Judging the RBI by the wrong rule

Has the RBI been biased toward tight monetary policy, or is its rate-setting framework flawed?

According to several commentators, the Reserve Bank of India needs to bring about radical changes in its monetary policy conduct. At the minimum, it must change its rate-setting approach. It might even need to abandon inflation-targeting altogether. Are these claims valid?

 

Their argument, in essence, runs like this. Ever since the RBI adopted inflation targeting, real interest rates have stayed far too high, thereby undermining economic growth. This has happened because the RBI follows a flawed approach: It sets rates on the basis of past inflation, even when price pressures are already on the cusp of dissipating.

 

These are strong claims, and they deserve careful scrutiny. Let’s start with the main one: Inflation targeting has biased the RBI towards a tight monetary policy. The data says otherwise.

 

Since the RBI adopted inflation targeting in 2015, the real repo rate (nominal repo rate minus headline consumer price index-based inflation) has averaged just 1.1 per cent (see chart). This is strikingly low for a country growing as rapidly as India. It is also below the neutral rate that the RBI itself estimates at 1.5-2 per cent. And it has been too low to achieve the 4 per cent inflation target, given that inflation, since 2015, has averaged 5 per cent.

Why, then, have commentators been complaining about high real rates? That is because some periods have seen real rates exceed 2 per cent, including recently. In 2025-26, the real rate hit 3.5 per cent.

 

Could this have occurred because the RBI’s framework focuses only on past inflation, as the commentators claim? Vaishali Garga of the Boston Fed and I analysed the data thoroughly (Garga V and Sengupta, R, “Do Actions Match Words? Reassessing the Taylor Rule in an Emerging-market Context”, Federal Reserve of Boston Working Papers, August 2026). We found that the RBI does consider past inflation when setting the repo rate. But it also puts considerable weight on its inflation forecast — exactly what inflation targeting requires.

 

The problem, then, is not the framework. It is the accuracy of the forecasts. If the inflation forecast is too high, the real rate will be too high as well, and this is what happened recently. In its April 2025 monetary policy statement, the RBI’s forecast of CPI inflation for 2025-26 was 4 per cent. Actual inflation came in at 2 per cent. The RBI did not revise its forecast down to that level until the December statement. A wrong forecast, acted on in good faith, produces exactly the outcome commentators point to: Real rates that turn out, in hindsight, to have been too tight — or too loose.

 

Of course, the RBI knows this. As a result, it is extremely pragmatic. In exceptional periods, when the direction of inflation and growth is relatively more obvious, monetary policy can afford to be more aggressive. For example, when Covid hit, it was obvious that aggregate demand would collapse. So, the RBI was comfortable reducing the repo rate to very low levels. But outside such periods, the RBI has been cautious about reacting to new forecasts, keeping the repo rate in a narrow band of 5.25-7 per cent, tighter than the range of inflation (2-6.8 per cent). This makes sense. When monetary policy is partly forward-looking yet forecasts are error-prone, it is prudent to practise caution.

 

In summary, it is not correct to assert that the RBI has been biased towards tight monetary policy or that its rate-setting framework is flawed. The real task for the central bank is to improve its inflation forecasts. This is where the debate should move.

 

Forecasts will never be precise, because the future is always unknown, and it is also true that the global economic environment is unusually uncertain right now. Even so, there are many ways to reduce the size of the RBI’s forecast errors.

 

To start with, India’s macroeconomic data has well-documented weaknesses, often giving conflicting signals about how the economy is actually doing. Complicating matters further, whenever the statistical agency introduces a new series for gross domestic product and inflation, it does not release much of a back series, leaving the RBI without the historical data it needs to build accurate models of how the new data behaves.

 

Then, there is the question of household inflation expectations. Existing surveys need better language, because it may not be easy to understand what “inflation rate” means. It is also important to collect data on inflation expectations of firms. Once this is done, deeper research is needed into how inflation expectations actually form and how they feed into long-term contracts. This will also require better data on wages and rents.

 

A central bank that sets monetary policy based on forecasts is only as credible as its forecasts. That is where the real work needs to go, not in reopening the case for or against the inflation targeting framework itself.

---

The writer is associate professor of economics, IGIDR, Mumbai