Opinion New CPI overstates inflation

Rs 116,000 crore in 2009-10. An idea of how staggeringly large this shortfall is can be gleaned from the fact that it exceeded the entire higher education budget of Rs 16,000 crore by more than seven times,or more than 600 per cent! If that doesn’t catch somebody’s attention,perhaps the fact that this shortfall was more than 11 per cent of the Central government’s total revenue in the same year should have been a telltale sign. Or the fact that if this was the “corruption” shortfall in one ministry alone,and if one adds the corruption money alleged by the CAG for the other Sibal ministry (telecom) of Rs 176,000 crore,one arrives at the conclusion that corruption in two ministries alone was more than 5 per cent of GDP in 2009-10.

The CAG is an honourable institution,but its honour is beginning to wear thin. That it does not even begin to do the most basic checks on its mountain-top discoveries is revealed by the latest embarrassing stumble. It turns out that the CAG interpreted the HRD budget statement of lakhs as crores. That is it. A 100 times error,but how does that matter when the larger purpose is to fight corruption?

This is an extremely sorry state of affairs. Numbers can be confusing,and we all make mistakes. But I have yet to hear about a report being tabled in Parliament,alleging large-scale corruption,and done so without the most basic of checks.

Of late,there have been several data “mistakes”,and unfortunately,in ministries whose job it is to produce data that policymakers rely on to make policy. Data that allows for proper functioning,and data that negates the likelihood of the blind leading the blind. The important data on India’s industrial production has suffered from so much questioning that its pronouncements now hold as much sway as missives from the CAG.

There is only one other monthly statistic that has as much importance as the index of industrial production,or IIP. This is the index of inflation. One of the most important policy parameters in any modern economy is the rate of inflation. The RBI governor,Duvvuri Subbarao,while announcing a surprise 50 basis point repo rate cut a few weeks ago,emphasised that this cut might be the last because both growth and inflation were expected to surge in the new fiscal year. Growth was to rise from the depths of 6.1 per cent observed in October-December 2011 to 7.5 per cent,and inflation was expecte d to surge back to the high 8-plus per cent levels observed over the last four years. The very next day,the new CPI series revealed that Subbarao was almost as accurate a forecaster as the venerable Alan Greenspan at his peak. This new CPI series registered close to double-digit inflation in March,9.5 per cent. There go any more rate cuts,though the larger question remains as to how the RBI governor will be able to achieve an acceleration in GDP growth of 1.4 per cent on the basis of more rate hikes to come!

Some two weeks ago,I had advanced the hypothesis that the new CPI series may be,just may be,comparable to the wonky IIP data. I showed that this new series was systematically registering excess inflation of over 2 per cent in each month of 2012. In other words,the “true” CPI was likely to be closer to 7 per cent than 9.5 per cent. Undoubtedly,still a worrisome and high inflation number,and one whose causes remain somewhat under-investigated. Nevertheless,7-odd per cent is consistent with a decline in inflation from more than the 9-plus per cent levels registered in each of the three years 2008-10.

This article met with considerable scepticism. But constructive suggestions were offered. Most importantly,the critics and the sceptics pointed that the new series had new updated 2004-05 consumption pattern,many new centres from where data

were gathered,and some new items of consumption,for example,computers,laptops,mobile phones,etc. So,given all these differences,it was not surprising that the new,technologically improved CPI should show a different rate of inflation. True. But different,and higher,and higher by more than two percentage points each month? Different weights,

etc can be expected to have a one-time effect,but a persistent inflation effect was likely as uncommon as a ministry spending seven times more than its allocation.

Of the three determinants,the last two are expected to show lower,not higher,inflation. New centres normally mean that the poorer regions are a higher percentage of the total,and poorer regions tend to have both lower price levels and the same or lower inflation. Ditto arguments for newer goods. Such goods have a higher technology content,and technological progress tends to generate lower than average inflation. Thus,one is left with a change in the weights as a major explanatory for differences in inflation rates between the old CPI series and the new series.

The old and new rural CPI series can be compared to reveal contributions of each determinant. The new series identifies 21 broad categories with 2004-05 expenditure weights

derived from NSS household expenditure data. NSS expenditures for these 21 items are averaged for

the years 1983 and 1993-94 to approximate the old CPI expenditure weights of 1986-87. These weights allow us to identify the contribution of different factors.

The table details the results for the three months,January-March 2012. The contribution of weights is about 0.6 percentage points. However,more than half of this contribution resides in two ambiguous categories accounting for only 5.6 per cent of the total basket,household requisites (4.5 per cent) and others (1.1 per cent). Net of these items,that is,for 94.4 per cent of the basket,the average difference in inflation due to differences in weights is only 0.27 percentage points. These two items have averaged an annual inflation rate of 20.4 per cent in the last three months,well above the average excluding these items of 7.5 per cent.

It is possible that Subbarao’s prescient observation on resurgent inflation was based on prior knowledge of the new CPI data. If so,it underlies the fact that data affects policy,as it should. But bad data affecting policy is not something the Indian economy can afford as it lurches backward at ever slower rates of growth.

The writer is chairman of Oxus Investments,an emerging market advisory firm