Inflation has been raging at 10%-plus for several weeks now. To what extent is the Reserve Bank of India (RBI) to blame? This is a tough question, partly because the surge in inflation is global this year. Indeed, some are of the view that inflation has risen due to RBI’s lack of adequate autonomy. It was overruled on many decisions (ban on participatory notes, external commercial borrowing limits etc.) by the finance ministry, until the rupee rose sharply last April, when the ministry backtracked.
Rajeev Malik, executive director, JPMorgan Chase Bank, Singapore, has often defended RBI. Mint columnist V. Anantha Nageswaran recently stressed that RBI is being made the convenient scapegoat for the finance ministry’s expansionist policies, such as directives to banks (Mint, 8 July). There is some truth to this view.
Granted these considerations, RBI can still be criticized for not pre-emptively reacting to rising crude oil prices over the last two years. This article evaluates another important aspect of current (RBI) policy — the stated 5-5.5% year-on-year WPI (Wholesale Price Index) inflation target which RBI somewhat relies upon. Evidence shows that this measure is technically flawed.
Many have stressed repeatedly (including me) that CPI (Consumer Price Index) and not WPI inflation should be targeted. CPI corresponds to what consumers buy, and captures the hardship due to food price inflation better. Nothing new. Nevertheless, it should be noted that the WPI focus of the finance ministry and RBI has not been expedient, but consistent. Late last year when WPI inflation was well below 4%, and also below CPI inflation, policymakers said inflation was in the comfort zone. But with WPI inflation well above 10%, neither the ministry nor RBI has downplayed the situation by pointing to lower CPI inflation. To their credit, they have admitted to a serious problem.
The key problem with the year-on-year (y-o-y) measure of inflation (whether WPI or CPI) for India is that it is too volatile to capture underlying inflation. The price level in India is subject to huge shocks. Hence, the inflation rate also swings a lot, rising too much in the month (or week) when the price level jumps and then falling in the same month the following year. This is the well-known, much-talked-about, and yet inadequately understood base effect.
The magnitude of the base effect can be observed from the chart. The intra-year swings in inflation are typically 300-400 basis points. Indeed, for the last fiscal year, the inflation range (high minus low) is the same as the mean: 4.7%. The month in which inflation peaks in a (fiscal) year is likely to be followed by a trough for the next year, e.g., a low of 3.9 % in August 2003, followed by a high of 8.7% in August 2004 etc. Statistical tests confirm there is a whopping base effect. Things are not very different when we look at CPI.
These conclusions are based on direct calculations, not rocket science. Yet, surprisingly, those advocating a y-o-y inflation target have not paid attention to these huge intra-year swings that are staring us in the face. If RBI took the 5-5.5% inflation target seriously, then it should have been moving interest rates up and down every year by 300-400 basis points for the last five years. As policy goes, this would have been a disaster.
I am in broad agreement with inflation targeting, philosophically. However, those explicitly for an inflation rule (Raghuram Rajan and others), must concretely specify at what frequency this should be done, and accept the interest rate swings that would result. The devil is in the detail; very much so.
When there is a strong base effect, then a three-year (moving) average of y–o-y inflation is a sound measure of underlying inflation (Vivek Moorthy, Economic & Political Weekly, 6-13 October, 2001). Even this statistic has generally moved about 100 basis points intra-year, and surprisingly reached a low of 4.2% in August 2007. Other smoothed inflation measures need to be explored in connection with inflation targeting.
However, suppose that RBI tries to keep its interest rate corridor 200-300 basis points above three- year inflation. (The reverse repo rate is the lower bound of the corridor and the repo rate the upper bound). Then the reverse repo rate, based on the three-year average should seldom have been below 6%, and most of the time above 7% for the last five years. And the reverse repo rate should be 9% now since the three-year WPI inflation is 7.1%.
In short, although the current 11% inflation may fall next July because of the current high base, still monetary policy all along should have been, and now needs to be, much tighter. Way to go.
This article is not meant to suggest that RBI strictly target three- year inflation. A pre-emptive approach to inflation is far better than an explicit inflation target. (“Too late to tackle inflation”, Vivek Moorthy, Mint, 3 June). Among Asian countries, Thailand and Malaysia, which followed a 3% inflation rule, are now suffering the most as their retail energy prices have been decontrolled.
One suggested remedy to the huge intra-year inflation swings is the use of ostensibly timely, intra- year, seasonally adjusted data. However, seasonal adjustment for India does not resolve the huge inflation swings—a technical working paper is needed to show this.
Vivek Moorthy is professor of economics at IIM, Bangalore. Comment at firstname.lastname@example.org