India’s CPI Reform Still Misses the House Rent Reset Problem

India’s consumer price index (CPI) overhaul, formalised in the January 2026 Expert Group report, represents the most consequential modernisation of inflation measurement in more than a decade. The housing component — long debated — has been substantially revised. Rural rents are now included alongside urban rents. Employer-provided accommodation has been excluded to avoid unnecessary distortions. Dwelling-type weights will be derived from Census 2011. Dissemination has been expanded across rural, urban and combined segments. These are meaningful improvements that strengthen transparency and international comparability.

Yet despite these reforms, one structural feature of rental markets continues to be ignored in our computation of the housing index: rent stickiness embedded in housing leases.

Most Indian leases include predetermined escalation clauses — typically 5-10% annually. When tenants continue, rents rise mechanically according to these pre-agreed escalation terms. When tenants vacate, rents reset to prevailing demand–supply conditions. That reset is what captures the real inflationary or disinflationary pressures in the housing market.

The divergence between house rent changes faced by continuing and new tenants has a real and significant impact on how we measure house price inflation. An index constructed from a fixed sample of dwellings is likely to contain a high share of continuing tenants. Changes in the index, therefore, are more likely driven by contractual escalations than by real changes in housing market conditions. In short, the index measures the average rent paid by incumbent tenants rather than the marginal rent faced by new entrants.  

A widely cited media report in the not-so-distant past described “the curious case of rising rents and falling housing inflation,” observing that CPI housing inflation “does not align with the real rise in rentals seen on the ground.” When brokers and tenants report double-digit rent increases, but official inflation appears subdued, the credibility of the inflation signal gets questioned.

International evidence reinforces the concern. Academic researchers have shown that US shelter indices systematically underrepresent new tenants and overrepresent sitting tenants. They estimate that survey-based rent measures can lag market rent indices by close to one year. During the Global Financial Crisis, this lag was economically meaningful. Official inflation was estimated to be overstated by between 1.7 and 4.1 percentage points in 2008-09 because shelter inflation declined far more slowly than market rents. Monetary easing therefore occurred later than underlying housing conditions warranted.

India’s revised CPI addresses several earlier weaknesses. But by ignoring a key feature in rental markets, it continues to capture the average rent paid by incumbent tenants rather than the marginal rent faced by new entrants. The resultant lag in capturing real rental inflation could impact our monetary policy setting significantly.  Housing carries significant weight in the CPI basket. If housing inflation lags market dynamics, headline inflation will also lag at cyclical turning points. Policy tightening may be delayed during rapid rent increases, and easing may be postponed when markets soften. Over time, such lags can influence expectations formation and the credibility of the inflation target itself.

By explicitly identifying tenant changes within the sample and separating rent resets from continuing contracts, we could mitigate some of this lag. Statistical models could infer current market rent for dwellings whose leases have not turned over, using nearby reset transactions as benchmarks. The index would then combine observed reset rents with model-implied market rents. In fact, such an index is already being published for the commercial office rentals (bit.ly/4rjy7CI). Extending this to housing is challenging but feasible, and it would give policymakers a cleaner read of the inflation renters face when they move.

A longer-term redesign would embed tenant changes directly into the sampling design. Rent resets would be tracked as primary observations. For continuing leases, market rent would be estimated using comparable properties within the same micro-market. Location-level averages would then be aggregated nationally. This approach recognises that rental markets are spatial and dynamic rather than static contractual averages.

Reforming the CPI is not merely a technical exercise. Housing carries a large weight in the consumption basket, and delays in measuring rent inflation can blur turning points – when markets are heating up or cooling down. If the next CPI series records tenant turnover and treats rent resets as the primary signal, supported by model-based estimates for dwellings that have not turned over, India’s inflation statistics will track housing cycles more faithfully. That would sharpen monetary policy signals and, equally important, strengthen public confidence that the CPI reflects the costs households actually face.

(Venkatesh Panchapagesan is Professor of Finance and Head, Real Estate Research Initiative, IIM Bangalore and Soudeep Deb is Associate Professor of Decision Sciences, IIM Bangalore. Views are personal)