India should adopt differentiated inflation targeting

The early seeds of inflation targeting as a monetary policy tool has been in vogue since the days of Paul Volcker (FED chairman) in the 1980s. Persistent double-digit inflation was the order of the day which was dysfunctional. High interest rates over 8-10 years did tame the inflation in US.

Although RBI adopted the Urjit Patel committee recommendations on inflation targeting in early 2014, RBI’s tightening had started at least 3 years before that to tame the inflation arising out of expansionary interventions undertaken to tackle 2007-08 crisis.

The Critique

Many commentators have laid blame on tight inflation targeting for fall in our growth rates and persistent unemployment. Some assumptions of our Inflation targeting need serious re-examination. The choice of CPI instead of WPI, the level of 4 per cent, equal band on both sides, and choice of single 4 per cent are some of the more serious ones. This article is about the latter two.

The central target of 4 per cent is supposedly a trade-off between protecting the interests of pensioners on the one hand and fixed income earners and debtors, investors and employment and growth on the other. Physical asset investors would prefer higher inflation (lower real interest rates) which would facilitate more investments and job creation but harm the fixed income earners and salaried class.

A percentage increase in inflation from 2 to 3 per cent is more bearable for the fixed income class than when it goes up from say 7 to 8 per cent. On the other side, the social impact when unemployment goes up from say 7 to 8 per cent are far more serious than when it goes up from say 3 to 4 per cent. The equal band on both sides of 4 per cent needs serious re-examination. Unfortunately, the 4 per cent +/- 2 per cent assumes this to be equal and linear. This is seriously debatable.

The case for differentiated and multi-level targeting 

RBI has chosen to target inflation at a central rate of 4 per cent with a tolerance of 2 per cent on both sides. Let us see the case for differentiated targeting.

 The signalling strength of the developing inflationary or deflationary pressures differs from item to item and sector. A 2-3 per cent more than normal variation in consumer durables which is a discretionary expenditure tells us more about the people’s confidence in job stability, sustaining their income levels, capacity utilisation and demand pressure. A similar variation in food and beverages (46 per cent in combined and 54 per cent in Rural basket) could well be a day-to-day occurrence in many markets. Similarly, house prices and rents in urban areas which accounts for a sizable percent have a stronger messaging than fuel or petrol.

A 5 per cent variation (say from 2 per cent to 7 per cent) in food inflation would result in a 2 per cent uptick in inflation pushing it beyond the band from central target, while a 10 per cent increase in household furniture, appliances and utensils which all are clear signs of developing overheating may not trigger any action, since they would push inflation up by less than 0.5 per cent.

Secondly, an 8 per cent inflation in food and 2 per cent in all others would yield a 5 per cent inflation, well within control. But an 8 per cent persisting inflation can push many in rural areas, where more than 50 per cent are net buyers of food, into distress and below poverty line. The distress of a person with higher weightage in his consumption basket compared to the average is more when its inflation is more than average. Of course, this is the curse of all averages. But its impact on lower strata teetering on the brink of poverty is far more than when one is well above average.  This can be partially offset by lower limits for food inflation especially those sub-items which vulnerable sections consume. 

Third, the dependence on formal credit and transmission (both per cent and speed) of control measures on different items and sectors varies vastly.   

Fourth, inflation in certain sectors like construction, transport, health, house rents get passed on through factor markets (like wages) and become very sticky while inflation in items like food and fuel do not stay sticky that long and often do not affect factor markets that strongly in short term.  

Export oriented sectors will get hit by domestic monetary action whereas the overseas prices hardly move in tandem, affecting their operations. The existing schemes of concessional finance for exports cover largely the credit for the final stage, but the changes in all previous stages which may be around 3/4th does impact them.

A single size fits all inflation target may be more acceptable where inequality is low or where the minimum income of people covers most essential items so the risk of slipping back into poverty or distress is low. But in low-income societies with high inequities, it might seem callous. Average inflation levels may be alright for financial markets which can switch amongst instruments or from one market to another at the click of mouse but inappropriate for the real economy with stickier consumption habits.  

It would be more purposeful to have lower 3-4 per cent for consumer durables and discretionary items before pressing control actions. Within food, the staples which poor depend on can be targeted at 4 per cent upper limit to kick start action including some agreement with the government on MSPs. For fuel and lighting it can be 8 per cent given its non-stickiness. Exporters would need to be compensated for prior stage interest increases.      

In the minimum it would serve well to observe the targets and band for each category level in CPI individually so that undue pressures do not build up without adequate action losing vital time.

The writer is author of Making Growth Happen in India (sage)