If we had included bank deposits, the growth rate would have been higher: TCA Anant

In a chat with ET Now, TCA Anant, Chief Statistician of India, says the directional impact of the trajectory could be also slightly sharpened when we go to into the full year.

Edited excerpts:

How would you explain the 7.1% estimated growth figure that you have come up with?

TCA Anant: The 7.1% growth is essentially a projection based on substantial data for the month of October which has been projected to the full year using the standard methods of projection which CSO has used all along.

Mythili Bhusnurmath: The 7.1% figure reflects a slowdown from the first half growth. Why is that even without taking into account the impact of demonetisation, there definitely is a decline in GDP growth as compared to the first half of the year. Which particular sector is it? Is it trade hotels and is there any specific reason that you can attribute for that decline over there?

TCA Anant: The trade hotels number is influenced by two sorts of indicators. To a limited extent, it is a projection of the first half using the corporate statistics because there is no further corporate statistics on that segment but it also reflects the sales tax collections figures which are available for October.

In other areas, it is partly influenced by the index of industrial production which also came in for October and that too came in less than what had been seen earlier. So these two are the sorts of numbers which should have gone into showing why the growth post September has come in lower. There is another element. The trajectory in the second quarter when it is extended to the full year also used the relationship between the first two quarters and the full year and so the directional impact of the trajectory could be also slightly sharpened when we go to into the full year.

Mythili Bhusnurmath: CSO’s estimates have been criticised very strongly by a lot of people who perhaps are not aware of the intricacies of GDP calculation and how you can only go purely by data. But the fact is that as far as financial sector is concerned, you have ignored the volatility in bank deposits. But given the fact that these deposits are likely to remain with the banking system because there are certain restrictions on withdrawal, could it possibly happen that when the year ends, we finally see the CSO’s advance estimates being realised despite the downward drag of demonetisation. But this number would be realised not because of any real increase in economic activity but purely because of the statistical blip caused by increase in deposits which again has been driven by demonetisation.

So can the GDP estimates take this finely nuanced view? Or is it your job just to calculate GDP and leave it at that?

TCA Anant: GDP itself is a statistical creation and to say what is real in it, becomes an extremely complicated task. Essentially, I can say with some degree of certainty is that had we included bank deposits, the growth rate would have been higher and it is likely that the growth rate in banking and financial sector when the full effect is worked out and if we move to a long-term trajectory where deposits remain high and credits also improve because of higher availability of funds in the banking sector, the contribution of banking sector to GVA could go up. We do not have enough of a time series to assess what the contribution of banking to GVA will be in this new normal at the moment. We may be able to do a little better job when more data comes in, let us wait for that data.

Mythili Bhusnurmath: Agreed entirely that you have a very tough job and GDP calculation is a very technical exercise but at a time when you are “forced” to come out with advance estimates based on just six months data as against normal nine months data, do you think it will be good for CSO as an institution to come out with the data because that is what you are required to as per government demand but add a number of caveats saying this is how this should be read, this is how it could be? Or should you leave it at that right now and subsequently if there is a huge difference between your advance estimates and your provisional and your revised estimates, then try to explain why those differences are there, would caveats have been order at this particular point of time?

TCA Anant: The difference in data between the January and February estimates is not as large as you make out. In the January estimate we have actually seven months data, in February we would essentially have eight months data, it is not that it changes a great deal. More numbers do become available in February but I am not sure that is such a big thing. In fact, we had done an exercise internally looking at the likely scenarios, had we done a forecast in January in previous years the estimates are not very different from what we would have released in the month of February.

That is principally about forecasting in January vis-à-vis February is concerned. What we do is give out details of the data available and the method of forecasting and we try to stay with the stable method of assessment in all our exercises.

Largely revisions are a matter of concern with every statistical office and in this respect we are not unique. In fact, many countries including in more developed countries and advanced countries the issue of very sharp revisions across different estimates has been noted, particularly when there is a considerable degree of turmoil in the underlying economy.

If you look at the estimates for OECD countries, if you look at the estimates of the US, all have seen official statisticians revising estimates sometimes significantly. But our overall confidence has been high despite all these changes in the overall scenario and the directional impacts have by and large been correct. So we are reasonably confident about the processes we follow but for the exact impact of structural changes, let us wait for more information to come in.

Mythili Bhusnurmath: Yes, absolutely I agree CSO has had a fairly commendable record so far but when you have these kind of black swan events as it were, would it help if CSO thought in terms of introducing more high frequency data in terms of monthly GDP, try to improve the kind of IIP data? Will those efforts help because I think increasingly now you are going to be asked to get advanced estimates in the month of January?

TCA Anant: I do not think monthly data is feasible because there is far too much issues about availability of numbers on monthly GDP but more high frequency data, absolutely. We should be getting a lot more high frequency data in different forms. We are working on this and we would like to continue working on it because it is independent of black swan events. Essentially, a lot more high frequency data will help us get a better picture of the economy. Some efforts are already there with our efforts in employment statistics. There are also efforts to make sure that data from early reporting of taxes can also be captured. These are exercises which are underway and I am hopeful will lead to more statistics being available for users in different ways.

Mythili Bhusnurmath: Finance Minister has come out with numbers for tax collections till December and these seem to suggest an economy is cheering on all fronts. Does this seem to be a contradiction to your numbers which would suggest an economy which is doing possibly not too well. Could you not have used that tax data would that have changed your calculation significantly?

TCA Anant: We do use tax data more than we used to use in the past and both sales tax and service tax figures are used in the new series much more extensively than were done in the old series. This is an ongoing exercise and we are certainly looking at it in future revisions as well. In general, the tax data reflects a variety of different considerations and its overall effect on GDP has to be factored in both by changes in the production structure, changes in collection effort.

We get much better information when we are able to match the tax data with accounting data which will become available later particularly from both government and company accounts. So let us take a look at all of those and we will get a much better picture of how the thing works but it is certainly not a negative set of data as far as we are concerned.

Mythili Bhusnurmath: Because it is not negative, it becomes that much more difficult to reconciles with your numbers. In fact, in other numbers that I was finding a little difficult to reconcile is the one that one looks at nominal GDP growth. Your nominal GDP growth for this year is 11.9% as against the budget estimate 11% and last year’s nominal growth of just 8%, given that the economy even as per your calculations has come down from 7.6 to 7.1 and inflation has definitely turned it down this year as compared to the previous year, how can the 11.9% growth in nominal GDP be explained?

TCA Anant: There arevtwo ways to think. Two facts which will help you put this number in perspective. First of all, the nominal GDP. First first look at the growth in nominal GVA. The difference between nominal GVA and GVA in constant prices is principally attributable to the changes in the inflation rate where the GVA deflator is weighted average of WPI and CPI. That will be pretty much along lines of what the WPI and CPI numbers have come in.

When we go from GVA to GDP, we come into this phenomena of net indirect taxes now in national accounts we have had convention on how we deal with net indirect taxes. The growth in nominal net indirect taxes is converted to a growth in real net indirect taxes by a relationship which is dependent on the growth in the real segment which is the GVA real growth. The rest of it is attributed in some sense to a price effect on the net indirect taxes front. Typically, when you talk about WPI and CPI deflators, you do not include the deflator on account of the tax front. The GDP deflator includes all of these and that is why some of the commentators have missed this point.

Mythili Bhusnurmath: Another little puzzle that you can help solve for us is in the case of gross fixed capital formation. This has been consistently coming down and yet the decline in GDP growth has not been as dramatic. How can an economy grow fairly well when your gross fixed capital formation which is really the basis for any kind of growth has been consistently coming down? Can you join the dots for us on that as well?

TCA Anant: The answer is simple. It also depends upon capacity utilisation and so on. Remember we had a very high investment boom before the financial crisis in 2007-2008 followed by a very sharp global slowdown. Under these circumstances, what is our status on capital utilisation? To what extent are we reaching constraints and therefore slow growth?

The factors leading to the slow growth in GFCF are also readily available and you can see that both in government account, company accounts as well as the IIP indicators but whether we are at the point where it will start constraining growth is not so easy to say because we do not really have a good sense of what the capacity utilisation structure looks like.