Some earlier elements of volatility and bias will be smoother: T C A Anant

Interview with Chief Statistician of India

Volatility of the Index of Industrial Production (IIP) in the earlier series was a concern. T C A Anant, chief statistician to the central government, discusses the issue and the new series fior it and the Wholesale Price Index (WPI), with Dilasha Seth & Indivjal Dhasmana. Edited excerpts:

The new series of the IIP is showing a brighter side of the economy on data for earlier years. Why?

A higher growth rate is partly a reflection of an updated base. This means the production structure is more representative and data are taken from more numbers of establishments than the old series. The earlier series suffered from a directional bias and that has been corrected. That was the problem when Gross Domestic Product (GDP) data was being interpreted on the basis of the old series of IIP. There will still be differences between GDP and IIP, due to inherent differences between value added and output.

Will the new series of IIP and WPI be incorporated in the GDP numbers for the fourth quarter and all of 2016-17?

Yes. 

Will the higher IIP numbers and lower WPI one in the new series as compared to the old one lead to a bump in GDP?

It is very difficult to answer this question because the IIP is used principally on the quasi-corporate side of manufacturing. How it will impact GDP will have to be worked out vis-a-vis the whole manufacturing sector. What it will do to growth rates depends on how it is incorporated in the GDP series. So, wait until the GDP numbers come out. 

Volatility of the IIP was an issue in old series. Is it eliminated in the new series?

There are two elements to volatility. One is natural volatility in production, as establishments might be affected due to climatic or social disturbances. If a part of the country is producing automobile parts and saw floods, production of these would be affected. That should be reflected in the index, as a natural change in output. 

Volatility should mean a disproportionate impact on the index due to weighing diagrams (or weighted diagrams, a reference to the weights allotted for various items) or because you have taken some and missed out many entities. The new series will eliminate the second lot. For example, an entity which was very important in producing an item is no longer important. It has shut down. But, many new entities have come in. As far as the index is concerned, that problem does not happen now. To ensure against recurrence, we have set up a technical committee. 

Capital goods output data was known to be too volatile in the old series. Will this change?

Capital goods, particularly those which take many months to be produced, used to have lumpiness in the production basket. That lumpiness will now be smooth, due to the changed method of measuring work in progress. In a way, it is a different index. We are not recording those items which have lumpiness at the time of completion but as a cumulation of work in progress. 

A technical committee has been set up to update the series. Will it not come in the way of comparability of data or will it be a chain-based index?

It is not yet a chain-based index. The index is constructed by weighing diagrams, formed at a three-digit level. This is the structure we will retain, for now. But, within three digits, we have considerable space and items will be updated. For instance, we have incorporated non-renewable energy, not from 2011-12 but from 2014-15 because the data started becoming available from that time. That principle will apply in the future as well. 

From the new series of IIP data, it seems manufacturing is still a concern, as the growth fell in March over February.

Whether manufacturing is still a concern or not is a separate issue. That will be explained better by the policy establishments concerned. 

The data also show that post-demonetisation blues are still faced in industrial production.

I caution you about taking demonetisation in purely a 'post hoc, ergo propter hoc' (after this and, so, because of this) basis, which is highly fallacious. James Tobin (Nobel Laureate) criticised economists for this approach. Demonetisation is a good example of not doing so. 

Why have indirect taxes have been excluded from the new series of WPI?

This is a conceptual issue. One of the major users of WPI is the Central Statistics Office, for the purposes of deflating various estimates. It has been a concern, that the prices we used have an element of taxes. We don't want to do so, in line with an international standard. Second, we would like India to instead replace the WPI, a somewhat obsolete index, with a Producers Price Index (PPI). One of the latter's elements is that the price part is taken separately from the tax part, as you want to know the pure-price part of the structure. This is not yet a PPI. 

But, the CPI will have indirect taxes.

That's because CPI is the price the final consumer pays.

Since the goods and services tax (GST) is being rolled out and it will be reflected in CPI, not WPI, will it not widen the differences between the two?

An index is about capturing what is happening at that particular point reflected by it. GST is going to be levied after all the netting at the final point of sale. You would like to see what is the difference between the price of an item produced and sold. To that extent, this is a correct way of compiling WPI and CPI.