A Price Index to Hide Inflation: Financial Express 19th February 2011

The new CPI is another addition to the series of changes one is witnessing in the data sets on the Indian economy. We have had a new WPI series followed by GDP, which chose 2004-05 as base year. The IIP is also being reworked on this basis, which would harmonise the base years for most critical economic variables. This move is definitely pragmatic as comparisons are possible across indicators without the limitation of having to qualify the base years.

The new CPI in its new form with 2010 as base looks at urban, rural and a combination of the two, and would probably replace the current sets on industrial workers, agricultural labourers and rural labourers, although it has been stated that the status quo would not change right now. While it may still be difficult to map the same, it could be assumed that there is some correlation between these groups. However, the major issue here is the choice of base year.

This index uses 2010 as the base year. Is it a fair assumption? Usually a base year should be a normal year where there was normal economic activity with few extraneous distortions. But, looking at 2010, it would probably not fit the billing. This was probably one of the most inflationary periods of our lives. Further, the global economy is still recovering from the financial crisis, which means that activity was not normal. Therefore, there will be a wrong start at the beginning of the index.

A high base year for the CPI in 2010 means that the future rates of changes in prices will tend to be lower as we move along. It is not surprising that the rates shown for January are of a lower order of 7% for rural folks, 4% for urban and 6% for all. The number of 4% will definitely not go down well with all those who will finally (when that for industrial workers is replaced) have their salaries adjusted for this kind of inflation, since the actual cost of living has gone up more significantly than this number suggests. In fact, looking at this number, one will get an impression that inflation was never a major issue to begin with, as January was the time when we had the highest number in food inflation. The WPI monthly index showed price increase of over 15%, while the CPI tells us that it was not really significant. Clearly, there is something wrong in the representation of prices.

A serious anomaly is that the base year has been taken as 2010, even though the weights are based on consumption patterns as per the NSS study for 2004-05. This being the case, they could have used 2004-05 as the base year rather than 2010.

Further, the new CPI has changed the weights of several components of the index quite drastically. Food items have a lower weight for urban folk when compared to the existing index for industrial workers by around 11.5 percentage points. The weight of housing has been increased by around 7 percentage points. What this does is that it would keep understating the price increase, since the category of housing is unlikely to show monthly changes, as such revisions take place only periodically. But then there could always be discussion on weights to other components also such as transport, clothing etc. What is the solution?

To strike a balance, the government should focus on bringing out a food price index and simultaneously revealing how this moves. The problem with having a composite consumption index is that these weights differ for different segments of society. Just as rural people spend more on food, so do the low-income earners in urban areas, who probably do not spend on other items like housing or transportation.

Finally, one needs to ask as to why at all are we having one more index on prices. If it will replace other indices, then it is okay. If it is to run parallel, then it will add to the confusion as each index can tell a different story. Already there is an anomaly between CPI and WPI numbers and one is questioning as to which one is right. Now with this new index having a statistical bias towards showing better numbers, we will get another picture. The existence of three indices here will only add to the plethora of presentations of inflation numbers—we already have week on week, month on month, year on year, average till date for the WPI and various CPIs. The same will hold for these new three CPIs.

While more is merrier in general, the tricky question is, which one will the policy makers be looking at?