Lies, damn lies and statistics

The WPI is an index of only wholesale prices not retail, which are invariably higher.
A fall in a price index does not mean prices have fallen.

By Paranjoy Guha Thakurta

Who believes the government when it states on a particular Friday that the official wholesale price index (WPI) has gone up by, say, 6.7 per cent? Certainly not the home-maker who has to shell out much more in the marketplace for virtually each and every article of daily necessity.
The credibility gap between the authorities put out and what is perceived by the person on the street widens for more than one reason. The WPI is an index of only wholesale prices not retail prices that are invariably higher. Then, a decline in a price index does not mean that prices have come down, merely that the rate at which prices are going up has slowed down or decelerated. Moreover, it can be argued that the way in which the WPI itself has been constructed is flawed and does not reflect changes in the cost of living that are brought about by inflation.

According to the ‘Economic Survey’ released by the Department of Economic Affairs in the Ministry of Finance, the WPI “which is available on a weekly basis, continues to be the most popular measure of headline inflation in India…(The) WPI is an economy-wide index covering 435 commodities. Weights of the commodities are derived based on the value of quantities traded in the domestic market. It is, therefore, the most comprehensive measure of economy-wide inflation available with high frequency…”

There are four consumer prices indices – CPI-IW for industrial workers, CPI-UNME for urban non-manual employees, CPI-AL for agricultural labourers and CPI-RL for rural (non-agricultural) labourers – that are specific to different groups of consumers and the commodity baskets for these indices are based on group-specific consumer expenditure surveys and the weights given to each commodity is proportionate to its expenditure. The CPI-IW is used for indexation of wages of employees working in the government or in the organized sector. The CPIs are supposed to be compiled according to standards and guidelines set by the International Labour Organization.
The WPI and the CPI-IW are compiled by the Central Statistical Organization in the Ministry
of Statistics & Programme Implementation whereas the other indices are put out on the basis of data gathered by another wing of the Union government, the Ministry of Labour and Employment. The two sets of figures in the WPI and the CPIs often do not match.

For instance, the WPI had reached a peak of 6.6 per cent in March 2007, then declined gradually to a low of 3.6 per cent in December 2007 before going up again. For the week ended March 1, 2008, the WPI went up by 5.1 per cent against 6.5 per cent a year earlier. During the week ended March 15, however, the WPI had surged to 6.7 per cent, the highest over the past thirteen months. In December, when the WPI was up by 5.6 per cent, various CPIs were registering a far faster rate of inflation – the CPI-AL had risen by nearly 9 per cent while the CPI-UNME had increased by almost 7 per cent. Annual inflation as measured by the CPI-IW for the month of January 2008 stood at 5.5 per cent against 6.7 per cent in January 2007.

The CPI-UNME had increased by 5.2 per cent in February 2008 over a year. These statistics do not reveal the fact that inflation is not uniform across the country. For instance, the CPI-UNME is compiled for 59 selected urban centres and in February 2008, annual inflation was as low as 1.2 per cent in Shimla whereas it was as high as 12.4 per cent in Shillong.

When average figures of price indices are disclosed in newspapers, the figures are often not disaggregated. But once the details are revealed, the true picture comes out. For example, during the calendar year 2006, the WPI reveals that the prices of pulses went up by more than 25 per cent, wheat prices increased by 15 per cent, prices of fruit by 12 per cent and milk by over 6 per cent. A look at the details of the CPI-UNME for February reveals that while the price of cereals went up by an average of 5.8 per cent, prices of edible oils and fats jumped by a huge 18.7 per cent and so did the prices of milk and milk products (by 10.5 per cent).

There is a fundamental issue about official statistics of the WPI and the CPIs that needs to be highlighted. Whereas the CPI-UNME, for instance, includes in the index items of expenditure such as medical care, education, recreation, transport and communications, personal care and household articles, such items are excluded from the WPI. Thus, a question arises as to whether the changes in the WPI truly reflect the shifts in standards of living of ordinary citizens.

In June 2007, two economists associated with the Multi-Commodity Exchange prepared a research paper pointing out serious flaws in the method of calculating inflation. The economists, V. Shunmugam and D.G. Prasad, writing in their personal capacity, called for changes in the way in which the government calculates inflation in keeping with the methodologies being used in developed economies. They first point out that consumer prices are most commonly used to calculate inflation, not wholesale prices. Countries such as the US, UK, Japan, France, Canada, Singapore and China all use consumer prices and the composition of the commodity basket used to arrive at inflation figures are reviewed every 4-5 years to factor in changes in consumption patterns.
According to the economists, more than 100 of the 435 commodities in the current WPI – that was constituted in 1993 -- are no longer important to be included in the index. For instance, the WPI includes coarse grain that is used mainly to feed cattle. Conversely, a host of items that are consumed regularly are excluded, notably education and health-care.

An important reason why the WPI is used the most in India for calculating inflation is simply because it is made available every Friday whereas the data contained in the CPIs are usually two months old. The challenge before the government is to revamp systems that would ensure that data on consumer prices are gathered and disseminated more expeditiously.