’re available on the ministry’s website. 3. Weight them by your actual spending shares, not the national weights. If you spend 40% on rent, 20% on food and 10% on petrol, use those. 4. If you are poorer, watch food and fuel; if you are urban and better-off, watch health, education and transport.

The official number tells you what’s happening to the country’s price level. To know what’s happening to yours, build your own index, with your own weights and your own items. That is the difference between a statistic and your monthly budget.