How not to "control" inflation

India is a fascinating country on attitudes to inflation. On one hand, there is a deeply ingrained belief that inflation hurts the poor most, and politicians are hawkish about inflation. But concern about inflation has not manifested itself in good measurement of inflation. Inflation measurement is bad - and has been bad for the longest time without any effort at fixing it. And, in the short run, there is lassitude in doing anything about inflation. The `Taylor Principle' is not understood, and there is little by way of institutional mechanism or institutional commitment to control inflation.

When inflation does spike up, the first response in socialist India is to look for ways in which government can manipulate prices of a few products.

Suppose there are N commodities that go into an inflation measure. There are two sets of N numbers: the price rise p_i for all commodities and their weights in the index w_i. There will always be a distribution of different price rises; i.e. all values p_i will not be the same. So if you sort p_i you will always be able to identify 10 commodities with the highest price rise. And if you sort p_i*w_i, you will always be able to identify the 10 commodities which have `contributed the most to inflation'. Statements then get made of the form `there is no real problem of inflation; it's just that there are just 10 weird commodities which account for xx% of the overall inflation; we have to solve these localised problems and we're fine'.

In socialist India, the government would then go after these 10 commodities with price controls, imports, release of inventory from government warehouses, ban on futures trading, etc.

There is a need to completely shift away from this mentality. Data processing, as described above, delivers no insight into where inflation is coming from. It is just an arithmetic fact that some commodities will have higher price rises than others. The economy suffers greatly from distortions when the government goes after the top 10 commodities every now and then in this fashion.

Ila Patnaik had a great article in Indian Express on 11th November where she argues that in a market economy, inflation should be seen as a phenomenon of macroeconomics, requiring a response from macro policy; arguing that it is time for the government to get out of manipulating prices of individual commodities.

This is an important part of the case for an inflation targeting central bank in India. Monetary policy based on inflation targeting is good in its own right; it is monetary policy which stabilises the boom and bust of GDP growth. But in India, it will address the goal of politicians who like low inflation, and remove the pressures that come up from time to time for government to get hands-on with manipulation of prices of specific commodities.

Update (10 Feb 2007): Indian Express has an editorial on difficulties of inflation measurement.