How the rate is built.

The headline Consumer Price Index is computed by the National Statistical Office from a basket whose weights derive from the Consumer Expenditure Survey of 2011–12. Those weights are a national average across rural and urban households at every income level. They cannot describe an urban household in 2024 whose budget allocates 30% to dining, transport and lifestyle services rather than 46% to food alone. The official rate is correct in aggregate and wrong for almost everyone individually.

Thirteen behavioural questions establish a tier (T1–T5), city class, dietary pattern, commute mode, AC usage, healthcare load, education load, and travel frequency. The tier sets the broad allocation envelope. The remaining answers resolve sub-baskets within each category. The final basket is normalised to sum to one.

Every basket item carries a confidence rating.

• HIGH — government-notified prices or directly observed commodities. Used as-is.
• MEDIUM — CPI sub-series with defined outlet sampling. Used directionally; actual may vary ±20%.
• LOW — proxied from adjacent series, or from outlet samples known to skew lower than urban retail. Presented as a floor estimate.

Private school fees in CPI are sampled from a basket that includes government schools and skews low; actual private fee inflation runs 1.5–2× the reported figure. Service prices are systematically under-sampled. Alcohol prices are state-driven; the national series understates volatility for any individual state. Fresh produce rates carry ±30% seasonal variance not captured by a YoY snapshot.

Not a financial advisor. Not a tax tool. Not a budgeting app. Not a chatbot. Not personalised over time. No survey responses are stored.