The WPI is an average of the current level of prices. The year-on-year change in it tells us how much higher or lower prices are than at the same time last year. But if year-on-year inflation according to the WPI rises, it does not necessarily mean that prices are rising faster; it could also mean that the rise in prices slowed down last year. Conversely, if year-on-year inflation according to the WPI falls, it may be because prices began rising faster last year. In other words, the information that changes in WPI inflation from week to week conveys is ambiguous; in particular, it tells us nothing about whether prices are rising, and how fast.

How, then, do we know whether prices are rising or falling? Each of us has an approximate estimate of inflation from whether we are spending more or less. If our expenditure on things we buy from week to week goes up, we know that prices are going up, whatever the WPI says. In fact, each of us could work out his own cost-of-living index by calculating the changes in the prices of things we buy regularly and multiplying them by their share in our expenditure.

If that is too much trouble, we can get the same information from the WPI and the cost-of-living indexes, which the government publishes with greater delay. But week-to-week changes are not very reliable; prices keep fluctuating, and we should ignore accidental or random changes.

So we can compare prices over the past two, three, four and more weeks; the larger the number of such comparisons that shows a rise, the greater the chance that prices are rising. But just the information that prices have been rising or falling is not very interesting; we would like to know how fast they are changing. It is possible to work that out from the comparison of any two successive WPI figures; the percentage change between the figures should be annualized.

If we did that, we would find these annualized figures going up and down like crazy; rates of change over short periods are extremely unstable. That is one reason why the government takes year-on-year changes. The other is that some prices go up and down seasonally. For example, prices fall when a harvest comes in; then they go up. Year-on-year changes are not affected by seasonal effects – or are less affected than other figures.

Should we, then, abandon our search for the correct figure and settle for the government’s figure of pseudo-inflation? We do not have to be so defeatist. All economic magnitudes are approximate. When the government publishes figures of gross domestic product every quarter, they are not based on any measurements. From a few available figures, it calculates output indexes for various industries, obtains rates of change from them, and multiplies them by the figures for the last quarter or year to get current figures. When the Reserve Bank publishes figures of the balance of payments, it invariably puts in an item that it calls errors and omissions to acknowledge that its figures are not consistent. In balance of payments estimates, it puts figures of imports and exports based on the foreign exchange the banks have paid and received for them. The ministry of foreign trade estimates its own figures based on declarations made to the customs department. The two estimates never match. They ideally should; in other countries they do. But in our country, the government propagates the homegrown philosophy of ‘chalta hai’ (it will do).

So we too should take estimates of inflation over different periods and make up our own minds. If I do so, I come to the conclusion that inflation has gone up alarmingly after January, and is bound to be reflected in the year-on-year index later in the year. People are paying higher prices now, but official figures will show that only in the second half of the year. Pay now and feel bad later.

The author is Consultant Editor of Businessworld.
This email address is being protected from spam bots, you need Javascript enabled to view it

(Businessworld issue 27 May-2 June 2008)